households who reported receiving a SAP by wave 2. June 15 is the latest SAP receipt date reported in wave 2; all respondents who had not received a SAP in wave 2 but received one between waves 2 and 3 are included in this category (no dates were recorded in this survey wave).\\n\\n19\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099637201162481468/pdf/IDU17a9f43a1155191400d18bb91fe2719ee4d34.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"Table 4: Severe Depression by SAP Receipt Timing and Survey Wave\\n\\n||(1)|(2)|(3)|(4)|\\n|---|---|---|---|---|\\n||Severe<br>Depression|Severe<br>Depression|Severe<br>Depression|Severe<br>Depression|\\n|Wave=3|0.24***|0.23***|0.24***|0.24***|\\n||(0.037)|(0.038)|(0.039)|(0.038)|\\n|non-4Ps who received SAP before Apr 21|0.0027<br>(0.017)||||\\n|non-4Ps who received SAP between Apr 21-28|-0.016*||||\\n||(0.0079)||||\\n|non-4Ps who received SAP between Apr 29-May 9|0.016||||\\n||(0.034)||||\\n|non-4Ps who received SAP between May 10-Jun 15|-0.016*||||\\n||(0.0079)||||\\n|non-4Ps who received SAP after Jun 15|-0.016*||||\\n||(0.0079)||||\\n|non-4Ps who received no SAP|-0.0064||||\\n||(0.012)||||\\n|Wave=3 _×_ non-4Ps who received SAP before Apr 21|0.13*|0.13*|0.12*|0.13*|\\n||(0.068)|(0.069)|(0.072)|(0.069)|\\n|Wave=3 _×_ non-4Ps who received SAP between Apr 21-28|0.076|0.079|0.068|0.074|\\n||(0.075)|(0.076)|(0.077)|(0.075)|\\n|Wave=3 _×_ non-4Ps who received SAP between Apr 29-May 9|0.017|0.022|0.010|0.023|\\n||(0.075)|(0.071)|(0.070)|(0.072)|\\n|Wave=3 _×_ non-4Ps who received SAP between May 10-Jun 15|0.21*|0.21*|0.20|0.21*|\\n||(0.12)|(0.12)|(0.12)|(0.12)|\\n|Wave=3 _×_ non-4Ps who received SAP after Jun 15|0.31*|0.31*|0.41**|0.43**|\\n||(0.17)|(0.17)|(0.17)|(0.21)|\\n|Wave=3 _×_ non-4Ps who received no SAP|0.047|0.041|0.038|0.037|\\n||(0.087)|(0.089)|(0.090)|(0.090)|\\n|Observations|2906|2906|2906|858|\\n|Baseline Mean Outcome|0.01|0.01|0.01|0.02|\\n|Fixed Efects|None|Household|Household|Individual|\\n|Controls|None|None|Age/Gender|N/A|\\n|Sample|All|All|All|Panel<br>indiv.|\\n\\nNotes: Standard errors (clustered at the municipality level) are in parentheses * p _<_ 0 _._ 1, ** p _<_ 0 _._ 05, *** p _<_ 0 _._ 01.\\n\\n20\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099637201162481468/pdf/IDU17a9f43a1155191400d18bb91fe2719ee4d34.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"that cash transfer timeliness played a role in protecting the 4Ps from even larger increases in depression.\\n\\nAmong the subset of households eligible for a second SAP, 4Ps beneficiaries also received their second SAP transfer earlier than non-4Ps households. Given that only 20% of households in our sample received a second SAP, this is unlikely to be a major driver of the different depression trajectories across the two groups. Indeed, when we restrict our sample to households who only received one SAP (Appendix Table A7), results are almost identical to our baseline results. In short, it seems that the timely receipt of the first transfer played a more important role than that of the second transfer in generating smaller increases in depression for 4Ps households.\\n\\n# **5 Discussion**\\n\\nWe compare depression rates among beneficiaries of the existing 4Ps CCT program to those of non-beneficiaries, at the end of 2019 and in July 2020. At baseline (prior to the pandemic), depression rates were extremely low and similar for both groups. During the pandemic, depression rates increased dramatically for the entire sample, but less so for 4Ps beneficiaries. After ruling out alternative explanations, we interpret this as evidence that the 4Ps program became more important for mental health during a crisis. One possible reason is that the continued CCTs provided to 4Ps beneficiaries have larger benefits in times of vulnerability. Another explanation is that the 4Ps program enhanced the effectiveness of the government’s emergency cash transfer response, allowing 4Ps to receive these transfers more quickly than non-4Ps did. These results highlight that the existence of a well-functioning social protection program can yield benefits when emergencies arise, amplifying the benefits of any emergency policy response for those who are already part of the social protection system.\\n\\n21\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099637201162481468/pdf/IDU17a9f43a1155191400d18bb91fe2719ee4d34.pdf\", \"pages\": [22]}}]}"}}},{"rowIdx":76,"cells":{"doc_id":{"kind":"string","value":"34239706"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34239706\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"authors\": {\"0\": {\"author\": \"Wahby,Sarah Mohammad Osman\"}, \"1\": {\"author\": \"Assaad,Ragui A\"}}, \"count\": \"Syrian Arab Republic,Jordan\", \"count_key\": \"82691,82556\", \"docna\": {\"0\": {\"docna\": \"Job Finding and Separation among Syrian\\n refugees in Jordan and Their Hosts during the COVID-19 Pandemic\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"Office of the Chief Economist (MNACE)\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34239706\"}, \"repnb\": \"WPS10670\", \"docdt\": \"2024-01-16T00:00:00Z\", \"datestored\": \"2024-01-16T06:21:39Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"Refugees face important barriers to\\n participation in the formal market, which locks them in\\n informal employment and makes them more vulnerable to\\n shocks. Using data from Jordan, this paper compares the job\\n finding and separation rates of Syrian refugees to those of\\n their hosts before and after the onset of the COVID-19\\n pandemic. The findings show the change in these rates over\\n time for Syrians to be similar to those of their Jordanian\\n hosts prior to the pandemic, with a significant divergence\\n after the start of the pandemic. Distinguishing between\\n Syrians living in camps and those living in host communities\\n shows that the Syrian disadvantage was entirely explained by\\n living in camps.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10670; COVID-19 (Coronavirus); MNACE labor and gender\\n research programs; PEOPLE\", \"display_title\": \"Job Finding and Separation among\\n Syrian refugees in Jordan and Their Hosts during the\\n COVID-19 Pandemic\", \"disclosure_date\": \"2024-01-16T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099821301162430228/text/IDU16c8b4f4a1d5c9142801a62b166575607c40c.txt\", \"listing_relative_url\": \"/research/2024/01/34239706/job-finding-separation-among-syrian-refugees-jordan-hosts-during-covid-19-pandemic\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099821301162430228/IDU16c8b4f4a1d5c9142801a62b166575607c40c\", \"new_url\": \"2024/01/34239706/IDU16c8b4f4a1d5c9142801a62b166575607c40c\", \"ext_pub_date\": \"2024-01-16T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Labor Markets m1326488 855,Labor and Employment Laws and Regulations\\n m1327247 848,Refugees and Jobs m1504221 2894,Coronavirus (COVID-19) m1939478 4828\", \"docm_id\": \"IDU-6c8b4f4a-d5c9-4280-a62b-66575607c40c\", \"chronical_docm_id\": \"IDU-6c8b4f4a-d5c9-4280-a62b-66575607c40c\", \"publishtoextweb_dt\": \"2024-01-16T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10670\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"guid\": \"099821301162430228\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"49\", \"author_keys\": {\"0\": {\"author_key\": \"2251252\"}, \"1\": {\"author_key\": \"606784\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099821301162430228/IDU16c8b4f4a1d5c9142801a62b166575607c40c\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10670 Job Finding and Separation among Syrian refugees in Jordan and Their Hosts during the COVID-19 Pandemic _Sarah Wahby Ragui Assaad_ Middle East and North Africa Region Office of the Chief Economist January 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10670\\n\\n# **Abstract**\\n\\nRefugees face important barriers to participation in the formal market, which locks them in informal employment and makes them more vulnerable to shocks. Using data from Jordan, this paper compares the job finding and separation rates of Syrian refugees to those of their hosts before and after the onset of the COVID-19 pandemic. The findings show the change in these rates over time for Syrians to be similar to those of their Jordanian hosts prior to the pandemic, with a significant divergence after the start of the pandemic. Distinguishing between Syrians living in camps and those living in host communities shows that the Syrian disadvantage was entirely explained by living in camps.\\n\\nThis paper is a product of the Office of the Chief Economist, Middle East and North Africa Region. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at wahby001@umn.edu and assaad@umn.edu _The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [{\"mention_name\": {\"text\": \"data from Jordan\", \"confidence\": 0.8537031412124634, \"start\": 213, \"end\": 229, \"clean_text\": \"data from Jordan\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 213, \"end\": 229, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 213, \"end\": 229, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 213, \"end\": 229}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 213, \"end\": 229, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.312415212392807, \"start\": 213, \"end\": 229}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.8537031412124634, \"start\": 213, \"end\": 229}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5899358987808228, \"start\": 213, \"end\": 229}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5899358987808228, \"start\": 213, \"end\": 229}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# **Job Finding and Separation among Syrian refugees in Jordan and Their Hosts during the COVID-19 Pandemic**<sup>**1**</sup>\\n\\nSarah Wahby<sup>2</sup> and Ragui Assaad<sup>3</sup>\\n\\n**JEL codes:** J15, J61\\n\\n**Keywords:** refugees, job separation, job finding, camps, COVID-19\\n\\n**Topics:** Refugees and Jobs, Labor Markets, Labor and Employment Laws and Regulations, Coronavirus (COVID-19)\\n\\n## **Acknowledgements**\\n\\nThis work was supported by the MENA Chief Economist Office under the labor and gender research programs (TTLs: Nelly Elmallakh and Nazmul Chaudhury). We appreciate the comments of participants in the MNACE authors’ workshop, particularly our discussant Hai-Anh Dang.\\n\\n> 1 This paper is a product of the Office of the Chief Economist, Middle East and North Africa region. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp.\\n\\n> 2 Corresponding author. Humphrey School of Public Affairs, University of Minnesota. Email: wahby001@umn.edu\\n\\n> 3 Humphrey School of Public Affairs, University of Minnesota. Email: assaad@umn.edu\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **Introduction**\\n\\nRefugees, as economic actors, are not inherently different from members of host communities. Nevertheless, they face a different set of constraints and opportunities in economic participation (Betts et al., 2017b). Although socio-economic rights are officially granted to refugees by the 1951 Convention Relating to Refugees, they are often restricted by host countries. Host governments of developing countries, which host 86 percent of refugees (UNHCR, 2021a), are typically reluctant to grant such rights in fear of increased competition over limited resources with their own citizens.\\n\\nThe right to work is one prominent case of conflict between refugee rights and the practices of host governments. While this right is officially recognized in the 1951 convention, regulations in host countries range from denying the right to work altogether to restricting eligibility at varying levels (Zetter & Ruaudel, 2018). Facing such barriers to entry to the formal economy along with limited humanitarian support, refugees turn to informal employment characterized by lack of job security, social insurance, and lower wages. Refugees are, therefore, particularly vulnerable to shocks (Hoseini & Dideh, 2022).\\n\\nWe study the differential employment outcomes of refugees and the Jordanian host community before and during the COVID-19 pandemic, focusing in particular on job finding and separation rates. Syrian refugees in Jordan face substantial administrative and practical barriers to obtaining work permits. In practice, they are mostly confined to specific sectors, such as agriculture, construction and manufacturing, where employment is primarily informal (Gordon, 2019). Accordingly, over 90 percent of Syrian youth in Jordan work in regular and irregular informal jobs (Assaad et al., 2021b). Moreover, Syrian refugees in Jordan live primarily in urban host communities, with only 19 percent living in officially recognized refugee camps (UNHCR, 2022). Refugees living in host communities may experience more freedom of movement and lower 2\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"restrictions on job search, but they do not benefit from the same humanitarian assistance refugees in camps benefit from (Betts et al., 2017b). We investigate the determinants of the vulnerability of Syrian refugees to the COVID-19 shock in Jordan. In particular, we examine whether Syrian disadvantage during the shock is completely explained by institutional barriers reflected in precarious jobs and residency in camps, or if they experience added vulnerability by virtue of their refugee status.\\n\\nWe use retrospective data on the job histories of young Jordanian and Syrian men between the ages of 16 and 30 years from the Survey of Young People in Jordan (SYPJ) (Assaad et al., 2021a; OAMDI, 2022) to construct a synthetic semi-annual panel dataset that tracks the job finding and separation experiences of respondents. We compare the trends of job finding and separation across the two populations before and after the onset of the COVID-19 pandemic. We find that Syrians have generally experienced lower job finding rates and higher job separation rates compared to Jordanians, although the differences were generally not statistically significant prior to the pandemic. With the onset of the pandemic, Syrians became significantly disadvantaged on both measures. Controlling for the type of employment showed that workers whose last job was informal were disadvantaged on both measures compared to formal workers regardless of nationality. Controlling for the Syrian’s residency in camps, however, showed that the Syrian disadvantage was completely driven by those living in camps. In fact, holding a work permit, contrary to expectations, was correlated with significantly lower job finding rates during the pandemic particularly because camp residents are more likely to hold a work permit. The negative effect of permits disappeared when we controlled for camp residence. Moreover, controlling for informal employment, we did not find significant difference in the experiences of refugees residing in host communities and those of Jordanians. These results are in line with previous research 3\", \"datasets\": [{\"mention_name\": {\"text\": \"Survey of Young People in Jordan\", \"confidence\": 0.9299076795578003, \"start\": 627, \"end\": 659, \"clean_text\": \"Survey of Young People in Jordan\"}, \"acronym\": {\"text\": \"SYPJ\", \"confidence\": 0.6968355774879456, \"start\": 661, \"end\": 665, \"clean_text\": \"SYPJ\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 627, \"end\": 659, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 627, \"end\": 659}, \"geography\": {\"text\": \"Jordan\", \"confidence\": 0.5687822103500366, \"start\": 241, \"end\": 247, \"clean_text\": \"Jordan\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.98172527551651, \"start\": 627, \"end\": 659}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.9299076795578003, \"start\": 627, \"end\": 659}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9356135129928589, \"start\": 627, \"end\": 659}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9356135129928589, \"start\": 627, \"end\": 659}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"comparing Syrian refugees in Türkiye to their hosts, which showed the native-refugee gap to be much lower than previously reported for developed countries, when accounting for the fact that Syrians in Türkiye, similar to those in Jordan, are primarily employed in the informal sector (Demirci & Güray Kırdar, 2021). We found Syrians living in camps, however, to be significantly disadvantaged compared to both Jordanians and Syrians not living in camps during the pandemic. Recent papers have specifically looked at Syrian refugees’ outcomes across living arrangements in Jordan prior to the pandemic. They found Syrians outside camps to be less likely to live under the national abject poverty line, possess more assets and report higher satisfaction with access to services and with life in general (Obi, 2021). Nonetheless, using the same data, (Ginn, 2018) found that the loss of income that camp residents face is more than offset by what they save in rent. He also found no difference in growth of employment between the two populations.\\n\\nThis paper contributes to two strands of literature. First, it contributes to the literature on the impact of the COVID-19 pandemic on the labor market outcomes of different demographic groups. Shifts in demand during the pandemic away from face-to-face services such as hospitality and retail sectors meant that the pandemic primarily affected those who cannot perform their jobs remotely (Abay et al., 2020). Evidence suggests that these are typically workers in low paying jobs (Bartik et al., 2020) who are less educated, younger, disproportionately female, and are more likely to belong to minority and immigrant groups (Angelucci et al., 2020; Borjas & Cassidy, 2020). In short, originally disadvantaged groups bore the brunt of the pandemic (Cortes & Forsythe, 2020). In developing countries, those who could perform their jobs remotely were essentially more educated workers (Gottlieb et al., 2020). In addition, social protection networks are less prevalent in these countries, exacerbating the vulnerability of workers who lost their jobs (Garrote Sanchez et al., 2021). A similar study to ours compared the outcomes of refugees in Kenya to those of 4\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"Kenyan nationals and found that the former group were harder hit by the pandemic and that their recovery has been slower. The gap between those two populations was not explained by demographic characteristics, residency in camps, previous employment or sector of employment and is likely the result of an “unobservable refugee factor” (Vintar et al., 2022). Evidence from the Middle East and North Africa (MENA) region showed that pre-pandemic employment was a strong predictor of workers’ ability to preserve their jobs during the pandemic and recover afterwards. In particular, workers in the formal sector, public and private, experienced greater stability compared to workers in informal employment, especially those in irregular jobs not attached to a fixed establishment (Krafft et al., 2022). Second, more broadly, this paper contributes to the nascent literature on refugee economies. This literature stresses the importance of constructing a separate theoretical framework and empirical evidence for the study of forced migration given the different choices and constraints that forced migrants face compared to other types of migrants (Betts et al., 2017a; Ruiz & Vargas-Silva, 2013; Verme, 2017; Verme & Schuettler, 2021). Evidence from this literature suggests that refugees are less likely to be employed and earn less than other migrants and hosts (Ruiz & Vargas-Silva, 2017). Nonetheless, the native-refugee gap is much smaller in developing countries compared to developed countries especially after controlling for type of employment (Demirci and Güray Kırdar 2021).\\n\\nThe paper underlines the necessity of translating theoretical refugee rights stipulated in international conventions and treaties into practical regulations in host countries, especially the right to work. The right to work is essential for refugee self-reliance, contribution to the economy of the host country and resilience in the face of shocks. With a majority of refugees hosted in poorer countries with a limited capacity to create jobs and to provide social protection, this highlights the importance of increased cooperation and shared responsibility from the international 5\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"community in managing refugee crises and securing refugee rights. Similarly, while we cannot causally ascribe our results to residency in camps since residential selection into camps vs. host community is endogenous, our results for Jordan underscore the importance of revisiting camp settings and the policies governing them. In particular, the confinement of refugees to camps and their strict closures during the pandemic, while justifiable on health grounds, appear to have resulted in adverse socio-economic outcomes for camp residents.\\n\\n# **Context**\\n\\nAs of May 2022, Jordan hosted 674,000 Syrian refugees as per UNHCR registration records (UNHCR, 2022). The actual number of Syrians in Jordan could exceed 1.3 million when accounting for the unregistered and those who do not consider themselves refugees (Krafft et al., 2019). That is more than one-tenth of the Jordanian population. The bulk of Syrian refugees arrived between 2012 and 2014, with the numbers stabilizing as of mid 2016 (Alhawarin et al., 2021). In the SYPJ sample, 95% of Syrians have been in Jordan for at least 5 years. Syrians in Jordan are mainly recognizable by their different accent.\\n\\nDespite being the second largest per capita refugee host in the world (UNHCR, 2021b), Jordan does not have a legal framework on refugees. The country is not a signatory of the 1951 Convention Relating to Refugees and the subsequent 1967 protocol and does not have clear domestic refugee legislation. The only framework managing refugees’ issues is the memorandum of understanding the Jordanian government signed with UNHCR in 1998. Nevertheless, in line with customary law, Jordanian law explicitly prohibits refoulement<sup>4</sup> and allows exemptions on the requirements for entry and residency for political asylum seekers (ILO, 2015). Accordingly, Jordan\\n\\n> 4 The forcible return of refugees or asylum seekers to a country where they are liable to be subjected to persecution.\\n\\n6\", \"datasets\": [{\"mention_name\": {\"text\": \"UNHCR registration records\", \"confidence\": 0.7687290906906128, \"start\": 619, \"end\": 645, \"clean_text\": \"UNHCR registration records\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.4749488830566406, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"UNHCR\", \"confidence\": 0.3338903486728668, \"start\": 619, \"end\": 624, \"clean_text\": \"UNHCR\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 619, \"end\": 645}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 619, \"end\": 645, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.3170991837978363, \"start\": 619, \"end\": 645}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7687290906906128, \"start\": 619, \"end\": 645}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.788287341594696, \"start\": 619, \"end\": 645}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.788287341594696, \"start\": 619, \"end\": 645}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"has generally maintained an open border policy toward refugees reflected in its tolerance toward Syrian refugees at the onset of the Syrian crises in 2011 (Beaujouan & Rasheed, 2020).\\n\\nIn addition to the lack of clear legislation regarding refugees, there is no clear policy concerning their right to work in the Jordanian territory. More generally the Jordanian constitution stipulates that the right to work is reserved for Jordanian citizens. In practice, the same regulations organizing the work of non-Jordanian migrants apply to refugees. A work permit approved by the ministry of labor is required. Such a permit is conditional on proving that the job in question requires experience or skills either unavailable or not sufficiently available among Jordanians. Accordingly, Syrian refugees are allowed to work in specific sectors that typically hire migrant workers, which include construction, agriculture, manufacturing, and a mix of low skill services (Gordon, 2019; ILO, 2015).\\n\\nIn 2016, the Jordan Compact<sup>5</sup> relaxed many of the restrictions related to work permits, leading to a substantial increase in the number of work permits issued. Nevertheless, those were mostly issued in the construction and agriculture sectors, i.e., sectors in which Syrians were already working without permits before the Jordan Compact. This raises questions around the extent to which the Compact has resulted in job creation versus legalizing already existing work situations (Beaujouan & Rasheed, 2020). Moreover, given the predominantly informal nature of these sectors, obtaining work permits seemed to translate into the formalization of the status of the worker and not necessarily the formalization of their work (Gordon, 2019). Syrians, also, sometimes voluntary abstain from getting work permits because of fear of losing access to humanitarian assistance. Given the insecure nature of jobs available to them, they do not want to\\n\\n> 5 An agreement with the European Union whereby the Government of Jordan received large funds and economic benefits in return for further integration of the Syrian refugees in the labor market. In the framework of the Jordan Compact the Government of Jordan pledged to provide 200,000 job opportunities for Syrian refugees.\\n\\n7\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"run the risk of losing both their job and the humanitarian assistance they receive, and end up without any source of livelihood (Razzaz, 2017).\\n\\nThe Jordanian economy’s capacity to create jobs has been limited. Growth has slowed down for more than a decade under the effect of external shocks beginning with the financial crises in 2008 and followed by the general instability in the region with the onset of Arab Spring uprisings in 2010, which culminated in the influx of Syrian refugees starting in 2012. Growth reached the decade’s lowest plateau of 2 percent between 2016 and 2018. In response to the stagnant growth, unemployment increased, reaching 19 percent in 2019, and labor force participation continued to fall, reaching 34.5 percent in the same year. This ranks labor market participation in Jordan among the lowest in the world (World Bank, 2020). Over the same period, the share of informal employment has increased, a reflection of the increasing share of non-national employment. Before the influx of Syrian refugees, the Jordanian labor market had a segmented structure, which for the most part allocates Jordanians to formal jobs and non-Jordanians to informal ones. In 2016, non-Jordanians made up 31 percent of employment, among whom only one-fifth were Syrians. In the same year, while 77 percent of non-Jordanians worked in informal jobs only 32 percent of Jordanians did (Assaad & Salemi, 2019).\\n\\nAn already difficult situation became even more challenging during the COVID-19 pandemic. Among Jordanian youth 13 percent were separated from their jobs during 2020, compared to 5 percent during 2016. Perhaps even more strikingly, over 31 percent of Syrian youth experienced job separation during 2020 compared to only 2 percent in 2016 (Assaad et al., 2021b). Such important discrepancies in job separation rates between Jordanians and Syrians are probably the result of the government’s prohibition of layoffs during the pandemic, which is only pertinent to 8\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"formal employment.<sup>6</sup> In fact, among the five MENA countries with available data on labor market outcomes during the pandemic, Jordan exhibited the greatest stability in employment over time (Krafft et al., 2022). Moreover, Jordan undertook strict closure policies especially for camps whereby entering and exiting refugee camps was prohibited, including for humanitarian workers providing assistance to refugees.\\n\\n# **Data and descriptive statistics**\\n\\nTo analyze the trends of employment across Jordanians and Syrians before and during the COVID19 pandemic we would ideally need panel data that covers this period. Such data does not exist. The Survey of Young People in Jordan (SYPJ) 2020 (Assaad et al., 2021a; OAMDI, 2022) offers a suitable alternative. Collected between August 2020 and October 2020<sup>7</sup> from a nationally representative sample of Jordanian and Syrian youth between 16 and 30 years old, the survey captures the labor market experience of this population before and during the pandemic through retrospective questions about their job history since they entered the labor market. We use data on the characteristics of the individual’s current job as well as data on all their previous jobs since they entered the labor market. Individuals can report as many jobs as they took with no upper limit. The maximum number of jobs reported in our sample was seven jobs.\\n\\nThe other source of micro data for Jordan during the pandemic is the COVID-19 MENA Monitor (CMM) data (OAMDI, 2021). This data was collected over three waves in February, June and August 2021 through a phone survey. It also has retrospective data on individuals’ employment\\n\\n> 6 Defence order number 6 stipulated, early April, that institutions subject to the labor law must allow their employees who were dismissed or whose services were terminated since the beginning of the pandemic to return to their work < https://www.jordantimes.com/news/local/pm-issues-defence-order-no-6-stipulating-labour-rightsunder-defence-law>.\\n\\n> 7 Additional data was collected between February and March 2021 from refugee camps.\\n\\n9\", \"datasets\": [{\"mention_name\": {\"text\": \"panel data\", \"confidence\": 0.5007688403129578, \"start\": 590, \"end\": 600, \"clean_text\": \"panel data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 590, \"end\": 600, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 590, \"end\": 600, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 590, \"end\": 600}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 590, \"end\": 600, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.48174044489860535, \"start\": 590, \"end\": 600}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5007688403129578, \"start\": 590, \"end\": 600}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.546370267868042, \"start\": 590, \"end\": 600}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.546370267868042, \"start\": 590, \"end\": 600}}, {\"mention_name\": {\"text\": \"Survey of Young People in Jordan\", \"confidence\": 0.8840844631195068, \"start\": 656, \"end\": 688, \"clean_text\": \"Survey of Young People in Jordan\"}, \"acronym\": {\"text\": \"SYPJ\", \"confidence\": 0.5915729403495789, \"start\": 690, \"end\": 694, \"clean_text\": \"SYPJ\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 656, \"end\": 688, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 656, \"end\": 688}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 656, \"end\": 688, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9982598423957825, \"start\": 656, \"end\": 688}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8840844631195068, \"start\": 656, \"end\": 688}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8069441914558411, \"start\": 656, \"end\": 688}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8069441914558411, \"start\": 656, \"end\": 688}}, {\"mention_name\": {\"text\": \"COVID-19 MENA Monitor\", \"confidence\": 0.5578464865684509, \"start\": 1470, \"end\": 1491, \"clean_text\": \"COVID-19 MENA Monitor\"}, \"acronym\": {\"text\": \"CMM\", \"confidence\": 0.7402021884918213, \"start\": 1493, \"end\": 1496, \"clean_text\": \"CMM\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1470, \"end\": 1491, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1470, \"end\": 1491}, \"geography\": {\"text\": \"Jordan\", \"confidence\": 0.5151874423027039, \"start\": 1436, \"end\": 1442, \"clean_text\": \"Jordan\"}, \"typology_tag\": {\"text\": \"microdata\", \"confidence\": 0.9999289512634277, \"start\": 672, \"end\": 693}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5578464865684509, \"start\": 672, \"end\": 693}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.8412466049194336, \"start\": 672, \"end\": 693}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8412466049194336, \"start\": 672, \"end\": 693}}, {\"mention_name\": {\"text\": \"retrospective data on individuals’ employment\", \"confidence\": 0.5366873145103455, \"start\": 1629, \"end\": 1674, \"clean_text\": \"retrospective data on individuals’ employment\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1629, \"end\": 1674, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1629, \"end\": 1674, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1629, \"end\": 1674}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1629, \"end\": 1674, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.5523881316184998, \"start\": 52, \"end\": 97}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.5366873145103455, \"start\": 52, \"end\": 97}, \"usage_context\": {\"text\": \"background\", \"confidence\": 0.411729097366333, \"start\": 52, \"end\": 97}, \"is_used\": {\"text\": \"False\", \"confidence\": 0.411729097366333, \"start\": 52, \"end\": 97}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"statuses in February 2020, right before the pandemic. We deemed the SYPJ data to be more suited to our question since it allows to look at the trend of employment outcomes before the pandemic as well as right after the shock in the summer of 2020 and therefore allows investigating the effect of the pandemic on the two populations. The CMM data would be better suited to capture the recovery from the pandemic (Krafft et al., Forthcoming). The overlapping portion of the CMM data with the SYPJ data (February 2020 to February 2021) did not allow for a comparable analysis to the one we did in this paper. The overlap is not perfect. CMM data was collected a few months after the bulk of the SYPJ data only allowing for a comparison between the first half of 2021 (CMM) to the second half of 2020 rates (SYPJ). This is not a reasonable comparison especially in a period with rapidly changing events like the pandemic time. In addition, the time intervals are different as CMM data allows for examining annual change (February 2020 to February 2021) as opposed to semi-annual, and that is only one period which does not allow to compare the trend.\\n\\nIn our analysis we focus on young males aged 15 and above who ever worked. Since those who never worked do not have job finding and job separation outcomes we believe including them in the sample will not add much information. Moreover, given low participation rates among women for both populations, the number of women who ever worked in the SYPJ sample is only 229 (75 of whom are Syrians). It is therefore difficult to conduct the analysis by sex. Given the difference in the experiences of women and men in the labor market, we limit the sample to male respondents for whom we have information on their first job.<sup>8</sup> Furthermore, since our analysis spans the period from 2016 to 2020, we limit the sample to those who were 15 or above in 2016. This results in a sample of 824 young males, among whom 314 are Syrians.\\n\\n> 8 We show the results of the full sample, men and women, in the appendix.\\n\\n10\", \"datasets\": [{\"mention_name\": {\"text\": \"SYPJ data\", \"confidence\": 0.6102789044380188, \"start\": 68, \"end\": 77, \"clean_text\": \"SYPJ data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.3050912916660309, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 68, \"end\": 77, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 68, \"end\": 77}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 68, \"end\": 77, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.46896952390670776, \"start\": 68, \"end\": 77}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6102789044380188, \"start\": 68, \"end\": 77}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8433218002319336, \"start\": 68, \"end\": 77}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8433218002319336, \"start\": 68, \"end\": 77}}, {\"mention_name\": {\"text\": \"CMM data\", \"confidence\": 0.4977554678916931, \"start\": 337, \"end\": 345, \"clean_text\": \"CMM data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 337, \"end\": 345, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 337, \"end\": 345, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 337, \"end\": 345}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 337, \"end\": 345, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.5949283838272095, \"start\": 337, \"end\": 345}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4977554678916931, \"start\": 337, \"end\": 345}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.74580317735672, \"start\": 337, \"end\": 345}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.74580317735672, \"start\": 337, \"end\": 345}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"<!-- Start of picture text --> o Jordanian o Syrian<br>oO oO<br>| |<br>oe en<br>te) I oak<br>ar ee eee te) | -<br>2 --\\\" —r—S __ re am2 | l- =.\\\" 7<br>= - wane! | = ta ee<br>se}Oo -14°0 --.!Li — se}Oo forcmy | _—<br>= ___l = | -—!<br>= ~! 5 a<br>(e)® _ ': (S)® _<br>oD a 5}<br>Ow ow —<br>N ' i ee<br>fo) fo)<br>0 2 4 6 8 10 0 2 4 6 8 10<br>Years since school exit (or age 15) Years since school exit (or age 15)<br>----- Less than Basic --------- Basic<br>—:'-—<br>— - Secondary —-—--— Higher education<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"As Table 1 suggests, Jordanians and Syrians have significantly different characteristics. Jordanians are significantly more educated by an average of three more years of schooling. Jordanians who have ever worked are therefore, on average, older than their Syrian counterparts, with fewer years since their first job. They also have a significantly lower total number of jobs. Moreover, considering the aforementioned segmentation of the market, while 40% of Jordanian youth have informal wage work, Syrians are almost entirely in informal wage employment.\\n\\n_Table 1: Sample characteristics of Jordanians and Syrians_\\n\\n||Jordanian|Syrian|Difference||\\n|---|---|---|---|---|\\n|Age|23.99|23.65|-0.34||\\n||{3.2}|{3.0}|(0.25)||\\n|Years of schooling|11.89|8.65|-3.24|***|\\n||{2.7}|{3.2}|(0.18)||\\n|Years since first job|4.50|5.08|0.58|**|\\n||{3.8}|{4.1}|(0.24)||\\n|Type of employment for last job:<br>informal wage|<br>0.41|0.94|0.53|***|\\n||{0.4}|{0.2}|(0.03)||\\n|Total number of jobs taken|1.25|1.41|0.16|***|\\n||{0.7}|{0.9}|(0.03)||\\n|N|510|314|824||\\n\\nEach row is a regression of the dependent variable on binary variables for each nationality. Standard deviations are reported in brackets and standard errors are reported in parentheses. Significance * .10; ** .05; *** .01.\\n\\n_Source: Authors’ calculations using SYPJ data_ The SYPJ has two main limitations. First, since the sample is restricted to youth it gives no information about older working populations. Second, retrospective data on job histories 12\", \"datasets\": [{\"mention_name\": {\"text\": \"retrospective data\", \"confidence\": 0.3983204960823059, \"start\": 1458, \"end\": 1476, \"clean_text\": \"retrospective data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1458, \"end\": 1476, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1458, \"end\": 1476, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1458, \"end\": 1476}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1458, \"end\": 1476, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.4095190167427063, \"start\": 492, \"end\": 510}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3983204960823059, \"start\": 492, \"end\": 510}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.37032005190849304, \"start\": 492, \"end\": 510}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.37032005190849304, \"start\": 492, \"end\": 510}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"potentially involve long recall periods, which could result in recall bias. This would affect accuracy of information on some job characteristics. Evidence suggests that we can draw fairly accurate information on labor market dynamics from retrospective data. More elaborate information about job characteristics such as fine distinctions between different employment statuses are less reliable (Assaad et al., 2018). In our analysis, we focus on transitions in and out of different jobs, so we have reasons to believe that the data is reliable for this purpose. Recall of job start and end dates might be trickier. This is evident in the important number of respondents who could not recall the exact month or year when they started or ended a job. Since information on starting and ending dates of a job are essential for constructing the synthetic panel, we excluded from our dataset 19 individuals who could not recall the year where they started or ended any of their jobs. It is, however, easier for respondents to recall the year they started or ended a job rather than the specific month. For example, one third of the individuals who reported the starting year of their first job had a missing starting month. To avoid losing those who did not recall the specific month, we assigned a random number between 1 to 12 when the month was missing. In doing so, we considered the following: a) Respondents who ended and started two different jobs in the same year and reported the month of ending the former job are assigned a starting month for the latter job that is equal to or greater than the month when they ended their former job.\\n\\nb) Respondents who ended and started two different jobs in the same year and reported the month of starting the latter job are assigned an ending month for their former job that is equal to or less than the month where they started their latter job c) Since the survey ended before the end of 2020, we make sure that in 2020 the month assigned is less than or equal to the month where the individual took the survey.\\n\\n13\", \"datasets\": [{\"mention_name\": {\"text\": \"retrospective data\", \"confidence\": 0.9337791800498962, \"start\": 240, \"end\": 258, \"clean_text\": \"retrospective data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 240, \"end\": 258, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 240, \"end\": 258, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 240, \"end\": 258}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 240, \"end\": 258, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6146049499511719, \"start\": 240, \"end\": 258}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9337791800498962, \"start\": 240, \"end\": 258}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8240013122558594, \"start\": 240, \"end\": 258}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8240013122558594, \"start\": 240, \"end\": 258}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"Bearing in mind that this data is then translated to semi-annual data, i.e. the specific month is not used, among those who reported the starting month of their different jobs, the probability of starting a job in a given half-year was approximately 50 percent. We, therefore, think that a random assignment that gives equal weight to all months (and to each of the two half years) is a sensible choice.\\n\\nWe construct a semi-annual synthetic panel. In doing so, we assume that each individual in the sample is observed over the period from the first half year of 2016 to the second half year of 2020. Individuals who entered the job market later than the first half of 2016 are considered nonemployed until entry. Using the starting and ending dates for all the jobs an individual has engaged in, we can identify the individual’s job finding and job separation events. In particular, we convert the starting and ending month of a job into the half year they belong to. For example, a person who started a job in May of 2017 is coded as found a job in the first half year of 2017.\\n\\nIn Table 2, we present the transitions by half-year, nationality of individual, their camp residency if they are Syrian and the imputation status. Overall, 27% of the transitions that occurred had an imputed month. Two things to keep in mind when considering this table. This does not show all the respondents-time periods in our sample (8,240 observations). Since imputation will only occur when there is a transition in or out of a job, we are only including in the table those transitions and not all the periods where the individual is staying in a certain job. The total number of transitions is 726 which is less than the total number of individuals in the sample. That is because for some individuals all transitions occur prior to 2016 so they are not captured in our synthetic panel.\\n\\n_Table 2 Transitions in and out of jobs by half, nationality, camp residency and imputation status of the month where the transition occurred_ <u>Syrian</u> Half <u>Jordanian camp=0 camp=1</u> 14\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"||imputed=0|imputed=1|Total|imputed=0|imputed=1|Total|imputed=0|imputed=1|Total|\\n|---|---|---|---|---|---|---|---|---|---|\\n|2016h1|22|16|38|8|1|9|10|11|21|\\n|2016h2|21|15|36|1|3|4|8|5|13|\\n|2017h1|23|7|30|6|4|10|8|5|13|\\n|2017h2|28|16|44|5|4|9|2|5|7|\\n|2018h1|30|17|47|7|3|10|13|7|20|\\n|2018h2|27|11|38|5|6|11|8|5|13|\\n|2019h1|39|6|45|7|4|11|10|4|14|\\n|2019h2|29|17|46|6|2|8|14|2|16|\\n|2020h1|67|7|74|24|6|30|18|2|20|\\n|2020h2|47|3|50|12||12|23|4|27|\\n|Total|333|115|448|81|33|114|114|50|164|\\n\\nTable 3 in the appendix, investigating the differences between individuals by their ability to recall all months of start and end of jobs they have taken, confirms that recall is systematic and associated to significantly different characteristics of respondents. Respondents who recalled the month are significantly more educated, younger and less likely to be informally employed. Systematic recall, however, does not necessarily contradict that the half year where the individual finds or loses a job is random and therefore does not invalidate our imputation strategy. In fact, systematic differences by recall ability means that not accounting for those who did not recall the month of start or end of their job will undermine the representativeness of our sample and omit important conclusions concerning specific groups in the population.\\n\\nFollowing Borjas & Cassidy (2020) we decompose the change in employment rate from one half year to another into its two main components, the job separation rate and the job finding rate.\\n\\nWhere 15\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"_: the number of persons employed at time t_ tt _: the total population_ EE _: the number of persons who were not employed at time t-1 but found a job by time t_ PP tt _: the number of persons who were employed at time t-1 but lost their job by time t_ FF tt _: the number of persons who are not employed at time t-1_ LL tt−1 _: the employment rate at time t-1_ NN tt−1 _: the fraction of persons out of work who found a job by time t i.e., the job finding rate_ ππ tt _: the fraction of employed persons who are not working by time t i.e., the job separation rate_ ff tt ff If an individual had a start date for a job in a certain half year, he is considered to have found a job in this half year. Similarly, if an individual had an end date for a job in a certain half year, he is considered to have separated from a job in this half year. Moreover, if an individual started a job and ended it within the same half year, he is considered to have both found and separated from a job in this particular half year. Using this procedure, we can estimate the rate of job finding in a certain half year, _ft_ , and the rate of job separation in that half year, _lt_ .\\n\\nFigures 2 to 4 present the descriptive statistics of the change in employment rate and its components, job separation and job finding rates, for Jordanian and Syrian males. Data in these figures is weighted to be nationally representative of Jordanian and Syrian male youth in Jordan. The data was also seasonally adjusted by subtracting from the rates for the second half-years the difference in means between the second half-year and the first half-year.\\n\\nJob separation rates generally coincide across nationality except that Jordanians have lower rates during the pandemic compared to Syrians. The divergence of outcomes is clearer in job finding rates where Jordanians have evidently higher rates during the pandemic. Similarly, when we condition on informal employment, we see overlapping trends for Jordanians and Syrians across 16\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"<!-- Start of picture text --> ao- Job separation so- Job finding so- Change in employment<br>50 + 50 + 50+<br>. .<br>40 40+ 4o-<br>5 5 AS<br>pm a eoa207 6a404<br>a a<br>204 20-4 Nd 20-4<br>. iTM «<br>105 10 i; .<br>.<br>ry<br>a ee 032 —<br>PrCn a half-yearar a oFa a FSShalf-year oFCn a SFhalf-yeareeSS<br>——@®—- Jordanian = —#®—\\\\ Syrian<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"<!-- Start of picture text --> so- Job Separation ao- Job finding ao- Change in employment<br>50 + 504 50<br>o .<br>40+ 40- 40+<br>55E<br>205 2 i> 0m<br>a aa<br>a a<br>20+ 20+ bal 204<br>.<br>* *<br>1-7 10 * oF<br>' * . ° .<br>“6 . . x) ee .<br>SK SF SS FF FS Pr ee eo<br>half-year half-year half-year<br>—-—@®—- Jordanian = —#®—\\\\ Syrian<br>so- Job Separation so- Job finding eo- ‘Change in employment<br>504 504 aay<br>40- .<br>40- 40- || @<br>F F a .<br>ey ey * e .<br>CT a a<br>a a C a 204 e<br>*<br>20 20 * e<br>* 104 *<br>ee<br>w- wo4<br>s ; e o-<br>Ce | s<br>FF K SF FF K & FF FS<br>half-year half-year half-year<br>——®—- Non-Camp —®#— Camp<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"# **Empirical strategy**\\n\\nUsing the synthetic semi-annual panel data we constructed, we estimate several models for each of the two outcomes: likelihood of being separated from a job from one period to another and the likelihood of finding a job. In all models we include individual and half year fixed effects. Adding individual fixed effects absorbs all time-invariant differences across individuals, including average differences between groups of workers, such as Syrians and Jordanians. We are thus measuring differences across these groups over time as they deviate from the group average. If we do not see significant differences, this does not mean that there are no differences between the two populations, rather that there are no differences in trends over time.\\n\\nThe first model investigates the relationship between Syrian nationality and the likelihood of finding (separating from) a job in each half year from the beginning of 2016 to the end of 2020. (1) iitt ii tt 1tt ii tt iitt yy = αα + γγ + ββ SS ∗HH + εε _Where_ : _equal 1 if individual i finds (separates from) a job between half-year t-1 and half-year t, 0 otherwise_ iitt yy : _Syrian indicator_ ii : _Half year dummy_ SS HHtt: _half year fixed effects_ tt: _individual fixed effects_ γγ ii : _error term_ αα iitt In the second model we use the same specification but, in addition, we control for the type of εε employment of the individual’s last job, that is their current job if they are currently employed and their last job if they are currently non-employed. Workers do not transition frequently between informal and formal jobs. In our sample only 80 individuals have changed their employment type 19\", \"datasets\": [{\"mention_name\": {\"text\": \"synthetic semi-annual panel data\", \"confidence\": 0.6409096121788025, \"start\": 36, \"end\": 68, \"clean_text\": \"synthetic semi-annual panel data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 36, \"end\": 68, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 36, \"end\": 68, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 36, \"end\": 68}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 36, \"end\": 68, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9691480398178101, \"start\": 36, \"end\": 68}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6409096121788025, \"start\": 36, \"end\": 68}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6998140215873718, \"start\": 36, \"end\": 68}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6998140215873718, \"start\": 36, \"end\": 68}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"at lease once since they took their first job. The formality status is therefore generally stable and equivalent to the formality of their last job. Moreover, since hardly any Syrians have formal jobs, this is equivalent to comparing each group of informal workers to formal Jordanian workers. Nonwage workers are included as a separate category in the regression, but we do not show their results given their limited sample size.\\n\\nS (2) iitt ii tt 1tt ii tt 2tt ii tt iitt yy = αα + γγ + ββ ∗HH + ββ JJSS ∗HH + εε S : _dummy for Syrian whose last job was private informal, 0 otherwise._ ii _: dummy for Jordanian whose last job was private informal, 0 otherwise._ ii JJSS In the third, fourth and fifth models we limit the sample to Syrians. We use an in-camp indicator to test the effect of living in camp among Syrians (3), a permit indicator for examining the effect of holding a work permit (4), we then account for both variables in the same model (5). (3) iitt ii tt 1tt ii tt iitt yy = αα + γγ + ββ CC ∗HH + εε : takes 1 if Syrian living in camp, omitted category is Syrian not living in camp ii CC (4) iitt ii tt 1tt ii tt iitt yy = αα + γγ + ββ PP ∗HH + εε : takes 1 if Syrian holding work permit, omitted category is Syrian not holding permit ii PP (5) iitt ii tt 1tt ii tt 2tt ii tt iitt yy = αα + γγ + ββ CC ∗HH + ββ PP ∗HH + εε Finally, using the full sample we include two dummy variables for being a Syrian who lives in camp or being a Syrian who does not live in camp to have a closer look at these categories in comparison to Jordanians. In this model we also control for formality of the last job. (6) iitt ii tt 1tt ii tt 2tt ii tt 3tt ii tt iitt yy = αα + γγ + ββ SSCC ∗HH + ββ SSnnCC ∗HH +ββ EE ∗HH + εε 20\", \"datasets\": [{\"mention_name\": {\"text\": \"in-camp indicator\", \"confidence\": 0.3159550726413727, \"start\": 753, \"end\": 770, \"clean_text\": \"in-camp indicator\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.3432369530200958, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 753, \"end\": 770, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 753, \"end\": 770}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 753, \"end\": 770, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9921531081199646, \"start\": 753, \"end\": 770}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3159550726413727, \"start\": 753, \"end\": 770}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8860774040222168, \"start\": 753, \"end\": 770}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8860774040222168, \"start\": 753, \"end\": 770}}, {\"mention_name\": {\"text\": \"permit indicator\", \"confidence\": 0.3889050781726837, \"start\": 829, \"end\": 845, \"clean_text\": \"permit indicator\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 829, \"end\": 845, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 829, \"end\": 845, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 829, \"end\": 845}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 829, \"end\": 845, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9890962243080139, \"start\": 118, \"end\": 134}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3889050781726837, \"start\": 118, \"end\": 134}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8898139595985413, \"start\": 118, \"end\": 134}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8898139595985413, \"start\": 118, \"end\": 134}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \": indicator for Syrian in camp, reference category is Syrian not in camp + Jordanian ii SSCC : indicator for Syrian not in camp, reference category is Syrian in camp + Jordanian ii SSnnCC : type of employment, reference group: formal employment ii EE In all models we are concerned with the vector of 1 coefficients. Those represent the marginal effect of being Syrian, informal, holding a permit or living in camp in each half year. A coefficient ββ is negative in the case of job finding and positive in the case of job separation if we expect the category in question to be experiencing increased disadvantage on the labor market in a certain half-year. Any time this coefficient has a different sign or is not significant we can refute our hypothesis of increased disadvantage for this category in this half-year.\\n\\nOur estimation models do not attempt to measure the effect of the pandemic on Syrian refugees or Syrians’ disadvantage compared to Jordanians. We are mainly concerned with estimating the differential trends between Syrians and Jordanians over time, including at the time of the pandemic, while accounting for the formality status of employment, and distinguishing between Syrians in camps and those in host communities.\\n\\n# **Results**\\n\\nWe present the results of the different models in Figures 5 to 10. As we mentioned earlier, these results concern men only. Results of the full sample are in the appendix.\\n\\nSyrians generally have lower job finding rates. The results of both populations, start to clearly diverge in 2019 where we see significantly lower job finding rates and higher job separation rates for Syrians and are further magnified during the pandemic (Figure 4). There is, however, no significant difference between Syrians’ job finding and job separation between each half-year and the next.\\n\\n21\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"<!-- Start of picture text --> N<br>:<br>:<br>1<br>~v.<br>i<br>©<br>i<br>Cc Cc Cc Cc Cc Cc Cc Cc Cc<br>© © © © © BS © © aS<br>oo a a<br>>>>>>>>>><br>”<br>* ”* A)* ”* ”* A)* A)* ”* 7)*<br>N - N - N - N - N<br>aon no nom aon ao ao com ao om<br>\\\\<o) MN Mw (oe) foe) oO (o>) (o) (o)<br>- - - - al - - N N<br>(o>) (o) (oo) (=) (o>) (oo) (=) oO (oo)<br>NNNNNNNNN<br>@Job separation ®@ Job finding<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"<!-- Start of picture text --> Oo ;<br>.<br>=a<br>.<br>i<o)<br>-<br>Cc fc f£& f& G&G f& f& f& f& G&G fF f& fF f& F&F f& CLC Cc<br>oC © © F& GF FC GC GCF GF GF FC GC GC GC GF GF G<br>oT oOo OC fF GF OC GO GO WD > DP DPD PD DD DDH DD<br>TRPRPPPPPPPLPYba PY PPP YP YP YL<br>Oo Oo 0a 0 0 0 OO Oa ON TN TN TN Tt AN<br>qxd7,cUlUmN Nr srl Nr Ne lcNe Ne cD i GS i GS GS GS GS GS HS LS LS<br>5FeSS,tFESreEeESeESSae Se [>(© 2sSESbn bnSSS© © SEE © © SEE © > EEESAYS © > SEE GD HEE GD)<br>ore en > es > ee > en > ee ee ><br>De rhr O DO DD DO DT NNN N NN WN NWN EN<br>[> ee> es >eeee cee> Pe e eePPee GP)|<br>NN NNNNN NN N<br>@ Job separation ®@ Job finding<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"<!-- Start of picture text --> N<br>oO<br>N<br>i<br>v.<br>1<br>©<br>i<br>(oe)<br>2<br>ioe Qa Qa ioe Qa Qa Qa Qa Qa<br>EE=EE==E=<br>OOOO OOo OOo O0 OO OO OO OOo<br>*********<br>N— N - N - N N<br>aon — nom aon Cc ao nom nom co<br>\\\\<o) Mw Mm (oe) foe) oO oO (oo) io)<br>- —- - - —- - - N N<br>(>) io) (=) (=) oO (o) (=) (=) (=)<br>NNNNNNNNN<br>@Job separation ®@ Job finding<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"<!-- Start of picture text --> N<br>(=)<br>N<br>1<br>wr.<br>1<br>i<o)<br>2<br>(oe)<br>2<br>=========<br>—— — — — — — po<br>oO f<b) <b) oO f<b) c<b) <b) f<b) oO<br>io io [om [om ioe [om ior io [om<br>wn wn Yn wn wn Yn wn wn Yn<br>Oo O O Oo O O Oo Oo O<br>as*as*Z*x*r*as*as*r*r*<br>aonN— N - N - N —_ N<br>\\\\<o) — nom aon ao com co ao com<br>Tc -» » (oe) foe) (o>) (o>) oO (=)<br>(=) c — Tc c— cT vc N N<br>N jo)N(o)N(=)N(o)N(=)N(=)NoO (=)<br>@Job separation ®@ Job finding<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"<!-- Start of picture text --> Ww<br>LO<br>—<br>i<br>FEEEEE<br>® © © © © O BO O<br>Oo0 op op Op oo op opEFFECof opoO90eee0 0 0eee0 0 EE0 0<br>x x x * * x x x x Q x ~ x ew x Q x A<br>STESGeESCeESEReE SE F&F fF FKL EEE<br>OF KY DO DO DH OOeFTKrF KrF DD 0 DOD DB GD<br>- —- —- —- —- —- - N N ~~ ~~ ~~ ~~ ~~ ~~ ~~ N N<br>So OO OO Ol OUlCOUlUlCUOOUCUCOUCUCOCUCOCUCON COCO OD COCO<br>aA at voananan#4w#4#ananNnwtinnnennnna<br>@ Job separation ® Job finding<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [27]}}, {\"page\": 28, \"chunk\": 28, \"input_text\": \"ww (oo) > a ? nvN ° : +r co “ [omelet ooo lloilo loll oll oll oll oll oll oll clio lilo lir—tir—tirtirTTT tirTTDtir_tirOO tirOD EEFEEEEEEEEEEEEEEEEETOOT TOTO TT HOTDOHTOHHHOBDHHADEELEEE EEEEEE EEE **EE** DCDODODDOOOOSS PSPPRPFFSESSSSSSS cccccececec **T** OTO<sup>HOaanannanwallOOOODDOQDOe</sup> ec FC FCCC CCLC LF LSEke kKLE eeEELkK KKEES SDD DADDDADaDRHeCOI KeITTKLEE OTTOKK ZOOKK NOOOO TNTOOOONT NANTONLS NDNNNNNHDHADRHDHDSEESESEROR SSE SEGEecrecornheDDBDADOO N RKN Rk ORN OR ORN OR (NeOR CESSa a SESSSESsssssssgygagA SeekerORR DADADAONNANANNHNANNec ec ec eee SsSSSSSPRRP aeNNnnnananaaNV SRR RRROnRNRNDADAAHDVOEL LKELELEEE bee sell eel wee well wel wee TE | ooono momonomone) NNANNANANAAAN @ Job separation ® Job finding\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [28]}}, {\"page\": 29, \"chunk\": 29, \"input_text\": \"To account for potential biases resulting from the random assignment of the month for individuals with missing start or end of their job we ran two robustness checks for which we show the results in the appendix. We repeated the analysis limiting the sample to those who reported the month omitting all observations with imputed month. We found that the results hold for this sample as well (Figures 16-22 in the appendix). Additionally, we limited the sample to the years 2019 and 2020 given that individuals are generally more likely to report the month for jobs taken in recent years than earlier ones. While the results are attenuated compared to the main analysis, the main conclusion holds: Syrians are significantly disadvantaged compared to Jordanians during the pandemic and this disadvantage is mainly driven by Syrians residing in camps.\\n\\n# **Conclusion**\\n\\nSyrian refugees, like other refugees, face substantial barriers to entry to the formal labor market and are therefore confined to informal employment. Even when they obtain work permits, this often merely translates into formalization of their right to work and not necessarily their employment. Concentration in informal employment means that Syrian refugees are particularly vulnerable to shocks. Moreover, Syrian refugees in camps may be experiencing an additional layer of vulnerability stemming from restrictions on movement and their inability to compete in the labor market like refugees in host communities.\\n\\nIn this paper we tried to explore the interplay of these factors for Syrian refugees during the COVID-19 pandemic. We compared the employment outcomes of Jordanians and Syrians over a period of four years before and during the COVID-19 pandemic. This was possible by constructing a synthetic semi-annual panel using the Survey of Young People in Jordan collected in 2020 and 2021. Running a fixed effects model, we found that Syrians and Jordanians experienced similar trends in job separation and job finding rates prior to the pandemic. With the onset of the pandemic, 28\", \"datasets\": [{\"mention_name\": {\"text\": \"Survey of Young People in Jordan\", \"confidence\": 0.856292188167572, \"start\": 1805, \"end\": 1837, \"clean_text\": \"Survey of Young People in Jordan\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1805, \"end\": 1837, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1805, \"end\": 1837, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1805, \"end\": 1837}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1805, \"end\": 1837, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9885968565940857, \"start\": 952, \"end\": 984}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.856292188167572, \"start\": 952, \"end\": 984}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.700021505355835, \"start\": 952, \"end\": 984}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.700021505355835, \"start\": 952, \"end\": 984}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [29]}}, {\"page\": 30, \"chunk\": 30, \"input_text\": \"however, Syrians were found to be significantly disadvantaged on both measures. Controlling for the type of employment showed that informal workers have seen their opportunities deteriorating in job separation and job finding compared to formal workers throughout the period regardless of nationality. The only exception was the second half of 2020 when Jordanian informal workers witnessed a significant recovery. Dissecting the Syrians by their residency in camps, we find that the Syrian disadvantage observed is completely driven by Syrians in camps.\\n\\nThese results highlight the importance of securing the right to work for refugees and dismantling barriers to formal employment to ensure their self-reliance and resilience in the face of shocks. Given the pressure on many host countries that are primarily developing countries with limited resources and social protection capacity, this calls for more coordinated efforts internationally to support refugees and alleviate the pressure on host country labor markets. Moreover, the results suggest the importance of revisiting camp settings, reconsidering camp closure policies, and providing job assistance to residents of camps.\\n\\n# **References**\\n\\nAbay, K. A., Tafere, K., & Woldemichael, A. (2020). Winners and Losers from COVID-19: Global Evidence from Google Search. _Winners and Losers from COVID-19: Global Evidence from Google Search_ , _June_ . https://doi.org/10.1596/1813-9450-9268 Alhawarin, I., Assaad, R., & Elsayed, A. (2021). Migration shocks and housing: Short-run impact of the Syrian refugee crisis in Jordan. _Journal of Housing Economics_ , _53_ , 101761. https://doi.org/10.1016/j.jhe.2021.101761\\n\\n- Angelucci, M., Angrisani, M., Bennett, D., Kapteyn, A., & Schaner, S. (2020). Remote Work and the Heterogeneous Impact of COVID-19 on Employment and Health. _National_\\n\\n_Bureau of Economic Research_ , 34.\\n\\n29\", \"datasets\": [{\"mention_name\": {\"text\": \"Google Search\", \"confidence\": 0.5696890354156494, \"start\": 1312, \"end\": 1325, \"clean_text\": \"Google Search\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1312, \"end\": 1325, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1312, \"end\": 1325, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1312, \"end\": 1325}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1312, \"end\": 1325, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6161355376243591, \"start\": 333, \"end\": 346}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5696890354156494, \"start\": 333, \"end\": 346}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7205677032470703, \"start\": 333, \"end\": 346}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7205677032470703, \"start\": 333, \"end\": 346}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [30]}}, {\"page\": 31, \"chunk\": 31, \"input_text\": \"Assaad, R., Krafft, C., & Sieverding, M. (2021a). _Survey of Young People in Jordan—UNICEF ACSS Dataverse V1.0_ (V1.0) [dataset]. ACSS Dataverse.\\n\\nAssaad, R., Krafft, C., & Sieverding, M. (2021b). _Youth Transitions to Adulthood in Jordan: High Aspirations, Challenging Realities_ (Issue November).\\n\\nAssaad, R., Krafft, C., & Yassin, S. (2018). Comparing retrospective and panel data collection methods to assess labor market dynamics. _IZA Journal of Development & Migration_ . https://doi.org/10.1186/s40176-018-0125-7 Assaad, R., & Salemi, C. (2019). The Structure of Employment and Job Creation in Jordan: 2010-2016. In C. Krafft & R. Assaad (Eds.), _The Jordanian Labor Market Between Fragility and Resilience_ . Oxford University Press.\\n\\nBartik, A. W., Bertrand, M., Lin, F., Rothstein, J., & Unrath, M. (2020). Measuring the labor market at the onset of the COVID-19 Crisis. _Brookings Papers on Economic Activity_ , _Special Edition_ , 239–268. https://doi.org/10.1353/eca.2020.0010 Beaujouan, J., & Rasheed, A. (2020). The Syrian Refugee Policy of the Jordanian Government.\\n\\nIn _Syrian Crisis, Syrian Refugees_ (pp. 47–63). https://doi.org/10.1007/978-3-030-35016-1 Betts, A., Bloom, L., Kaplan, J. D., & Omata, N. (2017a). _Refugee economies: Forced_ _displacement and development_ (First edition). Oxford University Press.\\n\\nBetts, A., Bloom, L., Kaplan, J., & Omata, N. (2017b). _Refugee Economies: Forced_ _Displacement and Development_ (First edit). Oxford University Press.\\n\\nBorjas, G. J., & Cassidy, H. (2020). The Adverse Effect of the COVID-19 Labor Market Shock on Immigrant Employment. _National Bureau of Economic Research_ , _27243_ .\\n\\nCortes, G. M., & Forsythe, E. (2020). The Heterogeneous Labor Market Impacts of the Covid-19 Pandemic. _SSRN Electronic Journal_ . https://doi.org/10.2139/ssrn.3634715 30\", \"datasets\": [{\"mention_name\": {\"text\": \"ACSS Dataverse\", \"confidence\": 0.48792558908462524, \"start\": 130, \"end\": 144, \"clean_text\": \"ACSS Dataverse\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 130, \"end\": 144, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 130, \"end\": 144, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 130, \"end\": 144}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 130, \"end\": 144, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.776412308216095, \"start\": 130, \"end\": 144}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.48792558908462524, \"start\": 130, \"end\": 144}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5478965640068054, \"start\": 130, \"end\": 144}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5478965640068054, \"start\": 130, \"end\": 144}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [31]}}, {\"page\": 32, \"chunk\": 32, \"input_text\": \"Demirci, M., & Güray Kırdar, M. (2021). _THE LABOR MARKET INTEGRATION OF SYRIAN_ _REFUGEES IN TURKEY_ (14973; IZA Discussion Paper Series, Issue December). Garrote Sanchez, D., Gomez Parra, N., Ozden, C., Rijkers, B., Viollaz, M., & Winkler, H.\\n\\n(2021). Who on Earth Can Work from Home? _World Bank Research Observer_ , _36_ (1), 67–100. https://doi.org/10.1093/wbro/lkab002 Ginn, T. D. (2018). _Prison or Sanctuary? An Evaluation of Camps for Syrian Refugees_ . https://www.semanticscholar.org/paper/Prison-or-Sanctuary-An-Evaluation-of-Campsfor-Ginn/1d968936d17d45bb2dc5901000253ed37d6178fb Gordon, J. (2019). Refugees and decent work lessons learned from recent refugee jobs compacts. In _Employment Working Paper_ (Issue 256).\\n\\nGottlieb, C., Saltiel, F., Gottlieb, C., & Poschke, M. (2020). _Working from Home in Developing Countries Working from Home in Developing Countries_ (13737; Discussion Paper Series, Issue 13737).\\n\\nHoseini, M., & Dideh, M. J. (2022). _How do shared experiences of economic shocks impact refugees and host communities? Evidence from Afghan refugees in Iran 1_ (Issue Policy Research Working Paper 9915).\\n\\nILO. (2015). Access to work for Syrian refugees in Jordan: A discussion paper on labour and refugee laws and policies. In _Beirut: ILO Regional Office for Arab States_ .\\n\\nKrafft, C., Assaad, R., Marouani, M. A., Cheung, R., & LaPlante, A. (2022). _Are labour markets in the Middle East and North Africa recovering from the COVID-19 pandemic?_ (ILO/ERF Working Paper).\\n\\n31\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [32]}}, {\"page\": 33, \"chunk\": 33, \"input_text\": \"Krafft, C., Assaad, R., Marouani, M. A., Cheung, R., & LaPlante, A. (Forthcoming).\\n\\nAre labour markets in the Middle East and North Africa recovering from the COVID-19 pandemic? _IZA Journal of Development and Migration._ Krafft, C., Razzaz, S., Keo, C., & Assaad, R. (2019). The Number and Characteristics of Syrians in Jordan: A Multi-source Analysis. In C. Krafft & R. Assaad (Eds.), _The Jordanian Labor Market between Fragility and Resilience_ (pp. 173–202). Oxford University Press. OAMDI. (2021). _COVID-19 MENA Monitor Household Survey (CMMHH)_ (Version 5.0 JordanCMMHH) [dataset]. http://www.erfdataportal.com/index.php/catalog.\\n\\nOAMDI. (2022). _Survey of Young People in Jordan—UNICEF_ (Version 1.0) [dataset]. Egypt: Economic Research Forum (ERF).\\n\\nObi, C. T. (2021). _The Impact of Living Arrangements (In-Camp versus Out-of-Camp) on the Quality of Life: A Case Study of Syrian Refugees in Jordan_ . https://doi.org/10.1596/18139450-9533 Razzaz, S. (2017). _A Challenging Market Becomes More Challenging: Jordanian Workers, Migrant Workers and Refugees in the Jordanian Labor Market_ .\\n\\nRuiz, I., & Vargas-Silva, C. (2013). The Economics of Forced Migration. _Journal of_ _Development Studies_ , _49_ (6), 772–784. https://doi.org/10.1080/00220388.2013.777707 Ruiz, I., & Vargas-Silva, C. (2017). Are Refugees’ Labour Market Outcomes Different from Those of Other Migrants? Evidence from the United Kingdom in the 2005-2007 Period: Refugees and Other Migrants Labour Market Outcomes: Evidence from the UK. _Population, Space and Place_ , _23_ (6), e2049. https://doi.org/10.1002/psp.2049 UNHCR. (2021a). Global Trends: Forced Displacement in 2020. _UNHCR Global Trends_ , 72.\\n\\n32\", \"datasets\": [{\"mention_name\": {\"text\": \"JordanCMMHH\", \"confidence\": 0.35960835218429565, \"start\": 566, \"end\": 577, \"clean_text\": \"JordanCMMHH\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 566, \"end\": 577, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 566, \"end\": 577, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 566, \"end\": 577}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 566, \"end\": 577, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.6963821649551392, \"start\": 566, \"end\": 577}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.35960835218429565, \"start\": 566, \"end\": 577}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.34779971837997437, \"start\": 566, \"end\": 577}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.34779971837997437, \"start\": 566, \"end\": 577}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [33]}}, {\"page\": 34, \"chunk\": 34, \"input_text\": \"UNHCR. (2021b). _Refugees in Jordan face escalating challenges as Syria conflict marks 10-year anniversary_ . https://www.unhcr.org/jo/14617-refugees-in-jordan-face-escalatingchallenges-as-syria-conflict-marks-10-year-anniversary.html UNHCR. (2022). _Syria Regional Refugee Response_ . https://data2.unhcr.org/en/situations/syria Verme, P. (2017). _The Economics of Forced Displacement: An Introduction_ . World Bank, Washington, DC. https://doi.org/10.1596/1813-9450-8038 Verme, P., & Schuettler, K. (2021). The impact of forced displacement on host communities: A review of the empirical literature in economics. _Journal of Development Economics_ , _150_ . https://doi.org/10.1016/j.jdeveco.2020.102606 Vintar, M., Beltramo, T., Delius, A., Egger, D., & Pape, U. J. (2022). _Impact of COVID-19 on Labor Market Outcomes of Refugees and Nationals in Kenya_ (9960; Policy Research Working Paper).\\n\\nWorld Bank. (2020). _Jordan Economic Monitor: Weathering the storm_ .\\n\\nhttps://doi.org/10.1126/science.370.6514.278 Zetter, R., & Ruaudel, H. (2018). Refugees’ right to work and access to labour markets: Constraints, challenges and ways forward. _Forced Migration Review_ , _58_ (June), 4–7.\\n\\n33\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821301162430228/pdf/IDU16c8b4f4a1d5c9142801a62b166575607c40c.pdf\", \"pages\": [34]}}]}"}}},{"rowIdx":77,"cells":{"doc_id":{"kind":"string","value":"34239717"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34239717\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"admreg\": \"Other,Other\", \"admreg_key\": \"912972,912972\", \"authors\": {\"0\": {\"author\": \"Horton,Susan E.\"}, \"1\": {\"author\": \"Fracassi,Patrizia\"}, \"2\": {\"author\": \"Siekmans,Kendra\"}, \"3\": {\"author\": \"Kato,Tomoko\"}, \"4\": {\"author\": \"Hanley-Cook,Giles\"}}, \"count\": \"World\", \"count_key\": \"517191\", \"docna\": {\"0\": {\"docna\": \"Estimating Economic Costs of Unhealthy Diets\\n : A Proposed Methodology\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"Agriculture and Food GE (SAGGL)\", \"projn\": \"1W-Repurposing Agrifood Public Policies And\\n Support For Healthy Di -- P180735\", \"subsc\": \"Other Agriculture, Fishing and Forestry\", \"trustfund\": \"TF0C1256-Curbing the rise of NCDs\\n through repurposing agrifood support,TF0C4026-Country-level trajectories for\\n sustainable and nutritious diets\", \"trustfund_key\": \"2290096,2393118\", \"theme\": \"Urban Infrastructure and Service Delivery,Mitigation,Human Development and Gender,Economic Policy,Food Security,Nutrition,Trade Policy,Fiscal Policy,Environment and Natural Resource Management,Climate change,Urban and Rural Development,Trade,Adaptation,Public Expenditure Policy,Nutrition and Food Security,Urban Development\", \"prdln\": \"Advisory Services & Analytics\", \"prdln_key\": \"1416466\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34239717\"}, \"repnb\": \"WPS10671\", \"docdt\": \"2024-01-16T00:00:00Z\", \"datestored\": \"2024-01-16T06:39:48Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"Healthy diets have been characterized as\\n responding to four universal principles—nutrient\\n adequacy, dietary diversity, macronutrient balance, and\\n moderation. With rising incomes, diet concerns globally have\\n shifted from inadequacy of nutrients and lack of diversity,\\n to lack of balance and moderation. This has occurred\\n alongside declining rates of stunting and wasting in\\n children under five and increasing rates of overweight and\\n obesity across a broad age span. Calculations undertaken for\\n low- and middle-income countries for policy and advocacy\\n purposes of the economic cost of unhealthy diets have used\\n nutritional status as a proxy or have made estimates of the\\n impact of noncommunicable diseases by simply adding up known\\n risks of individual diet factors. Both these methods have\\n problems. This paper proposes a new methodology, taking\\n advantage of recent, more holistic, measures of diet\\n quality. Preliminary regression results are presented using\\n cross-country data and the Global Dietary Recommendations\\n Score and Minimum Dietary Diversity for women. The results\\n suggest that diet quality variables generally have the\\n expected signs, but there are also clear limitations of\\n using cross-country data. The methodology could be applied\\n in future to a limited number of broadly representative low-\\n and middle-income countries data sets containing both diet\\n recall data as well as measures of noncommunicable disease\\n risk status. The analysis suggests that this work could\\n inform policies such as the repurposing of existing agrifood\\n policies to complement existing public health policies, to\\n reduce the economic and health burdens imposed by unhealthy diets.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10671; PEOPLE\", \"display_title\": \"Estimating Economic Costs of\\n Unhealthy Diets : A Proposed Methodology\", \"disclosure_date\": \"2024-01-16T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099839401162431298/text/IDU152ca6b541085314ca41a22a1bf67750bec89.txt\", \"listing_relative_url\": \"/research/2024/01/34239717/estimating-economic-costs-unhealthy-diets-proposed-methodology\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099839401162431298/IDU152ca6b541085314ca41a22a1bf67750bec89\", \"new_url\": \"2024/01/34239717/IDU152ca6b541085314ca41a22a1bf67750bec89\", \"ext_pub_date\": \"2024-01-16T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Malnutrition m1627088 4737\", \"docm_id\": \"IDU-52ca6b54-0853-4ca4-a22a-bf67750bec89\", \"chronical_docm_id\": \"IDU-52ca6b54-0853-4ca4-a22a-bf67750bec89\", \"publishtoextweb_dt\": \"2024-01-16T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10671\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"projectid\": \"P180735\", \"guid\": \"099839401162431298\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"47\", \"author_keys\": {\"0\": {\"author_key\": \"773306\"}, \"1\": {\"author_key\": \"1057102\"}, \"2\": {\"author_key\": \"NA\"}, \"3\": {\"author_key\": \"947273\"}, \"4\": {\"author_key\": \"NA\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099839401162431298/IDU152ca6b541085314ca41a22a1bf67750bec89\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10671 Estimating Economic Costs of Unhealthy Diets A Proposed Methodology _Susan Horton Patrizia Fracassi Kendra Siekmans Tomoko Kato Giles Hanley-Cook_ Agriculture and Food Global Practice January 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10671\\n\\n# **Abstract**\\n\\nHealthy diets have been characterized as responding to four universal principles—nutrient adequacy, dietary diversity, macronutrient balance, and moderation. With rising incomes, diet concerns globally have shifted from inadequacy of nutrients and lack of diversity, to lack of balance and moderation. This has occurred alongside declining rates of stunting and wasting in children under five and increasing rates of overweight and obesity across a broad age span. Calculations undertaken for low- and middle-income countries for policy and advocacy purposes of the economic cost of unhealthy diets have used nutritional status as a proxy or have made estimates of the impact of noncommunicable diseases by simply adding up known risks of individual diet factors. Both these methods have problems. This paper proposes a new methodology, taking advantage of recent, more holistic, measures of diet quality. Preliminary regression results are presented using cross-country data and the Global Dietary Recommendations Score and Minimum Dietary Diversity for women. The results suggest that diet quality variables generally have the expected signs, but there are also clear limitations of using cross-country data. The methodology could be applied in future to a limited number of broadly representative low- and middle-income countries data sets containing both diet recall data as well as measures of noncommunicable disease risk status. The analysis suggests that this work could inform policies such as the repurposing of existing agrifood policies to complement existing public health policies, to reduce the economic and health burdens imposed by unhealthy diets.\\n\\nThis paper is a product of the Agriculture and Food Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at sehotrton@uwaterloo.ca.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [{\"mention_name\": {\"text\": \"cross-country data\", \"confidence\": 0.8388298749923706, \"start\": 1010, \"end\": 1028, \"clean_text\": \"cross-country data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1010, \"end\": 1028, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1010, \"end\": 1028, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1010, \"end\": 1028}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1010, \"end\": 1028, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.4719136655330658, \"start\": 1010, \"end\": 1028}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8388298749923706, \"start\": 1010, \"end\": 1028}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6954019069671631, \"start\": 1010, \"end\": 1028}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6954019069671631, \"start\": 1010, \"end\": 1028}}, {\"mention_name\": {\"text\": \"Global Dietary Recommendations Score\", \"confidence\": 0.8261899352073669, \"start\": 1037, \"end\": 1073, \"clean_text\": \"Global Dietary Recommendations Score\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1037, \"end\": 1073, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1037, \"end\": 1073, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1037, \"end\": 1073}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1037, \"end\": 1073, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.4719136655330658, \"start\": 1037, \"end\": 1073}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8261899352073669, \"start\": 1037, \"end\": 1073}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6954019069671631, \"start\": 1037, \"end\": 1073}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6954019069671631, \"start\": 1037, \"end\": 1073}}, {\"mention_name\": {\"text\": \"diet recall data\", \"confidence\": 0.3035598397254944, \"start\": 1412, \"end\": 1428, \"clean_text\": \"diet recall data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1412, \"end\": 1428, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1412, \"end\": 1428, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1412, \"end\": 1428}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1412, \"end\": 1428, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.30784347653388977, \"start\": 505, \"end\": 521}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3035598397254944, \"start\": 505, \"end\": 521}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.48414644598960876, \"start\": 505, \"end\": 521}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.48414644598960876, \"start\": 505, \"end\": 521}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# Estimating Economic Costs of Unhealthy Diets: A Proposed Methodology\\n\\nSusan Horton,<sup>1</sup> Patrizia Fracassi,<sup>2</sup> Kendra Siekmans,<sup>3</sup> Tomoko Kato,<sup>4</sup> Giles Hanley-Cook<sup>5</sup>\\n\\n## **JEL classification:** I12, I15, I18, I38, Q18\\n\\n**Keywords:** diet quality, food choices, undernutrition, overweight, obesity, non-communicable disease, healthcare costs, agrifood policy\\n\\n## **Acknowledgments**\\n\\nThis work is a product of the Programmatic ASA _‘Repurposing Agrifood Public Policies and Support for Healthy Diets’_ (P180735) led by Felipe Dizon (Senior Economist, SAGGL) and Kyoko Shibata Okamura (Senior Nutrition Specialist, HHNGE). The authors would like to thank colleagues from the Healthy Diets project for their support of related work, including Erkan Ozcelik, Ti Kian Seow, and Diana Carter from FAO, and Felipe Dizon, Kyoko Shibata Okamura, and Hina Sherwani from the World Bank. We thank Mireya Vilar, Andrea Cattaneo, and John Hoddinott for their careful review of the paper. We thank Lynnette Neufeld, Bridget Holmes, Harold Alderman, Patrick Webb, and participants in a joint World Bank-FAO seminar on September 14, 2023 for helpful comments on earlier drafts. We thank Sangmin Seo, Rachel Bryant, and Laura Jimenez for expert research assistance for the empirical work. All remaining errors are our own. The findings, interpretations and conclusions expressed herein are those of the authors and do not necessarily reflect the view of the World Bank Group, its Board of Directors, or the governments they represent.\\n\\n1 University of Waterloo. Corresponding author: sehorton@uwaterloo.ca 2 Food and Nutrition Division, FAO. Patrizia.Fracassi@fao.org 3 Consultant, Food and Nutrition Division, FAO. ksiekmans@nurturedev.com\\n\\n> 4 Food and Nutrition Division, FAO. <u>Tomoko.Kato@fao.org</u>\\n\\n5 Consultant, Food and Nutrition Division, FAO. Giles.HanleyCook@fao.org\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# 1. Introduction\\n\\nThis Working Paper aims to develop a methodology to assess the economic costs of unhealthy diets, which can be particularly informative for agrifood policy. We argue that existing work on the economic cost of malnutrition in all its forms has disadvantages, since several factors other than diet can lead to malnutrition, and even individuals with healthy weights are still at risk of the detrimental health impacts of unhealthy diets.\\n\\nOur understanding of what constitutes a healthy diet has evolved over time. The early literature on food and nutrition in low- and middle-income countries (LMICs) focused on undernutrition (with low heightfor-age and low weight-for-height of children under five being major concerns). Initially the major focus was adequate energy intake once an early debate over the importance of calories versus protein was resolved. Subsequently attention turned to other dietary components, especially micronutrients. One estimate of the cost of child stunting, which is affected by both energy and micronutrient intakes, suggested that it was associated with an 8% loss of global GDP over the period 1900-2000 (Horton and Steckel 2013), while FAO estimated that chronic undernutrition (referring to child stunting) caused losses of 3.5% of global GDP in the early 21<sup>st</sup> century (FAO, 2013). The 2013 _Lancet_ Maternal and Child Nutrition series estimated that 46% of deaths of children under five were attributable to malnutrition (also referring to undernutrition) (Black et al. 2013).\\n\\nIn the past two decades, however, there has been rising concern about increases in rates of overweight and obesity globally (Popkin 2004), termed the “nutrition transition”, related to unhealthy diets. A large literature discusses the adverse effects of overweight and obesity on health and economic outcomes. While this literature initially focused on impacts in high-income countries, subsequent estimates have also been made for LMICs where the prevalence of overweight and obesity has been rising quickly (Schneider et al. 2020). One recent estimate for 161 countries (comprising high-income countries, as well as LMICs) is that the global economic losses attributable to overweight and obesity amount to 2.2% of global GDP (Okunogbe et al. 2022). The corresponding human cost was estimated at 4.7 million deaths in 2017 (Dai et al. 2020); an update using 2019 data associated this cost with 5 million deaths due to noncommunicable diseases (NCDs), 77% of these occurring in LMICs (Abbafati et al. 2020). Similarly, the World Bank (2020) estimated that 4 million deaths globally were attributable to obesity and projected that by 2035 the costs of obesity in LMICs will amount to $7 trillion.\\n\\nThere is also literature linking dietary factors directly to NCDs (e.g., GBD 2017 Diet Collaborators 2019), with some dietary factors protecting against and others increasing risks of NCDs. GBD 2017 Diet Collaborators (2019) use a bottom-up methodology for each of 15 dietary factors, adding up projected effects and estimating that these dietary risks accounted for 11 million deaths globally in 2017. The effects of some of the dietary risks (e.g., excess consumption of sugar and fat) may be mediated through obesity; while others may operate independently (e.g., GBD 2017 Diet Collaborators, 2019, estimate that 3 million deaths were attributable to excess sodium intake in 2017). Arguably using a more holistic measure of diet quality is preferable to adding up the risks due to multiple individual dietary factors, and less prone to double counting.\\n\\nAnother literature, even broader in scope, ties malnutrition with climate considerations (e.g., The Food and Land Use Coalition – FOLU – 2019). FOLU (2019) presents estimates of the economic costs of, both what they term, undernutrition (stunting, wasting, and underweight among children under five years of age) as 2.0% of global GDP, and of overweight and obesity as 3.1% of global GDP, based on their SYSTEMIQ 2\", \"datasets\": [{\"mention_name\": {\"text\": \"2019 data\", \"confidence\": 0.3889330327510834, \"start\": 2403, \"end\": 2412, \"clean_text\": \"2019 data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2403, \"end\": 2412, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2403, \"end\": 2412, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2403, \"end\": 2412}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2403, \"end\": 2412, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.23734337091445923, \"start\": 745, \"end\": 754}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3889330327510834, \"start\": 745, \"end\": 754}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7295032739639282, \"start\": 745, \"end\": 754}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7295032739639282, \"start\": 745, \"end\": 754}}, {\"mention_name\": {\"text\": \"SYSTEMIQ 2\", \"confidence\": 0.7149497866630554, \"start\": 4002, \"end\": 4012, \"clean_text\": \"SYSTEMIQ 2\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4002, \"end\": 4012, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4002, \"end\": 4012, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4002, \"end\": 4012}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4002, \"end\": 4012, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.7207224369049072, \"start\": 788, \"end\": 798}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7149497866630554, \"start\": 788, \"end\": 798}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9101930856704712, \"start\": 788, \"end\": 798}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9101930856704712, \"start\": 788, \"end\": 798}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"model.<sup>1</sup> These costs are in turn components of their much larger $12 trillion (13.8% of global GDP) estimates of hidden costs of global agrifood and land use systems. These larger costs use a planetary rather than a human health framework, incorporating environmental costs.\\n\\nThe “double burden” of malnutrition refers to: (1) the co-existence of undernutrition (as evidenced by child stunting, child wasting, women thinness) and adult or child overweight at country level; (2) the co-existence of one or more individuals with wasting, stunting, or thinness and one or more individuals with overweight or obesity within the same household and; (3) the underlying developmental origins biology on what is termed the Developmental Origins and Health and Disease (DOHAD), originating in the work of Barker (2004). The DOHAD theory posits a biological basis for why diets with inadequate energy and nutrients levels during early pre-conception and pregnancy predispose offspring to higher risks of obesity and associated NCDs in adulthood (Wells et al. 2020).\\n\\nFurthermore, with recent global estimates that over half of preschool-aged children and two-thirds of nonpregnant women of reproductive age have at least one of three micronutrient deficiencies (Stevens et al. 2022), most LMICs are dealing in reality with a “triple burden” of malnutrition. Micronutrient deficiencies can occur “at either end of the anthropometric spectrum” as well as among individuals whose body mass index is within the healthy range (i.e., 18.5-24.9 kg/m2) (Mwangome and Prentice 2019).\\n\\nRecent UN statements have referred to “malnutrition in all its forms,” for example in Sustainable Development Goal 2. Countries facing both stunting and obesity encounter more complications in policies; Hawkes et al. (2020) propose ten ways to reorient nutrition interventions to address both burdens simultaneously and Luo, Zyba, and Webb (2020) propose an index to measure nutrition in all its forms.\\n\\nThe discussion proceeds as follows. We first outline the important definitions being used throughout this Working Paper and present a basic conceptual framework. We then discuss issues of measurement, and review data sources, focusing on healthy diet metrics collected at national-level across a wide array of countries. This is followed by a brief summary of previous related costing studies, mainly focused on the costs of malnutrition in all its forms. We then propose a methodology for estimating the economic costs of unhealthy diets, along with preliminary cross-country empirical work. The concluding section summarizes key findings in terms of etiological pathways, nutrition and health outcomes, and economic effects, and notes key gaps and areas for future work.\\n\\n# 2. Definitions and Conceptual Framework\\n\\nThe 2021 UN Food Systems Summit used the following definition: “A healthy diet is health-promoting and disease-preventing. It provides adequacy without excess, of nutrients and health-promoting substances from nutritious foods and avoids the consumption of health-harming substances.” (Neufeld, Henriks and Hugas 2019). Healthy diets respond to the universal principles of nutrient adequacy, dietary diversity, macronutrient balance, and moderation (Verger et al. 2023). In what follows we consider three different diet aspects: energy sufficiency, nutrient adequacy, and food choices which promote health, using Figure 1 published by FAO (2020). These three aspects are also consistent with the “triple burden” of malnutrition discussed above.\\n\\n**Energy insufficient diets** have large economic consequences in terms of disability-adjusted life-year (DALY) losses from high child mortality and direct costs of treatment of acute malnutrition. Key pathways\\n\\n> 1 Conversions from the amounts in $ trillion in FOLU (2019), for 2018, to percentage of GDP were undertaken by the present authors, using $86.4 trillion as global GDP in 2018.\\n\\n3\", \"datasets\": [{\"mention_name\": {\"text\": \"healthy diet metrics\", \"confidence\": 0.448243111371994, \"start\": 2218, \"end\": 2238, \"clean_text\": \"healthy diet metrics\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2218, \"end\": 2238, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2218, \"end\": 2238, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2218, \"end\": 2238}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2218, \"end\": 2238, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.25978103280067444, \"start\": 672, \"end\": 692}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.448243111371994, \"start\": 672, \"end\": 692}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6230496764183044, \"start\": 672, \"end\": 692}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6230496764183044, \"start\": 672, \"end\": 692}}, {\"mention_name\": {\"text\": \"cross-country empirical work\", \"confidence\": 0.5885743498802185, \"start\": 2543, \"end\": 2571, \"clean_text\": \"cross-country empirical work\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2543, \"end\": 2571, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2543, \"end\": 2571, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2543, \"end\": 2571}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2543, \"end\": 2571, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9973986148834229, \"start\": 141, \"end\": 169}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5885743498802185, \"start\": 141, \"end\": 169}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4475010633468628, \"start\": 141, \"end\": 169}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4475010633468628, \"start\": 141, \"end\": 169}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"to assess are food energy supply, leading to child stunting and wasting, and subsequently mortality and treatment costs (Figure 2). Child stunting is also associated with cognitive deficits and reduced educational attainment, leading to reduced productivity and hence economic costs (Figure 2). For further documentation of the effects of stunting and wasting in early childhood, see systematic reviews by Black et al. (2013) and Olofin et al. (2023). Diets which are insufficient in energy, by definition, cannot be adequate in all nutrients.\\n\\n**Nutrient inadequate diets** (which are, however, sufficient in energy) have numerous consequences including reduced immunity to infection, increased rates of stunting, cognitive deficits, and birth defects. Various deleterious effects of nutrient inadequacies are potentially mediated through stunting (e.g., macronutrient and zinc deficiencies), while other nutrient inadequacies such as iodine and folic acid, are not. Anaemia in pregnancy has adverse effects on birth outcomes (Abu-Ouf and Jan 2015) and, if severe, increases the risk of maternal mortality (Daru et al, 2018). Furthermore, in adults, anemia has adverse effects on worker productivity (Marcus Schauer and Zlotkin 2021). For further documentation of the effects of iron deficiency in early childhood, see systematic reviews by Carter et al. (2010) and GranthamMacGregor and Ani (2001). Policy response to the limited availability of key micronutrients in commonly consumed foods have often included provision of micronutrient supplements to young children and pregnant women, fortification of key foods, and biofortification. There is a large literature on the costs and benefits of supplementation and fortification. The costs of individual micronutrient deficiencies are not included in our study, which focuses on food choices that individuals and households make in their diets.\\n\\n**Diets characterised by excess intakes of unhealthy foods** (which are, however, sufficient in energy) and inadequate intakes of healthy foods have their highest economic effect in terms of cost-of-illness from NCDs. Overweight and obesity are mediators for many NCDs (i.e., indirect effects), but there are also NCD risks directly related to specific diet components (e.g. salt, processed meats). Obesity in adults is associated with reduced adult life expectancy, reduced productivity, increased health care costs, early retirement, and reduced length of disability-free life years (Schneider et al. 2020). Key pathways to measure are from including food choices which increase health protection and avoiding food choices which increase health risk (emphasizing dietary balance and moderation), to NCD risks, hence NCD treatment costs and loss of productivity from morbidity and premature mortality.\\n\\nThere are differences in the pathways which are important at different stages in the life cycle (Table 1). More attention has been paid to date on children under 5, and women of reproductive age, not only in terms of research studies, but also development of dietary metrics, and data collection (discussed in the next section).\\n\\n**Non-diet factors** also affect nutritional status and health risks (Figure 2). For example, some individuals may be stunted or wasted despite a healthy diet, because of infection or overweight due to a particular genetic predisposition. Some individuals may have healthy weight but increased NCD risk because of exposure to health compromising behaviors like smoking. Some individuals may be overweight or obese due to their exposure to obesogenic environments, whereby low physical activity may be a significant influencing factor together with other health compromising conditions and behaviors. These aspects are taken up later in the presentation of a proposed methodology.\\n\\nIndividual countries may contain a mix of individuals and households in the four different diet categories (insufficient energy intake, inadequate nutrient intake, unhealthy food choices, and healthy diets), depending on levels of income, underlying inequities in terms of food availability and affordability, and disparities in consumption. While Latin America and the Caribbean, Eastern Europe and Central Asia, and East Asia (led by China and Indonesia) show growing rates of overweight and obesity among the poor, in Sub-Saharan Africa and South Asia the largest increases in prevalence are among higher-wealth households 4\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"in urban areas. In all LMICs, other than in South Asia and Sub-Saharan Africa, rural overweight and obesity is growing faster than in urban areas, albeit originally from lower levels (Popkin, Corvaland and GrummerStrawn 2020). These different patterns have implications for agrifood and health policies.\\n\\n# 3. Metrics for Healthy/Unhealthy Diets (Diet Quality)\\n\\nTable 2 summarizes information on some commonly used dietary metrics available at different life-cycle stages, discussed in this section. Further details on sources, availability and methods of measurement are given in Appendix Tables 1, 2 and 3.\\n\\nThe growing burden of overweight and obesity, as well as NCDs, has required the development of novel healthy diet metrics to complement the more well-established metrics for energy sufficiency and nutrient adequacy of the diet (Table 2). The focus of the measures of (in)sufficiency and (in)adequacy has primarily been in children under 5 and in women of reproductive age (15-49), perhaps because these are the periods of greatest vulnerability to malnutrition and greatest probability of adverse outcomes. However, the outcomes of unhealthy diets affect all age groups, and new efforts to collect these data therefore also include adolescents (a period where eating habits may be established and maintained into adulthood) and adult men (who are also at risk of diet-related NCDs). In LMICs there has been to date less attention to older adults, while high-income countries with larger proportions of this age-group are already collecting data. There are very few data on younger school-age children (ages 5 through 12).\\n\\nDetailed definitions and sources for diet variables are described in the Appendix Tables 1, 2 and 3, (corresponding to the lifecycle stages distinguished in Table 2) along with information on the number of countries/LMICs for which data are currently available. Since our focus is global, we have not included metrics specific to individual countries or regions. Data on metrics of nutritional and health status are well established and not reviewed here; UNICEF and WHO maintain global databases in this area (see Appendix Tables 4 and 5). Schooling outcomes and standardized test outcomes are available from UNESCO and key ones are included in the World Bank World Development Indicators database.\\n\\nTo help assess **energy sufficiency** , data on per capita food supply have been collated regularly at national level by FAO (and give some indication as to whether availability meets nutrient requirements). Measures compare food available and distributed and what is needed, according to the composition of the population by age, gender and physical activity levels. The Prevalence of Undernourishment (PoU, one of the suite of food security measures maintained by FAO) provides an estimate on how many people lack enough dietary energy for a healthy and active life. The prevalence of moderate or severe food insecurity based on the Food Insecurity Experience Scale (FIES, also maintained by FAO) provides an estimate on how many people do not have access to nutritious and sufficient food due lack of money or other resources. Energy sufficiency can also be assessed at the individual-level using quantitative dietary recall data.\\n\\nTo assess **nutrient adequacy** , quantitative data from dietary recall is considered to be the preferred source. However, given the difficulties in maintaining nationally-representative, up-to-date, quantitative data, metrics currently in widespread use employ recall to measure whether or not individuals consumed food from each of various food groups. One widespread summary population-level dietary indicator is that of **minimum diet diversity (MDD)** ; however, this has been developed only for young children and women. For children 6-23 months of age, MDD is defined as consumption of at least five out of eight food groups including breastmilk in the preceding 24 hours (WHO and UNICEF, 2021). For non-pregnant women of reproductive age (15-49 years), MDD-W is defined as consumption of at least five out of 10 food groups over the last 24 hours (FAO 2021). MDD-W has been validated as a proxy indicator for a minimally 5\", \"datasets\": [{\"mention_name\": {\"text\": \"World Bank World Development Indicators database\", \"confidence\": 0.4873239994049072, \"start\": 2283, \"end\": 2331, \"clean_text\": \"World Bank World Development Indicators database\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2283, \"end\": 2331, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2283, \"end\": 2331, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2283, \"end\": 2331}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2283, \"end\": 2331, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.8705809116363525, \"start\": 613, \"end\": 661}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4873239994049072, \"start\": 613, \"end\": 661}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7458481788635254, \"start\": 613, \"end\": 661}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7458481788635254, \"start\": 613, \"end\": 661}}, {\"mention_name\": {\"text\": \"data on per capita food supply\", \"confidence\": 0.534427285194397, \"start\": 2374, \"end\": 2404, \"clean_text\": \"data on per capita food supply\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2374, \"end\": 2404, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2374, \"end\": 2404, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2374, \"end\": 2404}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2374, \"end\": 2404, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.41937604546546936, \"start\": 704, \"end\": 734}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.534427285194397, \"start\": 704, \"end\": 734}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5426687598228455, \"start\": 704, \"end\": 734}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5426687598228455, \"start\": 704, \"end\": 734}}, {\"mention_name\": {\"text\": \"Food Insecurity Experience Scale\", \"confidence\": 0.3105698823928833, \"start\": 2969, \"end\": 3001, \"clean_text\": \"Food Insecurity Experience Scale\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2969, \"end\": 3001, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2969, \"end\": 3001, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2969, \"end\": 3001}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2969, \"end\": 3001, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9352437257766724, \"start\": 478, \"end\": 510}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3105698823928833, \"start\": 478, \"end\": 510}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8489221334457397, \"start\": 478, \"end\": 510}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8489221334457397, \"start\": 478, \"end\": 510}}, {\"mention_name\": {\"text\": \"quantitative dietary recall data\", \"confidence\": 0.6360040903091431, \"start\": 3234, \"end\": 3266, \"clean_text\": \"quantitative dietary recall data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3234, \"end\": 3266, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3234, \"end\": 3266, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3234, \"end\": 3266}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3234, \"end\": 3266, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.24761009216308594, \"start\": 743, \"end\": 775}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6360040903091431, \"start\": 743, \"end\": 775}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6459454298019409, \"start\": 743, \"end\": 775}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6459454298019409, \"start\": 743, \"end\": 775}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"acceptable level of dietary adequacy for 11 micronutrients (Martin-Prevel et al. 2017). Diet diversity data have been collected for a sufficient number of years and countries such that there is a body of research linking these measures with nutritional status.\\n\\nResearch suggests that **MDD for young children (6-23 months)** is associated with micronutrient adequacy (Molani-Gol et al. 2023), improved linear growth in some contexts (Gassara and Chen 2021), and lower risk of anemia (Donkor et al. 2021; Kathuria et al. 2023; Molla et al. 2020), but there is no strong indication of links with overweight or obesity. Recently, **egg and flesh food consumption** (which is a component of MDD) was included as an indicator of nutrient-sufficient diets among young children 6-23 months (WHO and UNICEF 2021) and data are widely available for LMICs (UNICEF 2022). Observational studies suggest a strong association for consumption of animal-source foods (ASF) with improved linear growth (Headey, Hirvonen and Hoddinott 2018; Krasevec et al. 2017; Zaharia et al. 2021), while systematic reviews of controlled trials found mixed results (Eaton et al. 2019; Shapiro et al. 2019).\\n\\nData collection began more recently for **MDD for non-pregnant women of reproductive age (15-49 years).** MDD-W is now being collected through large-scale DHS surveys in at least 24 countries (although, at the time of writing, results were available for only a limited number) as well as through Gallup World Poll surveys using telephone-based interviews (2021-2022 results available for women 15-49 years for 55 countries plus Palestine). As for children, MDD-W has a strong correlation with nutrient adequacy of diets (Martin-Prevel et al. 2017) but dietary diversity (i.e., not focusing on MDD-W only) has not shown consistent relationship with overweight or obesity and NCDs (Salehi-Abargouei et al. 2016; Vadiveloo, Dixon and Parekh 2013, Verger et al. 2021).\\n\\nTo assess whether a diet is **healthy,** quantitative dietary intake data are viewed as the reference, using detailed data on foods consumed. However, there is only very limited availability of validated metrics based on quantitative data that are well-suited for monitoring healthy diets in LMICs (Miller Micha and Mozaffarian 2020; WHO and UNICEF 2021). When quantitative dietary survey data are not available, the most common alternative approach in LMIC is to assess consumption of specific food groups (both protective and risk-inducing) across the previous day (or other such time interval). These data have only been collected relatively recently, and various alternative metrics have been suggested, such that the methodology for assessing these variables and their link with health and nutritional outcomes is still evolving. Another alternative method of data collection is to use food frequency questionnaires: these have the disadvantages that they are not quantitative and individuals may not recall as accurately over a longer period, but the advantage that better reflect the long-run intake of an individual. What is or is not healthy is frequently defined relative to WHO recommendations for a healthy diet for individuals 15-65 (WHO 2019, reproduced for convenience in Appendix Table 6).\\n\\nVarious metrics have been developed and used to assess whether diets of **adults aged 15 years and older, both men and women** are healthy, a literature which is evolving. Metrics have been developed to assess consumption of specific food groups associated with increased risk of negative nutrition/health outcomes (e.g. ultra-processed foods, sugar-sweetened beverages), as well as combinations of food groups that are recommended either for increased or limited consumption. The strengths and limitations of these indicators for this work are summarized in Table 3.\\n\\nThe Global Diet Quality Project provides new nationally representative data on prevalence of consumption of food groups for 37 LMIC (and up to 54 countries by 2023) (Global Diet Quality Project, 2022), using self-reported consumption of foods from different groups based on non-quantitative 24-hour recalls (Herforth et al. 2020). These data can be used to construct several metrics at the population level. The **Global Dietary Recommendations (GDR) Score** has two components, NCD-Protect and NCD-Risk.\\n\\n6\", \"datasets\": [{\"mention_name\": {\"text\": \"Diet diversity data\", \"confidence\": 0.77632737159729, \"start\": 88, \"end\": 107, \"clean_text\": \"Diet diversity data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 88, \"end\": 107, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 88, \"end\": 107, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 88, \"end\": 107}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 88, \"end\": 107, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6930081844329834, \"start\": 88, \"end\": 107}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.77632737159729, \"start\": 88, \"end\": 107}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.521797239780426, \"start\": 88, \"end\": 107}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.521797239780426, \"start\": 88, \"end\": 107}}, {\"mention_name\": {\"text\": \"DHS surveys\", \"confidence\": 0.519887387752533, \"start\": 1331, \"end\": 1342, \"clean_text\": \"DHS surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1331, \"end\": 1342, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1331, \"end\": 1342, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1331, \"end\": 1342}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1331, \"end\": 1342, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9964655637741089, \"start\": 586, \"end\": 597}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.519887387752533, \"start\": 586, \"end\": 597}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5794041156768799, \"start\": 586, \"end\": 597}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5794041156768799, \"start\": 586, \"end\": 597}}, {\"mention_name\": {\"text\": \"Gallup World Poll surveys\", \"confidence\": 0.7117467522621155, \"start\": 1472, \"end\": 1497, \"clean_text\": \"Gallup World Poll surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1472, \"end\": 1497, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1472, \"end\": 1497, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1472, \"end\": 1497}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1472, \"end\": 1497, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9964655637741089, \"start\": 727, \"end\": 752}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7117467522621155, \"start\": 727, \"end\": 752}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5794041156768799, \"start\": 727, \"end\": 752}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5794041156768799, \"start\": 727, \"end\": 752}}, {\"mention_name\": {\"text\": \"quantitative dietary intake data\", \"confidence\": 0.7596921324729919, \"start\": 1983, \"end\": 2015, \"clean_text\": \"quantitative dietary intake data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1983, \"end\": 2015, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1983, \"end\": 2015, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"2021-2022\", \"confidence\": 0.4688776135444641, \"start\": 1532, \"end\": 1541}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1983, \"end\": 2015, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.23674242198467255, \"start\": 463, \"end\": 495}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7596921324729919, \"start\": 463, \"end\": 495}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9727144837379456, \"start\": 463, \"end\": 495}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9727144837379456, \"start\": 463, \"end\": 495}}, {\"mention_name\": {\"text\": \"quantitative dietary survey data\", \"confidence\": 0.5910763144493103, \"start\": 2303, \"end\": 2335, \"clean_text\": \"quantitative dietary survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2303, \"end\": 2335, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2303, \"end\": 2335, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2303, \"end\": 2335}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2303, \"end\": 2335, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8174467086791992, \"start\": 783, \"end\": 815}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5910763144493103, \"start\": 783, \"end\": 815}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9118980169296265, \"start\": 783, \"end\": 815}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9118980169296265, \"start\": 783, \"end\": 815}}, {\"mention_name\": {\"text\": \"non-quantitative 24-hour recalls\", \"confidence\": 0.31746575236320496, \"start\": 4092, \"end\": 4124, \"clean_text\": \"non-quantitative 24-hour recalls\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4092, \"end\": 4124, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4092, \"end\": 4124, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4092, \"end\": 4124}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4092, \"end\": 4124, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.7503314018249512, \"start\": 893, \"end\": 925}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.31746575236320496, \"start\": 893, \"end\": 925}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7686870694160461, \"start\": 893, \"end\": 925}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7686870694160461, \"start\": 893, \"end\": 925}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"These measures have been validated for two countries (US and Brazil) against an index designed to reflect WHO healthy diet recommendations (2020) along with two additional recommendations from IARC (2018) and constructed using diet recall data (Herforth et al. 2020). Validation for a further eight countries is in the process of publication (Kennedy et al. 2022). The authors state that these metrics, combined with the MDD-W, can be used to assess current performance and track progress towards healthy diets. The only published study found to date that links these measures with outcomes is an observational study in China that assessed the association of the GDR scores with overweight and obesity in children 7-18 years of age, finding a positive association between the NCD-Risk score and obesity as well as a negative association between the overall GDR score and obesity (Wang et al. 2022).\\n\\nThe **Global Diet Quality Score (GDQS)** is a food-based metric composed of 25 food groups (16 healthy, 7 unhealthy and 2 unhealthy when consumed in excessive amounts) that are important contributors to nutrient intake and/or NCD risk (Bromage et al. 2021). It uses self-reported consumption based on a semiquantitative 24-hour recall. It is currently one of the only diet quality metrics that has been validated against both nutrient adequacy and diet-related NCD risk. A recent study found that the GDQS was better than other metrics (e.g. MDD- W and AHEI, the Alternative Healthy Eating Index from 2010) at predicting changes in weight and waist circumference among women in Mexico (Angulo et al. 2021). Another study for China found that the GDQS was inversely associated with metabolic syndrome and nutrient inadequacy or both (He et al. 2021). The use of the two submetrics allows one to distinguish between diets which are nutrient inadequate due to insufficiency in the diet, and diets which are nutrient-inadequate due to unhealthy food choices. However, this metric has not yet been widely integrated in national surveys and data availability for LMIC are very limited.\\n\\nThere are also two diet quality metrics derived from modeling approaches. The **Global Dietary Database** has sought to consolidate existing quantitative data and produce modeled estimates of individual food and nutrient intake for 188 countries (Miller et al. 2021). These estimates have been used to describe dietary quality at country, regional and global levels using various metrics (Miller et al. 2022) and quantify disease burdens attributable to diet, such as for type 2 diabetes (O’Hearn et al. 2023).\\n\\nThe **Global Burden of Disease Project** uses a set of **15 dietary risk factors** : diets low in: fiber; fruits; legumes; nuts and seeds; polyunsaturated fatty acids (PUFAs); seafood omega-3 fatty acids; vegetables and whole grains; milk; calcium; and diets high in: processed meat; red meat; sodium; sugar-sweetened beverages; and trans fatty acids; these include most of the recommendations from WHO with some additional ones for cancer from IARC (GBD 2017 Diet Collaborators, 2019). They develop Global Burden of Disease estimates of the contribution of dietary risks to the burden of NCDs in adults over the age of 25 across 195 countries, along with their associated impact on mortality and DALYs.\\n\\nOne critique of these modeled estimates points to the fact that evidence on dietary intake from representative populations in LMIC is currently inadequate to produce robust and reliable estimates of intake across countries (Beal et al. 2021). The relevance and face validity of modeled data for country-level work are also often questioned by national stakeholders. Another concern is that GBD 2017 Diet Collaborators’ (2019) estimates of adverse effects of dietary factors (11m deaths annually) far exceed the estimates from other studies cited above of adverse effects of overweight and obesity (4m globally). Some dietary risks affect health directly without necessarily being mediated by obesity, for example consumption of processed meats increases risk for colon cancer. There is also likely an indeterminate amount of double counting in the mortality estimates, given the “adding up” methodology employed (i.e., single dietary components are considered as independent and correlated structure or substitution effects not adequately considered). Diets do not tend to be randomly deficient in protective factors, or randomly contain single dietary risk factors. Rather, many households which have insufficient purchasing power are unable to consume several 7\", \"datasets\": [{\"mention_name\": {\"text\": \"diet recall data\", \"confidence\": 0.6498364210128784, \"start\": 227, \"end\": 243, \"clean_text\": \"diet recall data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 227, \"end\": 243, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 227, \"end\": 243, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 227, \"end\": 243}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 227, \"end\": 243, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.2848759889602661, \"start\": 227, \"end\": 243}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6498364210128784, \"start\": 227, \"end\": 243}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.775554895401001, \"start\": 227, \"end\": 243}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.775554895401001, \"start\": 227, \"end\": 243}}, {\"mention_name\": {\"text\": \"Global Dietary Database\", \"confidence\": 0.6605807542800903, \"start\": 2161, \"end\": 2184, \"clean_text\": \"Global Dietary Database\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2161, \"end\": 2184, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2161, \"end\": 2184, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2161, \"end\": 2184}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2161, \"end\": 2184, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.5902814865112305, \"start\": 568, \"end\": 591}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6605807542800903, \"start\": 568, \"end\": 591}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5276240110397339, \"start\": 568, \"end\": 591}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5276240110397339, \"start\": 568, \"end\": 591}}, {\"mention_name\": {\"text\": \"Global Burden of Disease estimates\", \"confidence\": 0.35210946202278137, \"start\": 3093, \"end\": 3127, \"clean_text\": \"Global Burden of Disease estimates\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3093, \"end\": 3127, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3093, \"end\": 3127, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3093, \"end\": 3127}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3093, \"end\": 3127, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9992792010307312, \"start\": 664, \"end\": 698}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.35210946202278137, \"start\": 664, \"end\": 698}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7819300889968872, \"start\": 664, \"end\": 698}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7819300889968872, \"start\": 664, \"end\": 698}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"protective dietary factors, in the same way that households with high energy intakes and high levels of ultraprocessed food consumption are likely to consume various of the higher-risk items and too few of the protective items.\\n\\nAnother summary measure based on quantitative diet data focuses on **ultra-processed foods** (Monteiro, Cannon, Lawrence et al. 2019; Monteiro, Cannon, Levy et al. 2019). Ultra-processed foods are defined by food processing methods as defined by the Nova classification and include foods such as convenience foods, fast foods, and sugar-sweetened beverages. Monteiro et al. (2019) summarize the relationship between amount of ultra-processed food in the diet and presence of diet risk factors in the diet (positive relation with saturated fat, sugar, trans fats, sodium and energy density) as well as the absence of protective factors in the diet (negative relation with protein, fiber and potassium). Monteiro et al. (2019) also summarize the relationship of amount of ultra-processed food in the diet, and suggest it is associated with increased obesity or obesity-related outcomes, CVD outcomes, increased prevalence of selected cancers. Recent systematic reviews also have found evidence of an increased risk of overweight and obesity, diabetes and other NCD risk factors associated with higher ultra-processed food consumption (Lane et al. 2021; Moradi et al. 2021).\\n\\nMetrics for unhealthy diets in children are at an early stage of development and currently rely on the monitoring of food groups that are shown to be associated with increased risk of overweight and obesity. In **young children 6-23 months** , zero vegetable or fruit consumption in the previous day is used to measure diet risk, a practice that is linked to both insufficient intake of nutrients presently as well as low intake later in life (WHO & UNICEF, 2021). Similarly, data on child consumption of sweet beverages and sentinel unhealthy foods<sup>2</sup> in the previous day/time interval are now included as unhealthy eating indicators in largescale surveys (i.e. MICS and DHS) based on evidence showing increased overweight/obesity risk among children consuming unhealthy snack foods and sugar-sweetened beverages (SSB) (Pries, Filteau, et al. 2019; Pries, Rehman, et al. 2019; Rousham et al. 2022).\\n\\nThere are no validated metrics for unhealthy diets in children 2-5 and 6-10 years. In LMIC, consumption of **unhealthy foods and SSBs** , as well as **insufficient intake of fruits and/or vegetables among adolescents** are monitored as health risk behaviors through Global School-based Student Health Surveys administered to school-going children 12-17 years of age (Beal et al. 2019). Eating fruit (Parra et al. 2021) or vegetables is associated with lower risk of overweight/obesity in children 12-17 years, while fast food and soft drinks consumption are associated with higher risk of overweight/obesity (Mahumud et al. 2021).\\n\\nThere are various **other characteristics of diet quality that are not assessed by the measures above** and are considered out of the scope of our work. These include several nutrition and health promoting components such as micronutrients added to fortified foods, dietary supplements, higher levels of nutrients in biofortified food products, and non-nutritive bioactive compounds (e.g. probiotics, polyphenols) present in maternal milk (Andreas et al. 2015; Badillo-Suárez et al. 2017) unprocessed or functional foods (Konstantinidi and Koutelidakis 2019). Similarly, food safety, including food borne contaminants, is beyond the scope of our work.\\n\\n> 2 “Sentinel unhealthy foods” are foods that are likely to be consumed by infants and young children and are high in sugar, salt and/or unhealthy fats. These include four categories of unhealthy foods: 1) candies, chocolates and other sugar confections; 2) frozen treats like ice cream, popsicles, etc.; 3) cakes, pastries, sweet biscuits, etc.; 4) chips, crisps, cheese puffs, French fries, instant noodles, etc.\\n\\n8\", \"datasets\": [{\"mention_name\": {\"text\": \"quantitative diet data\", \"confidence\": 0.9261667132377625, \"start\": 262, \"end\": 284, \"clean_text\": \"quantitative diet data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 262, \"end\": 284, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 262, \"end\": 284, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 262, \"end\": 284}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 262, \"end\": 284, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.4884527921676636, \"start\": 262, \"end\": 284}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9261667132377625, \"start\": 262, \"end\": 284}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5421572923660278, \"start\": 262, \"end\": 284}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5421572923660278, \"start\": 262, \"end\": 284}}, {\"mention_name\": {\"text\": \"data on child consumption\", \"confidence\": 0.693195104598999, \"start\": 1878, \"end\": 1903, \"clean_text\": \"data on child consumption\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1878, \"end\": 1903, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1878, \"end\": 1903, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1878, \"end\": 1903}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1878, \"end\": 1903, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.26033681631088257, \"start\": 1056, \"end\": 1081}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.693195104598999, \"start\": 1056, \"end\": 1081}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.46316760778427124, \"start\": 1056, \"end\": 1081}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.46316760778427124, \"start\": 1056, \"end\": 1081}}, {\"mention_name\": {\"text\": \"Global School-based Student Health Surveys\", \"confidence\": 0.3994998037815094, \"start\": 2578, \"end\": 2620, \"clean_text\": \"Global School-based Student Health Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2578, \"end\": 2620, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2578, \"end\": 2620, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2578, \"end\": 2620}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2578, \"end\": 2620, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9902830123901367, \"start\": 932, \"end\": 974}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3994998037815094, \"start\": 932, \"end\": 974}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7575883269309998, \"start\": 932, \"end\": 974}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7575883269309998, \"start\": 932, \"end\": 974}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"# 4. Previous Costing Studies\\n\\nA scoping review undertaken in parallel with this Working Paper (Siekmans et al, forthcoming) did not identify any published systematic reviews that bring together the evidence for LMIC on the economic costs of unhealthy diets. Candari and colleagues (2017) conducted a review of the literature to estimate the economic costs associated with unhealthy diets and low physical activity. They identified a total of six studies from Australia, China, the United Kingdom, and the United States that varied widely in their definition of unhealthy diets, including single food items, a list of specific food groups, and a dietary pattern compliant with national dietary recommendations. They reported a wide range of cost estimates that were sensitive to both the diet measure and study methodology.\\n\\nSince that review, several new studies have been published that estimate the burden of unhealthy food choice by quantifying the impact of individual dietary factors (dairy products, salt, sugar, sugar-sweetened beverages, ultra-processed foods) or poor-quality diets (measured by healthy eating indices) on diet-related NCD outcomes (cardio-vascular diseases, diabetes, chronic disease). For LMIC, several studies estimate the cost (mortality and DALYs) of cardiovascular disease or the NCD burden attributable to the dietary risk factors, using the Global Burden of Disease dataset for 204 countries (Liu et al. 2022; Qiao et al. 2022; Zhang et al. 2023) or in specific country contexts (Brazil, Mexico) (Dávila-Cervante, 2020; Machado et al. 2022). Four studies (three in Brazil, one in Costa Rica) estimated the costs associated with excess salt/sodium consumption, one using the Global Burden of Disease dataset (Guedes et al. 2022) and the others assessed health care costs associated with hypertension (Nilson, da Silva et al. 2020; Nilson, Metlzer, et al. 2020; Vega-Solano et al. 2023). An additional eight studies were found that estimated costs (usually mortality and DALYs) associated with specific dietary risk factors such as diets high in red meat (Liu et al. 2022) or processed meat (Rocha et al. 2023), high in sugar-sweetened beverage intake (Bardach et al. 2023; Li et al. 2021), low in fiber (Zhuo et al. 2022), or low in dairy (Javanbakht et al. 2018). One problem with these studies of individual dietary risks, is that it is not possible to add them up to obtain overall estimates of the costs of unhealthy diets, since the risk factors may be correlated and/or interact.\\n\\nA global Cost of Not Breastfeeding tool is also available to support country quantification of the human and economic losses, including loss of life, lost productivity and increased costs to health systems. The total annual global losses associated with inadequate breastfeeding are estimated to be between US$ 257-341 billion, or 0.37-0.70% of global gross national income (Walters et al. 2019).\\n\\nThere are a sizeable number of studies which estimate the cost of undernutrition, as measured by stunting and/or wasting among children under five years of age. A couple of recent examples, comprehensive in geographic scope, include ECLAC and WFP (2017) and the Cost of Hunger in Africa series, supported by African Union and led by at least 21 individual countries performing national studies (https://www.wfp.org/publications/cost-hunger-africa-series). The main costs included are costs of morbidity, mortality, and cognitive losses associated with undernutrition, measured as the costs of foregone future productivity. To these, the costs of treating severe acute malnutrition or community management of acute malnutrition can be added. These estimates typically focus on children (0-23 months, or 0-59 months) where the impact of diet on the life course is primarily set, and then make lifetime projections on DALY and productivity costs. The mother’s diet during pregnancy is also a factor, since that impacts birthweight (i.e., through premature delivery and/or intra-uterine growth retardation).\\n\\nEstimating the cost of overweight and obesity in LMICs has been undertaken more recently, with examples being ECLAC & WFP (2017), Jackson-Morris and Meyer (2023), Okunogbe et al. (2022) and Brero et al. (2023). The main costs considered are mortality and morbidity associated with noncommunicable diseases, 9\", \"datasets\": [{\"mention_name\": {\"text\": \"healthy eating indices\", \"confidence\": 0.5526431202888489, \"start\": 1105, \"end\": 1127, \"clean_text\": \"healthy eating indices\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1105, \"end\": 1127, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1105, \"end\": 1127, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1105, \"end\": 1127}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1105, \"end\": 1127, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9398890137672424, \"start\": 1105, \"end\": 1127}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5526431202888489, \"start\": 1105, \"end\": 1127}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.702231228351593, \"start\": 1105, \"end\": 1127}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.702231228351593, \"start\": 1105, \"end\": 1127}}, {\"mention_name\": {\"text\": \"Global Burden of Disease dataset\", \"confidence\": 0.9449612498283386, \"start\": 1708, \"end\": 1740, \"clean_text\": \"Global Burden of Disease dataset\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1708, \"end\": 1740, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1708, \"end\": 1740, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1708, \"end\": 1740}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1708, \"end\": 1740, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9165948033332825, \"start\": 844, \"end\": 876}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.9449612498283386, \"start\": 844, \"end\": 876}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8371632695198059, \"start\": 844, \"end\": 876}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8371632695198059, \"start\": 844, \"end\": 876}}, {\"mention_name\": {\"text\": \"Cost of Hunger in Africa series\", \"confidence\": 0.38526734709739685, \"start\": 3180, \"end\": 3211, \"clean_text\": \"Cost of Hunger in Africa series\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3180, \"end\": 3211, \"clean_text\": \"\"}, \"producer\": {\"text\": \"African Union\", \"confidence\": 0.3295508623123169, \"start\": 3226, \"end\": 3239, \"clean_text\": \"African Union\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3180, \"end\": 3211}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3180, \"end\": 3211, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.8011071085929871, \"start\": 865, \"end\": 896}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.38526734709739685, \"start\": 865, \"end\": 896}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.9582886695861816, \"start\": 865, \"end\": 896}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9582886695861816, \"start\": 865, \"end\": 896}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"for which overweight/obesity is a major risk factor. These costs include treatment costs, disability-adjusted life-year (DALY) losses, and productivity losses. Table 4 summarizes the major costs included in previous studies, with brief descriptions of how they are calculated. It is proposed to use similar standard methods in the methodology proposed in the next section.\\n\\n# 5. A Proposed Methodology for Estimating Economic Costs of Unhealthy Diets at the Country Level\\n\\nMalnutrition in all its forms (as measured by undernutrition/overweight/obesity) is not synonymous with unhealthy diets. There are individuals who have impaired growth despite nutrient-adequate diets because of other factors. Infection (itself related to vaccine status and the water and sanitation environment) is a key factor (Figure 2). Similarly, there are individuals who are overweight/obese despite their diet because of other factors, such as exposure to an obesogenic environment (Figure 2). Finally, having a healthy weight does not imply having a healthy diet. Individuals may maintain a healthy weight while consuming a diet that is inadequate in micronutrients or not balanced or not moderated. One example of individuals with healthy weight but unhealthy diets is those consuming excess salt which can increase hypertension and risks of cardiovascular disease.\\n\\nHence, although it is useful to employ cost methodologies similar to those of existing studies of undernutrition and overweight/obesity, adjustments are required since healthy weight does not imply the diet is healthy, nor is the converse true.\\n\\nFigure 3 illustrates this. The figure illustrates a hypothetical distribution of nutritional status in the population (in this case the population mean weight is in the healthy range, but that is not necessarily the case). The vertical lines delineate the population fractions who are with underweight, healthy weight, and overweight/obese. The dotted horizontal lines separate those within each nutritional status who have healthy diets and those who do not. We will use these fractions to estimate the associated economic costs, using data on stunting in children under 5 years of age to project productivity costs across the life course, and using data on adults 18 years and over on overweight/obesity since the vast majority of productivity losses are in this age group. Hence while the FOLU (2019) study equates the combined costs of underweight and overweight/obesity on morbidity and mortality to human costs of the food system, Figure 3 shows that the situation is more complex. However, estimating the relative proportions of those with undernutrition, of healthy weight and overweight/obese with healthy diets, is not trivial.\\n\\nWe propose the following methods. To distinguish the share of child undernutrition associated with inadequate dietary intake and that associated with other factors such as infection, we will use the Lives Saved Tool (LiST: <u>https://www.livessavedtool.org). We can then attribute this cost share (projecting</u> forward to lifetime productivity losses) to estimate the proportion of the costs of child undernutrition attributable to unhealthy diets.\\n\\nNCD costs (both treatment costs and DALY/productivity costs) among the adult population (18 years and older) are the major cost associated with unhealthy diets for those who are healthy weight and those who are overweight/obese. To estimate the proportion of NCD costs attributable to unhealthy diets we propose to use the following two equations:\\n\\n- (1) O&O = f(food choice, other)\\n\\n- (2) NCD risk = g(O&O, food choice, other)\\n\\n10\", \"datasets\": [{\"mention_name\": {\"text\": \"data on stunting in children\", \"confidence\": 0.4588017463684082, \"start\": 2132, \"end\": 2160, \"clean_text\": \"data on stunting in children\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2132, \"end\": 2160, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2132, \"end\": 2160, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2132, \"end\": 2160}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2132, \"end\": 2160, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8854606747627258, \"start\": 385, \"end\": 413}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.4588017463684082, \"start\": 385, \"end\": 413}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8041273951530457, \"start\": 385, \"end\": 413}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8041273951530457, \"start\": 385, \"end\": 413}}, {\"mention_name\": {\"text\": \"data on adults 18 years and over\", \"confidence\": 0.38441798090934753, \"start\": 2246, \"end\": 2278, \"clean_text\": \"data on adults 18 years and over\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2246, \"end\": 2278, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2246, \"end\": 2278, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2246, \"end\": 2278}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2246, \"end\": 2278, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8854606747627258, \"start\": 499, \"end\": 531}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.38441798090934753, \"start\": 499, \"end\": 531}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8041273951530457, \"start\": 499, \"end\": 531}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8041273951530457, \"start\": 499, \"end\": 531}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"where O&O stands for overweight and obesity, f and g represent functions and food choice is a measure of how healthy the diet is. This will help to identify the proportion of overweight/obesity attributable to unhealthy diet, and also those NCD risks not mediated by overweight and obesity.\\n\\nThere are existing studies linking dietary risks to burden of NCDs in LMICs, reviewed in the previous section. Most of these studies either look at a single NCD, or a single dietary factor. The Global Burden of Disease project has however a comprehensive study, aiming to link 15 dietary factors with the NCD burden (GBD 2017 Diet Collaborators, 2019). Various concerns have been raised about these estimates. First, the relative ranking of different dietary factors changes somewhat markedly between the analysis using the 2017 data and that using the 2019 data (analyzed by Qiao et al. 2022): Stanton et al. (2022) comment on differences regarding red meat. Second, evidence on dietary intake from representative populations in LMIC are currently inadequate to produce robust and reliable estimates of intake across countries (Beal et al. 2021). A third concern conceptually is how the rest of the diet is adjusted in the model when one individual dietary factor is varied in order to understand its impact on NCD risk. GBD 2017 Diet Collaborators (2019) state that the underlying studies they use are unclear on this adjustment. Finally, GBD 2017 Diet Collaborators (2019) estimates of adverse effects of dietary factors (11m global deaths annually) far exceed the estimates from other studies cited above of adverse effects of overweight and obesity (4m globally); and GBD 2017 Diet Collaborators (2019) imply that the 11m do not include the 4m. One likely explanation for the last point is that of double counting. The diets that individuals consume do not necessarily have only one dietary risk factor, and hence these individuals are at higher risk for more than one NCD. Adding the risks across multiple trials for individual NCDs is therefore likely to overestimate dietary risks of NCDs in aggregate.\\n\\nIt may be possible to reduce the double-counting on the side of dietary factors by using a more parsimonious indicator of dietary quality, such as the GDRS-protect and GDRS-risk scores (for cross-country data), or the two GDQS sub-scores (for national data). We propose to try this approach.\\n\\nThis does not solve the problem of multiple NCDs, however. Although deaths from NCDs list a primary cause of death, individuals suffering from (and being treated for) NCDs may have multiple conditions, but the individual should only be counted once when calculating treatment costs and lost productivity costs. The pragmatic approach suggested here is to focus on two of the top two conditions in terms of cause of death globally, namely cardiovascular disease and diabetes. Cancer could be included in future, as ongoing prospective cohort studies provide more information on the dietary risks for a broader range of cancers. This may yield an overly conservative estimate of the impact of unhealthy diets on NCDs, but the underlying rule on such economic calculations is always to err on the side of conservatism in such estimates.\\n\\n# 6. Preliminary Test of the Methodology Using Cross-Country Data\\n\\nSome preliminary regressions were undertaken using cross-country data, while recognizing that analysis using data for individuals and households within one country would be preferable (discussed in the section immediately following this one). This section discusses results of using cross-country data, benefiting from the recent availability of data on several diet quality measures across over 50 LMIC. Data definitions are given in Table 5. Stata 17 ((https://stata.com) was used for the analysis, and the dataset created using data from public databases is available through the World Bank Microdata Library (https://dimewiki.worldbank.org/Microdata_Catalog). Cross-country data have several disadvantages, among them:\\n\\n- Variables are not always available for the same year (see Table 5);\\n\\n11\", \"datasets\": [{\"mention_name\": {\"text\": \"2019 data\", \"confidence\": 0.4733637571334839, \"start\": 845, \"end\": 854, \"clean_text\": \"2019 data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 845, \"end\": 854, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 845, \"end\": 854, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 845, \"end\": 854}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 845, \"end\": 854, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.36484771966934204, \"start\": 29, \"end\": 38}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.4733637571334839, \"start\": 29, \"end\": 38}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8077048659324646, \"start\": 29, \"end\": 38}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8077048659324646, \"start\": 29, \"end\": 38}}, {\"mention_name\": {\"text\": \"GDRS-risk scores\", \"confidence\": 0.4864804446697235, \"start\": 2272, \"end\": 2288, \"clean_text\": \"GDRS-risk scores\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2272, \"end\": 2288, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2272, \"end\": 2288, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2272, \"end\": 2288}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2272, \"end\": 2288, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9974848031997681, \"start\": 660, \"end\": 676}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4864804446697235, \"start\": 660, \"end\": 676}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8173341155052185, \"start\": 660, \"end\": 676}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8173341155052185, \"start\": 660, \"end\": 676}}, {\"mention_name\": {\"text\": \"prospective cohort studies\", \"confidence\": 0.3597998023033142, \"start\": 2919, \"end\": 2945, \"clean_text\": \"prospective cohort studies\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2919, \"end\": 2945, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2919, \"end\": 2945, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2919, \"end\": 2945}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2919, \"end\": 2945, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.8157877326011658, \"start\": 502, \"end\": 528}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3597998023033142, \"start\": 502, \"end\": 528}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4569537937641144, \"start\": 502, \"end\": 528}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4569537937641144, \"start\": 502, \"end\": 528}}, {\"mention_name\": {\"text\": \"Cross-Country Data\", \"confidence\": 0.9344232082366943, \"start\": 3279, \"end\": 3297, \"clean_text\": \"Cross-Country Data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3279, \"end\": 3297, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3279, \"end\": 3297, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3279, \"end\": 3297}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3279, \"end\": 3297, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6916797757148743, \"start\": 862, \"end\": 880}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9344232082366943, \"start\": 862, \"end\": 880}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5006943941116333, \"start\": 862, \"end\": 880}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5006943941116333, \"start\": 862, \"end\": 880}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"- Variables are not always available for the same population group, for example the most complete recent data on underweight/overweight (measured using BMI) in adults are for non-pregnant women 20-49, whereas hypertension and diabetes prevalence are available for all adults 18+;\\n\\n- It is impossible to control for confounding variables;\\n\\n- There may be differences in reporting methods for the same variables;\\n\\n- Current prevalence of overweight in adults does not only depend on current diet, but on the diet trajectory over the life course to date;\\n\\n- Risks of NCD similarly do not only depend on current diet, but on the diet trajectory over the life course to date; and\\n\\n- Time trends (in per capita calorie supply and overweight/obesity for example) may mean that different age cohorts in the population have different health risks, such that analysis should be undertaken for narrower age cohorts and not for all adults combined.\\n\\nHence, the results from cross-country regressions presented here are only exploratory, pending more careful analysis using individual country data. Sample sizes for the regressions are modest (generally around 40 countries). A variety of specifications were tried and those presented here had the best fit (using AIC, Akaike Information Criterion to decide among different specifications using the same observations). Independent variables which were not significant and whose inclusion worsened the overall fit, were removed.\\n\\nThe initial results (summarized in tables 6, 7 and 8) are generally consistent with the literature on diet quality, although not all the independent variables are significant. Child stunting and women’s low BMI decrease as per capita calorie supply increases (significantly in both cases) and as diet diversity increases (significant for women but not children). Child wasting is not responsive to per capita calorie supply but decreases significantly as diet diversity increases. Wasting tends to reflect recent factors (e.g. seasonal hunger or epidemics) more sensitively than does stunting. The diet diversity variable may reflect circumstances at time of collection of anthropometric data, while calorie availability represents an annual average.\\n\\nWomen’s and children’s overweight is positively associated with per capita calorie supply (unsurprisingly) and also with dietary diversity (MDD and MDD-W), although not significantly. This latter finding is consistent with Herforth et al. (2020) who cite two previous studies noting the positive association between dietary energy intake with dietary diversity. The coefficients on the GDRS sub-scores (NCD-Protect and NCD-Risk) have the expected sign for women’s overweight but are not significant. There are regional differences, in particular the Southeast Asia region has significantly higher stunting and wasting for children and low BMI for women, controlling for other factors (compared to the Africa region). The Eastern Mediterranean region has significantly more overweight both in children and women, controlling for other variables (compared to the Africa region). The Gini coefficient, a measure of inequality, is not significant in any of the regressions. The effect of calorie supply per capita is non-linear and has the strongest impact at the lower levels of calorie availability.\\n\\nResults for hypertension and diabetes are mixed, perhaps because the data disadvantages mentioned above are more severe (less consistency of date, lack of age cohort data, and inconsistencies in the population covered, e.g. dependent variables are for all adults whereas some of the independent variables are only for women, etc.). Hypertension and diabetes both go down as ln(per capita calorie supply) increases and (results not presented) average level of hypertension is highest in low income countries and decreases consistently from low to lower-middle to upper-middle to high income countries. In theory this variable is supposed to 12\", \"datasets\": [{\"mention_name\": {\"text\": \"individual country data\", \"confidence\": 0.7877562642097473, \"start\": 1061, \"end\": 1084, \"clean_text\": \"individual country data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1061, \"end\": 1084, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1061, \"end\": 1084, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1061, \"end\": 1084}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1061, \"end\": 1084, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.3732813596725464, \"start\": 251, \"end\": 274}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7877562642097473, \"start\": 251, \"end\": 274}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.6433143615722656, \"start\": 251, \"end\": 274}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6433143615722656, \"start\": 251, \"end\": 274}}, {\"mention_name\": {\"text\": \"anthropometric data\", \"confidence\": 0.7324308753013611, \"start\": 2139, \"end\": 2158, \"clean_text\": \"anthropometric data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2139, \"end\": 2158, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2139, \"end\": 2158, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2139, \"end\": 2158}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2139, \"end\": 2158, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.29027774930000305, \"start\": 446, \"end\": 465}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7324308753013611, \"start\": 446, \"end\": 465}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.49531298875808716, \"start\": 446, \"end\": 465}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.49531298875808716, \"start\": 446, \"end\": 465}}, {\"mention_name\": {\"text\": \"age cohort data\", \"confidence\": 0.9336865544319153, \"start\": 3472, \"end\": 3487, \"clean_text\": \"age cohort data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3472, \"end\": 3487, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3472, \"end\": 3487, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3472, \"end\": 3487}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3472, \"end\": 3487, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.6311134696006775, \"start\": 17, \"end\": 32}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9336865544319153, \"start\": 17, \"end\": 32}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.35607004165649414, \"start\": 17, \"end\": 32}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.35607004165649414, \"start\": 17, \"end\": 32}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"measure both untreated hypertension (using blood pressure in the survey sample) but also to include treated hypertension (presumably by asking the respondent about medications at the time blood pressure is measured). It is possible that surveys are less good at capturing treated hypertension. The NCD-Protect score has the expect effect on reducing hypertension (significant) and is negative but not significant in the equation for diabetes. There are significant regional differences for diabetes. The NCD variables are likely to be affected by the quality of the health system. NCDs are more likely to be detected and treated in higher income countries. Thus, although one might expect incidence of NCDs to rise with income, premature mortality from NCDs (prior to age 70) is higher in LMICs. The results, given the data constraints noted, provide some encouragement for finding the expected effects of diet quality variables within individual countries.\\n\\n# 7. Potential Testing of the Methodology Using National Datasets\\n\\nNational health and nutrition survey data provide potential for further testing of the methodology. National data have several advantages. Within countries there is less variation in the ability of health services to screen and treat diseases including NCDs, although there remain problems of socio-economic gradients and rural/urban differences affecting decisions to seek care and obtain access. In a large dataset, it is possible to conduct analysis separately for different age cohorts. Given the progression of the nutritional transition, different age cohorts will face different conditions throughout their lifespan, and cross section data can permit modeling of this. Six datasets were noted above as being potentially suitable, namely health and nutrition surveys for Brazil, Mexico, China, Philippines, Andhra Pradesh (India), and South Africa. These include three lower-middle income countries/states and three upper-middle income countries.\\n\\nA disadvantage is that, to our knowledge, there are no low-income countries with such datasets. A suggested methodology for how to apply this to low-income countries in the same region is as follows. One could use the health and nutrition dataset from a given region to estimate the probability of cardiovascular risk and diabetes risk (via hypertension and blood glucose) among adults over 25 by BMI, by age cohorts and by socioeconomic group. These rates could then be applied to similar socioeconomic groups in low-income countries who, it is predicted, will follow similar paths.\\n\\nThis is possible for cardiovascular disease, the number one cause of death globally, as well as for diabetes. Although it would be desirable to extend this to the number two risk, all cancers combined, making estimates is more difficult. Unlike cardiovascular disease, there are no readily measurable biomarkers indicating longer-term risk, and the term cancer covers a large number of heterogeneous conditions affecting different sites such that prospective studies of cancer require very large populations. There are fewer large scale prospective cohort studies of cancer than there are involving CVD. Our understanding of the dietary risks for cancer is still evolving.\\n\\n# 8. Potential for Testing Using Prospective Cohort Studies\\n\\nProspective cohort studies have become more common as researchers note their usefulness for NCD conditions whose antecedents begin even prior to adulthood, but which mainly manifest in disease outcomes three or more decades later. Ongoing prospective cohort studies which include dietary recall and health measures include PURE (in 27 high, middle and low income countries: <u>https://www.phri.ca/research/pure/),</u> ELSA-Brazil (began 2008-2010: Aquino et al. 2012), the US Nurses Health Study (https://nurseshealthstudy.org), the EPIC study <u>(https://iarc.epic.fr) in a number of European countries,</u> 13\", \"datasets\": [{\"mention_name\": {\"text\": \"National Datasets\", \"confidence\": 0.8256396055221558, \"start\": 1007, \"end\": 1024, \"clean_text\": \"National Datasets\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1007, \"end\": 1024, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1007, \"end\": 1024, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1007, \"end\": 1024}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1007, \"end\": 1024, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.3617156445980072, \"start\": 1007, \"end\": 1024}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8256396055221558, \"start\": 1007, \"end\": 1024}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5383461117744446, \"start\": 1007, \"end\": 1024}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5383461117744446, \"start\": 1007, \"end\": 1024}}, {\"mention_name\": {\"text\": \"National health and nutrition survey data\", \"confidence\": 0.6946585774421692, \"start\": 1026, \"end\": 1067, \"clean_text\": \"National health and nutrition survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1026, \"end\": 1067, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1026, \"end\": 1067, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1026, \"end\": 1067}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1026, \"end\": 1067, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9920487999916077, \"start\": 183, \"end\": 224}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6946585774421692, \"start\": 183, \"end\": 224}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.6951128244400024, \"start\": 183, \"end\": 224}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6951128244400024, \"start\": 183, \"end\": 224}}, {\"mention_name\": {\"text\": \"health and nutrition dataset\", \"confidence\": 0.9252433776855469, \"start\": 2198, \"end\": 2226, \"clean_text\": \"health and nutrition dataset\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2198, \"end\": 2226, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2198, \"end\": 2226, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2198, \"end\": 2226}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2198, \"end\": 2226, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9678220152854919, \"start\": 399, \"end\": 427}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.9252433776855469, \"start\": 399, \"end\": 427}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7215175032615662, \"start\": 399, \"end\": 427}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7215175032615662, \"start\": 399, \"end\": 427}}, {\"mention_name\": {\"text\": \"PURE\", \"confidence\": 0.3501338064670563, \"start\": 3623, \"end\": 3627, \"clean_text\": \"PURE\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3623, \"end\": 3627, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3623, \"end\": 3627, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3623, \"end\": 3627}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3623, \"end\": 3627, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.4183133840560913, \"start\": 989, \"end\": 993}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3501338064670563, \"start\": 989, \"end\": 993}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7381734251976013, \"start\": 989, \"end\": 993}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7381734251976013, \"start\": 989, \"end\": 993}}, {\"mention_name\": {\"text\": \"ELSA-Brazil\", \"confidence\": 0.36492234468460083, \"start\": 3718, \"end\": 3729, \"clean_text\": \"ELSA-Brazil\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3718, \"end\": 3729, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3718, \"end\": 3729, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3718, \"end\": 3729}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3718, \"end\": 3729, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.45019835233688354, \"start\": 1084, \"end\": 1095}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.36492234468460083, \"start\": 1084, \"end\": 1095}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6267544031143188, \"start\": 1084, \"end\": 1095}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6267544031143188, \"start\": 1084, \"end\": 1095}}, {\"mention_name\": {\"text\": \"US Nurses Health Study\", \"confidence\": 0.6054548025131226, \"start\": 3773, \"end\": 3795, \"clean_text\": \"US Nurses Health Study\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3773, \"end\": 3795, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3773, \"end\": 3795, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3773, \"end\": 3795}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3773, \"end\": 3795, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8966956734657288, \"start\": 1139, \"end\": 1161}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6054548025131226, \"start\": 1139, \"end\": 1161}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.34257709980010986, \"start\": 1139, \"end\": 1161}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.34257709980010986, \"start\": 1139, \"end\": 1161}}, {\"mention_name\": {\"text\": \"EPIC study\", \"confidence\": 0.42888200283050537, \"start\": 3833, \"end\": 3843, \"clean_text\": \"EPIC study\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3833, \"end\": 3843, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3833, \"end\": 3843, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3833, \"end\": 3843}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3833, \"end\": 3843, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.8826554417610168, \"start\": 1199, \"end\": 1209}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.42888200283050537, \"start\": 1199, \"end\": 1209}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5939505100250244, \"start\": 1199, \"end\": 1209}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5939505100250244, \"start\": 1199, \"end\": 1209}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"focusing on cancer, the Young Lives study in 4 LMICs (https://www.younglives.org.uk) and HELTI - (https://cihr <u>irsc.gc.ca/e/49510.html) with linked studies in four middle and high-income countries, among</u> others. It might be worth exploring some of these regarding cancer risks.\\n\\nProspective cohort studies do have strengths and limitations: they may aim to measure “usual” consumption of food groups, rather than a 24-hour recall as in most cross-sectional surveys; however, unless the cohort is recruited in early adulthood, current diet may not be a good reflection of previous diet. Prospective cohorts have advantages over cross-sectional studies, where an individual’s diagnosis with elevated blood pressure or blood glucose may have led the individual already to change diet and thus attenuate the estimated link between diet and NCD risk.\\n\\n# 9. Data Limitations\\n\\nWhile we propose a life cycle approach, we acknowledge that certain age groups are not adequately covered in terms of evidence and dietary data. These are: children from 2 to 5 years, primary school children (6 to 10-12 years) and older adults (over 65 years). Evidence and data for adolescents is limited but growing. Dietary data for adult males (15 and above) are significantly less covered than for women.\\n\\nData on MDD-W are growing both through DHS and the Diet Quality Project (2022) but there appear to be some differences in the national prevalence estimates, some of which may reflect different sampling and data collection methods (face-to-face, as compared to by telephone). Few countries currently have opendata access for DHS MDD-W. MDD-W was not designed to measure risks of obesity and NCDs, and Herforth et al (2021) advise that caution is needed in using this in high income countries to indicate nutrient adequacy. The two GDRS subcomponent scores (NCD-Protect or NCD-Risk) perform better than MDDW in relation to compliance with WHO dietary recommendations (Herforth et al. 2020). In general, the underlying continuous food score data are preferable as they provide more information than a dichotomous summary of a score, although this may require going back to the original datasets rather than the summary statistics available in online databases.\\n\\nCaution should be exercised when linking diet metrics with the burden of malnutrition and diet-related NCDs as many of these dietary metrics still need additional validation with health outcomes. This issue should be further explored when applying the approach at the country level.\\n\\n# 10. Discussion and Conclusions\\n\\nEvidence on the hidden costs of undernutrition in children is well understood, and there is a large literature on the cost-benefit of investing in reducing undernutrition in young children, implemented in actions aimed at reaching the Millennium and Sustainable Development Goals. It is also well understood that unhealthy diets are associated with increased NCD risk. However, despite policy efforts, there were no examples of countries which have successfully reduced overweight and obesity out of 188 countries studied over 33 years from 1980 to 2013 (Ng et al. 2014). Short- and medium-term solutions to reduce undernutrition have in some cases failed to emphasize the basic principles of a healthy diet, namely adequacy, diversity, balance, and moderation.\\n\\nAlthough current methodologies for estimating the hidden costs of child undernutrition and of adult overweight and obesity are informative, it is important to note that other risk factors are at play beyond unhealthy diets. Anthropometric measurements such as weight cannot entirely be associated with diets. For 14\", \"datasets\": [{\"mention_name\": {\"text\": \"LMICs\", \"confidence\": 0.3385308086872101, \"start\": 47, \"end\": 52, \"clean_text\": \"LMICs\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.3214167058467865, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 47, \"end\": 52, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 47, \"end\": 52}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 47, \"end\": 52, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.5610897541046143, \"start\": 47, \"end\": 52}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3385308086872101, \"start\": 47, \"end\": 52}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7535260915756226, \"start\": 47, \"end\": 52}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7535260915756226, \"start\": 47, \"end\": 52}}, {\"mention_name\": {\"text\": \"DHS MDD-W\", \"confidence\": 0.5381919741630554, \"start\": 1612, \"end\": 1621, \"clean_text\": \"DHS MDD-W\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1612, \"end\": 1621, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1612, \"end\": 1621, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1612, \"end\": 1621}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1612, \"end\": 1621, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.331621915102005, \"start\": 815, \"end\": 824}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5381919741630554, \"start\": 815, \"end\": 824}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.56683748960495, \"start\": 815, \"end\": 824}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.56683748960495, \"start\": 815, \"end\": 824}}, {\"mention_name\": {\"text\": \"continuous food score data\", \"confidence\": 0.7615926265716553, \"start\": 2004, \"end\": 2030, \"clean_text\": \"continuous food score data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2004, \"end\": 2030, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2004, \"end\": 2030, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2004, \"end\": 2030}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2004, \"end\": 2030, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.30612748861312866, \"start\": 417, \"end\": 443}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7615926265716553, \"start\": 417, \"end\": 443}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8073082566261292, \"start\": 417, \"end\": 443}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8073082566261292, \"start\": 417, \"end\": 443}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"example, despite being given an adequate diet, children may still face an impaired growth or not gain weight because of underlying inflammations linked to infection. Conversely, individuals with healthy weight may have unhealthy diets and be at high risk of NCDs. Hence, estimating the hidden costs of unhealthy diets differs from estimating the cost of malnutrition in all its forms.\\n\\nThe new suites of healthy diet metrics have promise for policy analysis. They have potential for tracking the impact of nutrition-sensitive agrifood system policies, especially if they go beyond a focus on supply chain and also consider food environments and consumer behaviors. Using a more holistic indicator of healthy diets can avoid the double counting problems inherent in trying to add up the costs of failing to reach a dozen or more dietary guidelines. Work is underway to further validate these new measures against health outcomes.\\n\\nRegressions undertaken for the current study using cross-country data suggest that increased calorie availability and greater dietary diversity, as expected, tend to be associated with less child stunting and wasting, and less underweight in women. Calorie availability and dietary diversity also are associated with increased overweight and obesity in both children and women, although these results are less significant than for undernutrition. The new GDR score’s submetrics, NCD-Protect and NCD-Risk, have the expected effects on obesity in women, although not significantly so. Finally, cross-country data on NCD risks (hypertension and diabetes) have measurement issues but have the expected negative relation (although significant only for hypertension) with the NCD-Protect score.\\n\\nIn this note we have suggested that a way to proceed further would be to partner with national researchers to obtain similar regression results using national (or state-level) health and nutrition survey data to elucidate the relationship between overweight and/or cardiovascular risk or diabetes, and the new diet quality variables from individual-level data. This could then inform analysis for other countries in the same region of policy interest, which have individual quantitative dietary recall data but not associated with data on NCD risk factors.\\n\\nThe new data quality variables could help identify how diets vary across different population subgroups, and what the key issues are, such as insufficient diet diversity, insufficient consumption of foods protective against NCDs, or excess consumption of foods with risks for NCDs.\\n\\nKnowledge of the economic costs of unhealthy food choices for different population subgroups could then inform policy priorities and be compared with estimates of the costs of different agrifood policies aimed at supporting better food choices. Examples of agrifood policies which were adopted with short- and mediumterm goals of increasing farmer income and meeting nutrient needs of undernourished populations are subsidies on sugarcane and palm oil. These subsidies aimed to provide inexpensive calories and fat. However, being cognizant of the long-run costs of these policies in the context of the nutrition transition, with growing risks of diabetes and risks associated with saturated fat, can help guide the evolution of agrifood policy. There are also implications for public health nutrition policy, in that information on costs can underscore the importance of policies such as nutrition information and communication, and nutrition and exercise interventions in school-age children and adolescents.\\n\\n15\", \"datasets\": [{\"mention_name\": {\"text\": \"cross-country data\", \"confidence\": 0.8948037624359131, \"start\": 981, \"end\": 999, \"clean_text\": \"cross-country data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 981, \"end\": 999, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 981, \"end\": 999, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 981, \"end\": 999}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 981, \"end\": 999, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.4381208121776581, \"start\": 981, \"end\": 999}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8948037624359131, \"start\": 981, \"end\": 999}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7523417472839355, \"start\": 981, \"end\": 999}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7523417472839355, \"start\": 981, \"end\": 999}}, {\"mention_name\": {\"text\": \"cross-country data on NCD risks\", \"confidence\": 0.5535698533058167, \"start\": 1522, \"end\": 1553, \"clean_text\": \"cross-country data on NCD risks\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1522, \"end\": 1553, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1522, \"end\": 1553, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1522, \"end\": 1553}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1522, \"end\": 1553, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.5140917897224426, \"start\": 646, \"end\": 677}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.5535698533058167, \"start\": 646, \"end\": 677}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.43182024359703064, \"start\": 646, \"end\": 677}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.43182024359703064, \"start\": 646, \"end\": 677}}, {\"mention_name\": {\"text\": \"health and nutrition survey data\", \"confidence\": 0.5813485980033875, \"start\": 1896, \"end\": 1928, \"clean_text\": \"health and nutrition survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1896, \"end\": 1928, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1896, \"end\": 1928, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1896, \"end\": 1928}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1896, \"end\": 1928, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9433243274688721, \"start\": 1020, \"end\": 1052}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5813485980033875, \"start\": 1020, \"end\": 1052}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6022519469261169, \"start\": 1020, \"end\": 1052}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6022519469261169, \"start\": 1020, \"end\": 1052}}, {\"mention_name\": {\"text\": \"individual-level data\", \"confidence\": 0.3515755534172058, \"start\": 2058, \"end\": 2079, \"clean_text\": \"individual-level data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2058, \"end\": 2079, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2058, \"end\": 2079, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2058, \"end\": 2079}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2058, \"end\": 2079, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9433243274688721, \"start\": 1182, \"end\": 1203}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3515755534172058, \"start\": 1182, \"end\": 1203}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6022519469261169, \"start\": 1182, \"end\": 1203}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6022519469261169, \"start\": 1182, \"end\": 1203}}, {\"mention_name\": {\"text\": \"individual quantitative dietary recall data\", \"confidence\": 0.46675050258636475, \"start\": 2183, \"end\": 2226, \"clean_text\": \"individual quantitative dietary recall data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2183, \"end\": 2226, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2183, \"end\": 2226, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2183, \"end\": 2226}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2183, \"end\": 2226, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.2851123809814453, \"start\": 390, \"end\": 433}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.46675050258636475, \"start\": 390, \"end\": 433}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6005488634109497, \"start\": 390, \"end\": 433}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6005488634109497, \"start\": 390, \"end\": 433}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"# **References**\\n\\nAbbafati, Cristiana, Kaja M. Abbas, Mohsen Abbasi-Kangevari, Foad Abd-Allah, Ahmed Abdelalim, Mohammad Abdollahi, Ibrahim Abdollahpour, et al. 2020. “Global Burden of 87 Risk Factors in 204 Countries and Territories, 1990–2019: A Systematic Analysis for the Global Burden of Disease Study 2019.” _The Lancet_ 396 (10258): 1223–49. https://doi.org/10.1016/S0140-6736(20)30752-2.\\n\\nAbu-Ouf, Noran M., and Mohammed M. Jan. 2015. “The Impact of Maternal Iron Deficiency and Iron Deficiency Anemia on Child’s Health.” _Saudi Medical Journal_ 36 (2): 146. https://doi.org/10.15537/SMJ.2015.2.10289.\\n\\nAndreas, Nicholas J., Beate Kampmann, and Kirsty Mehring Le-Doare. 2015. “Human Breast Milk: A Review on Its Composition and Bioactivity.” _Early Human Development_ 91 (11): 629–35. https://doi.org/10.1016/J.EARLHUMDEV.2015.08.013.\\n\\nAngulo, Erick, Dalia Stern, Analí Castellanos-Gutiérrez, Adriana Monge, Martín Lajous, Sabri Bromage, Teresa T. Fung, et al. 2021. “Changes in the Global Diet Quality Score, Weight, and Waist Circumference in Mexican Women.” _The Journal of Nutrition_ 151 (12 Suppl 2): 152S-161S. https://doi.org/10.1093/JN/NXAB171.\\n\\nAquino, Estela M. L., Sandhi M. Barreto, Isabela M. Benseñor, Marilia S. Carvalho, Dóra Chor, Bruce B. Duncan, et al. 2012. “Brazilian Longitudinal Study of Adult Health (ELSA-Brasil): Objectives and Design.” _American Journal of Epidemiology_ 175 (4): 315–24. https://doi.org/10.1093/aje/kwr294.\\n\\nBadillo-Suárez, Pilar Amellali, Maricela Rodríguez-Cruz, and Xóchitl Nieves-Morales. 2017. “Impact of Metabolic Hormones Secreted in Human Breast Milk on Nutritional Programming in Childhood Obesity.” _Journal of Mammary Gland Biology and Neoplasia_ 22 (3): 171–91. https://doi.org/10.1007/S10911-0179382-Y/METRICS.\\n\\nBardach, Ariel Esteban, Natalia Espínola, Federico Rodríguez Cairoli, Lucas Perelli, Darío Balan, Alfredo Palacios, Federico Augustovski, Andrés Pichón-Riviere, and Andrea Olga Alcaraz. 2023. “The Burden of Disease and Economic Impact of Sugar-Sweetened Beverages’ Consumption in Argentina: A Modeling Study.” _PLOS ONE_ 18 (2): e0279978. https://doi.org/10.1371/JOURNAL.PONE.0279978.\\n\\n- Barker, David J.P. 2004. “The Developmental Origins of Adult Disease.” _Journal of the American College of Nutrition_ 23 (sup6): 588S-595S. https://doi.org/10.1080/07315724.2004.10719428.\\n\\nBeal, Ty, Anna Herforth, Shelly Sundberg, Sonja Y. Hess, and Lynnette M. Neufeld. 2021. “Differences in Modelled Estimates of Global Dietary Intake.” _The Lancet_ . https://doi.org/10.1016/S01406736(21)00714-5.\\n\\nBeal, Ty, Saul S. Morris, and Alison Tumilowicz. 2019. “Global Patterns of Adolescent Fruit, Vegetable, Carbonated Soft Drink, and Fast-Food Consumption: A Meta-Analysis of Global School-Based Student Health Surveys.” _Food and Nutrition Bulletin_ 40 (4): 444–59. https://doi.org/10.1177/0379572119848287/ASSET/IMAGES/LARGE/10.1177_0379572119848287FIG6.JPEG.\\n\\nBertram, Melanie Y, Jeremy A. Lauer, Karin Stenberg and Tessa Tan Torres Edejer. 2021. Methods for the Economic Evaluation of Health Care Interventions for Priority Setting in the Health System: An Update From WHO CHOICE. _International Journal of Health Policy and Management_ , 10(11): 673–677 16\", \"datasets\": [{\"mention_name\": {\"text\": \"Global Burden of Disease Study 2019\", \"confidence\": 0.6255105137825012, \"start\": 276, \"end\": 311, \"clean_text\": \"Global Burden of Disease Study 2019\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 276, \"end\": 311, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 276, \"end\": 311, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 276, \"end\": 311}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 276, \"end\": 311, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6720085740089417, \"start\": 276, \"end\": 311}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6255105137825012, \"start\": 276, \"end\": 311}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7882522940635681, \"start\": 276, \"end\": 311}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7882522940635681, \"start\": 276, \"end\": 311}}, {\"mention_name\": {\"text\": \"Brazilian Longitudinal Study of Adult Health\", \"confidence\": 0.6472817063331604, \"start\": 1287, \"end\": 1331, \"clean_text\": \"Brazilian Longitudinal Study of Adult Health\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.3995097279548645, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1287, \"end\": 1331, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1287, \"end\": 1331}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1287, \"end\": 1331, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.49874112010002136, \"start\": 537, \"end\": 581}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6472817063331604, \"start\": 537, \"end\": 581}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7937005758285522, \"start\": 537, \"end\": 581}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7937005758285522, \"start\": 537, \"end\": 581}}, {\"mention_name\": {\"text\": \"Global School-Based Student Health Surveys\", \"confidence\": 0.6590362191200256, \"start\": 2739, \"end\": 2781, \"clean_text\": \"Global School-Based Student Health Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2739, \"end\": 2781, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2739, \"end\": 2781, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2739, \"end\": 2781}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2739, \"end\": 2781, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9769040942192078, \"start\": 337, \"end\": 379}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6590362191200256, \"start\": 337, \"end\": 379}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8060059547424316, \"start\": 337, \"end\": 379}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8060059547424316, \"start\": 337, \"end\": 379}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"Black, Robert E, Cesar G Victora, Susan P Walker, Zulfiqar A Bhutta, Parul Christian, Mercedes de Onis, Majid Ezzati, et al. 2013. “Maternal and Child Undernutrition and Overweight in Low-Income and Middle-Income Countries.” _The Lancet_ 382 (9890): 427–51. https://doi.org/10.1016/s01406736(13)60937-x.\\n\\nBrero, Mauro, Christina L. Meyer, Angela Jackson‐Morris, Garrison Spencer, Elizabeth Ludwig‐Borycz, Daphne Wu, Anabel Fiorella Espinosa De Candido, et al. 2023. “Investment Case for the Prevention and Reduction of Childhood and Adolescent Overweight and Obesity in Mexico.” _Obesity Reviews_ 24 (9). https://doi.org/10.1111/obr.13595.\\n\\nBromage, Sabri, Carolina Batis, Shilpa N. Bhupathiraju, Wafaie W. Fawzi, Teresa T. Fung, Yanping Li, Megan Deitchler, et al. 2021. “Development and Validation of a Novel Food-Based Global Diet Quality Score (GDQS).” _The Journal of Nutrition_ 151 (Suppl 2): 75S. https://doi.org/10.1093/JN/NXAB244.\\n\\n- Candari, Christine Joy, Jonathan Cylus, and Ellen Nolte. 2017. “Assessing the Economic Costs of Unhealthy Diets and Low Physical Activity.” Copenhagen, Denmark.\\n\\nhttps://www.euro.who.int/__data/assets/pdf_file/0004/342166/Unhealthy-Diets-ePDF-v1.pdf.\\n\\nCarter, R. Colin, Joseph L. Jacobson, Matthew J. Burden, Rinat Armony-Sivan, Neil C. Dodge, Mary Lu Angelilli, Betsy Lozoff, and Sandra W. Jacobson. 2010. “Iron Deficiency Anemia and Cognitive Function in Infancy.” _Pediatrics_ 126 (2): e427–34. https://doi.org/10.1542/peds.2009-2097.\\n\\nDai, Haijiang, Tariq A. Alsalhe, Nasr Chalghaf, Matteo Riccò, Nicola Luigi Bragazzi, and Jianhong Wu. 2020. “The Global Burden of Disease Attributable to High Body Mass Index in 195 Countries and Territories, 1990–2017: An Analysis of the Global Burden of Disease Study.” _PLoS Medicine_ . https://doi.org/10.1371/journal.pmed.1003198.\\n\\nDaru, Jahnavi, Javier Zamora, Borja M. Fernández-Félix, Joshua Vogel, Olufemi T. Oladapo, Naho Morisaki, Özge Tunçalp, et al. 2018. “Risk of Maternal Mortality in Women with Severe Anaemia during Pregnancy and Post Partum: A Multilevel Analysis.” _The Lancet Global Health_ 6 (5): e548–54. https://doi.org/10.1016/S2214-109X(18)30078-0.\\n\\nDávila-Cervantes, Claudio Alberto. 2020. “Cardiovascular Disease in Mexico 1990-2017: Secondary Data Analysis from the Global Burden of Disease Study.” _International Journal of Public Health_ 65 (5): 661– 71. https://doi.org/10.1007/S00038-020-01377-4.\\n\\nDonkor, William E.S., Seth Adu-Afarwuah, Rita Wegmüller, Helena Bentil, Nicolai Petry, Fabian Rohner, and James P. Wirth. 2021. “Complementary Feeding Indicators in Relation to Micronutrient Status of Ghanaian Children Aged 6–23 Months: Results from a National Survey.” _Life_ 11 (9). https://doi.org/10.3390/LIFE11090969/S1.\\n\\nEaton, Jacob C., Pamela Rothpletz-Puglia, Margaret R. Dreker, Lora Iannotti, Chessa Lutter, Joyceline Kaganda, and Pura Rayco-Solon. 2019. “Effectiveness of Provision of Animal-Source Foods for Supporting Optimal Growth and Development in Children 6 to 59 Months of Age.” _Cochrane Database of Systematic Reviews_ 2019 (5).\\n\\n- <u>https://doi.org/10.1002/14651858.CD012818.PUB2/MEDIA/CDSR/CD012818/IMAGE_N/NCD012818 CMP-003-02.PNG</u>\\n\\nEconomic Commission for Latin America, and World Food Programme (ECLAC & WFP). 2017. “The Cost of the Double Burden of Malnutrition .” https://docs.wfp.org/api/documents/WFP-0000110372/download/ Accessed July 7, 2023 17\", \"datasets\": [{\"mention_name\": {\"text\": \"Global Burden of Disease Study\", \"confidence\": 0.8159011006355286, \"start\": 1721, \"end\": 1751, \"clean_text\": \"Global Burden of Disease Study\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1721, \"end\": 1751, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1721, \"end\": 1751, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1721, \"end\": 1751}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1721, \"end\": 1751, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9020683765411377, \"start\": 959, \"end\": 989}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5113828182220459, \"start\": 959, \"end\": 989}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7217743992805481, \"start\": 959, \"end\": 989}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7217743992805481, \"start\": 959, \"end\": 989}}, {\"mention_name\": {\"text\": \"National Survey\", \"confidence\": 0.37545502185821533, \"start\": 2664, \"end\": 2679, \"clean_text\": \"National Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2664, \"end\": 2679, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2664, \"end\": 2679, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2664, \"end\": 2679}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2664, \"end\": 2679, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.819353461265564, \"start\": 244, \"end\": 259}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.37545502185821533, \"start\": 244, \"end\": 259}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7884107828140259, \"start\": 244, \"end\": 259}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7884107828140259, \"start\": 244, \"end\": 259}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"FAO. 2013. “The State of Food and Agriculture 2013: Food System for Better Nutrition.” Rome.\\n\\n- ———. 2021. “Minimum Dietary Diversity for Women: An Updated Guide to Measurement - from Collection to Action.” Rome.\\n\\n- FAO, IFAD, UNICEF, WFP, and WHO. 2020. “The State of Food Security and Nutrition in the World 2020. Transforming Food Systems for Affordable Healthy Diets.” Rome. https://doi.org/10.4060/ca9692en.\\n\\nGassara, Goudja, and Jihua Chen. 2021. “Household Food Insecurity, Dietary Diversity, and Stunting in SubSaharan Africa: A Systematic Review.” _Nutrients_ 13 (12): 4401. https://doi.org/10.3390/NU13124401/S1.\\n\\nGBD 2017 Diet Collaborators. 2019. “Health Effects of Dietary Risks in 195 Countries, 1990–2017: A Systematic Analysis for the Global Burden of Disease Study 2017.” _The Lancet_ 393 (10184): 1958–72. https://doi.org/10.1016/S0140-6736(19)30041-8.\\n\\nGlobal Diet Quality Project. 2022. “Measuring What the World Eats: Insights from a New Approach.” Geneva and Cambridge, MA. https://doi.org/10.36072/dqq2022.\\n\\nGrantham-McGregor, S., and Charles C Ani. 2001. “A Review of Studies on the Effect of Iron Deficiency on Cognitive Development in Children.” _The Journal of Nutrition_ 131 (2S-2). https://doi.org/10.1093/JN/131.2.649S.\\n\\n- Guedes, Larissa Fernanda Fonseca, Mariana Santos Felisbino-Mendes, Aline Siqueira Fogal Vegi, Adriana Lúcia Meireles, Mariana Carvalho de Menezes, Deborah Carvalho Malta, and Ísis Eloah Machado. 2022. “Health Impacts Caused by Excessive Sodium Consumption in Brazil: Results of the GBD 2019 Study.” _Revista Da Sociedade Brasileira de Medicina Tropical_ 55 (suppl 1). https://doi.org/10.1590/0037-86820266-2021.\\n\\n- Hawkes, Corinna, Marie T. Ruel, Leah Salm, Bryony Sinclair, and Francesco Branca. 2020. “Double-Duty Actions: Seizing Programme and Policy Opportunities to Address Malnutrition in All Its Forms.” _The Lancet_ 395 (10218): 142–55. https://doi.org/10.1016/S0140-6736(19)32506-1.\\n\\n- He, Yuna, Yuehui Fang, Sabri Bromage, Teresa T Fung, Shilpa N Bhupathiraju, Carolina Batis, Megan Deitchler, et al. 2021. “Application of the Global Diet Quality Score in Chinese Adults to Evaluate the Double Burden of Nutrient Inadequacy and Metabolic Syndrome.” _The Journal of Nutrition_ 151 (October): 93S-100S. https://doi.org/10.1093/jn/nxab162.\\n\\nHeadey, Derek, Kalle Hirvonen, and John Hoddinott. 2018. “Animal Sourced Foods and Child Stunting.” _American Journal of Agricultural Economics_ 100 (5): 1302–19. https://doi.org/10.1093/ajae/aay053.\\n\\n- Herforth, Anna W., Doris Wiesmann, Euridice Martínez-Steele, Giovanna Andrade, and Carlos A. Monteiro. 2020. “Introducing a Suite of Low-Burden Diet Quality Indicators That Reflect Healthy Diet Patterns at Population Level.” _Current Developments in Nutrition_ 4 (12). https://doi.org/10.1093/cdn/nzaa168.\\n\\nHorton, Sue, and Richard H. Steckel. 2013. “Malnutrition: Global Economic Losses Attributable to Malnutrition 1900–2000 and Projections to 2050.” In _How Much Have Global Problems Cost the World?_ , edited by B. Lomborg, 247–72. Cambridge: Cambridge University Press. https://doi.org/10.1017/CBO9781139225793.010.\\n\\n- International Agency for Research on Cancer (IARC). 2018. _Red Meat and Processed Meat IARC Monographs on the Evaluation of Carcinogenic Risks to Humans_ . _IARC Monographs on the Evaluation of Carcinogenic Risks to Humans_ . Vol. 114.\\n\\n18\", \"datasets\": [{\"mention_name\": {\"text\": \"Global Burden of Disease Study 2017\", \"confidence\": 0.3525206744670868, \"start\": 751, \"end\": 786, \"clean_text\": \"Global Burden of Disease Study 2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 751, \"end\": 786, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 751, \"end\": 786, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 751, \"end\": 786}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 751, \"end\": 786, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.7373083829879761, \"start\": 17, \"end\": 52}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3525206744670868, \"start\": 17, \"end\": 52}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8220694661140442, \"start\": 17, \"end\": 52}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8220694661140442, \"start\": 17, \"end\": 52}}, {\"mention_name\": {\"text\": \"GBD 2019 Study\", \"confidence\": 0.7852917313575745, \"start\": 1535, \"end\": 1549, \"clean_text\": \"GBD 2019 Study\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1535, \"end\": 1549, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1535, \"end\": 1549, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1535, \"end\": 1549}, \"geography\": {\"text\": \"Brazil\", \"confidence\": 0.329660564661026, \"start\": 1512, \"end\": 1518, \"clean_text\": \"Brazil\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.5901853442192078, \"start\": 801, \"end\": 815}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7852917313575745, \"start\": 801, \"end\": 815}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7292519211769104, \"start\": 801, \"end\": 815}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7292519211769104, \"start\": 801, \"end\": 815}}, {\"mention_name\": {\"text\": \"Global Diet Quality Score in Chinese Adults\", \"confidence\": 0.3703305721282959, \"start\": 2090, \"end\": 2133, \"clean_text\": \"Global Diet Quality Score in Chinese Adults\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2090, \"end\": 2133, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2090, \"end\": 2133, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2090, \"end\": 2133}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2090, \"end\": 2133, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.806047260761261, \"start\": 505, \"end\": 548}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3703305721282959, \"start\": 505, \"end\": 548}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8512440323829651, \"start\": 505, \"end\": 548}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8512440323829651, \"start\": 505, \"end\": 548}}, {\"mention_name\": {\"text\": \"IARC Monographs\", \"confidence\": 0.326229065656662, \"start\": 3215, \"end\": 3230, \"clean_text\": \"IARC Monographs\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3215, \"end\": 3230, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3215, \"end\": 3230, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3215, \"end\": 3230}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3215, \"end\": 3230, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.38818100094795227, \"start\": 814, \"end\": 829}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.326229065656662, \"start\": 814, \"end\": 829}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8070234060287476, \"start\": 814, \"end\": 829}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8070234060287476, \"start\": 814, \"end\": 829}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"Isanaka, Sheila, Nicholas A. Menzies, Jessica Sayyad, Mudasiru Ayoola, Rebecca F. Grais, and Stéphane Doyon. 2017. Cost analysis of the treatment of severe acute malnutrition in West Africa. _Maternal and Child Nutrition_ , _13_ (4), e12398. https://doi.org/10.1111/mcn.12398 Jackson-Morris, Angela, and Christina L Meyer. 2023. “The Case for Investing in the Prevention and Treatment of Overweight and Obesity among Children and Adolescents in Three Middle-Income Countries.” _The Lancet Global Health_ 11 (March): S18. https://doi.org/10.1016/S2214-109X(23)00101-8.\\n\\nJavanbakht, Mehdi, Ahmad Reza Jamshidi, Hamid Reza Baradaran, Zahra Mohammadi, Atefeh Mashayekhi, Farhad Shokraneh, Mohsen Rezai Hamami, et al. 2018. “Estimation and Prediction of Avoidable Health Care Costs of Cardiovascular Diseases and Type 2 Diabetes Through Adequate Dairy Food Consumption: A Systematic Review and Micro Simulation Modeling Study.” _Archives of Iranian Medicine_ 21 (5): 213– 22.\\n\\nKathuria, Nishtha, Prasanta Bandyopadhyay, Shobhit Srivastava, Priyanka Rani Garg, Kshetrimayum Surmala Devi, Kauma Kurian, Suresh Kumar Rathi, and Sunil Mehra. 2023. “Association of Minimum Dietary Diversity with Anaemia among 6–59 Months’ Children from Rural India: An Evidence from a CrossSectional Study.” _Journal of Family Medicine and Primary Care_ 12 (2): 313. <u>https://doi.org/10.4103/JFMPC.JFMPC_926_22.</u> Kennedy, Gina, Anna Herforth, Chris Vogliano, Cecilia Gonzalez, Kristina Sokourenko, Betul Uyar, Rhys Manners, Tesfaye Hailu Bekele, and Ty Beal. 2022. “New Tools and Data for Global Diet Quality Monitoring.” In _22nd IUNS-ICN International Congress of Nutrition, Tokyo_ .\\n\\nKent, Seamus, Francesco Fusco, Alastair Gray, Susan A. Jebb, Benjamin J. Cairns, and Borislava Mihaylova. 2017. Body mass index and healthcare costs: a systematic literature review of individual participant data studies. _Obesity Reviews_ 18: 869-879.\\n\\n- Konstantinidi, Melina, and Antonios E. Koutelidakis. 2019. “Functional Foods and Bioactive Compounds: A Review of Its Possible Role on Weight Management and Obesity’s Metabolic Consequences.” _Medicines_ 6 (3): 94. https://doi.org/10.3390/medicines6030094.\\n\\n- Krasevec, Julia, Xiaoyi An, Richard Kumapley, France Bégin, and Edward A. Frongillo. 2017. “Diet Quality and Risk of Stunting among Infants and Young Children in Low- and Middle-Income Countries.” _Maternal & Child Nutrition_ 13 (October): e12430. https://doi.org/10.1111/MCN.12430.\\n\\n- Lane, Melissa M., Jessica A. Davis, Sally Beattie, Clara Gómez‐Donoso, Amy Loughman, Adrienne O’Neil, Felice Jacka, et al. 2021. “Ultraprocessed Food and Chronic Noncommunicable Diseases: A Systematic Review and Meta‐analysis of 43 Observational Studies.” _Obesity Reviews_ 22 (3). https://doi.org/10.1111/obr.13146.\\n\\n- Li, Man, Xiaojie Li, Yanfang Zhao, Lu Zhang, Jing Yang, Maigeng Zhou, and Zhuoqun Wang. 2021. “The Burden of Ischemic Heart Disease and Type 2 Diabetes Mellitus Attributable to Diet High in SugarSweetened Beverages in China: An Analysis for the Global Burden of Disease Study 2017.” _Journal of Diabetes_ 13 (6): 482–93. https://doi.org/10.1111/1753-0407.13132.\\n\\nLiu, Dan, Qingyang Shi, Gaiping Cheng, Qiaorong Huang, and Sheyu Li. 2022. “Worldwide Burden Attributable to Diet High in Red Meat from 1990 to 2019.” _Archives of Medical Science_ , October. https://doi.org/10.5114/aoms/156017.\\n\\n19\", \"datasets\": [{\"mention_name\": {\"text\": \"individual participant data studies\", \"confidence\": 0.3871881365776062, \"start\": 1850, \"end\": 1885, \"clean_text\": \"individual participant data studies\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1850, \"end\": 1885, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1850, \"end\": 1885, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1850, \"end\": 1885}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1850, \"end\": 1885, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.4671091139316559, \"start\": 205, \"end\": 240}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3871881365776062, \"start\": 205, \"end\": 240}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.727092444896698, \"start\": 205, \"end\": 240}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.727092444896698, \"start\": 205, \"end\": 240}}, {\"mention_name\": {\"text\": \"Global Burden of Disease Study 2017\", \"confidence\": 0.612123429775238, \"start\": 3032, \"end\": 3067, \"clean_text\": \"Global Burden of Disease Study 2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3032, \"end\": 3067, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3032, \"end\": 3067, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3032, \"end\": 3067}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3032, \"end\": 3067, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.7244924902915955, \"start\": 611, \"end\": 646}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.612123429775238, \"start\": 611, \"end\": 646}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6797305941581726, \"start\": 611, \"end\": 646}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6797305941581726, \"start\": 611, \"end\": 646}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"Luo, Hanqi, Sarah J. Zyba, and Patrick Webb. 2020. “Measuring Malnutrition in All Its Forms: An Update of the Net State of Nutrition Index to Track the Global Burden of Malnutrition at Country Level.” _Global Food Security_ 26 (September): 100453. https://doi.org/10.1016/j.gfs.2020.100453.\\n\\nMachado, Ísis Eloah, Magda Do Carmo Parajára, Larissa Fernanda Fonseca Guedes, Adriana Lúcia Meireles, Mariana Carvalho de Menezes, Mariana Santos Felisbino-Mendes, Eliseu Verly-Junior, and Deborah Carvalho Malta. 2022. “Burden of Non-Communicable Diseases Attributable to Dietary Risks in Brazil, 1990-2019: An Analysis of the Global Burden of Disease Study 2019.” _Revista Da Sociedade Brasileira de Medicina Tropical_ 55. https://doi.org/10.1590/0037-8682-0282-2021.\\n\\nMahumud, Rashidul Alam, Berhe W. Sahle, Ebenezer Owusu-Addo, Wen Chen, Rachael Lisa Morton, and Andre M.N. Renzaho. 2021. “Association of Dietary Intake, Physical Activity, and Sedentary Behaviours with Overweight and Obesity among 282,213 Adolescents in 89 Low and Middle Income to High-Income Countries.” _International Journal of Obesity 2021 45:11_ 45 (11): 2404–18. https://doi.org/10.1038/s41366-021-00908-0.\\n\\nMarcus, Hannah, Claudia Schauer, and Stanley Zlotkin. 2021. “Effect of Anemia on Work Productivity in Both Labor- and Nonlabor-Intensive Occupations: A Systematic Narrative Synthesis.” _Food and Nutrition Bulletin_ 42 (2): 289–308.\\n\\nhttps://doi.org/10.1177/03795721211006658/ASSET/IMAGES/LARGE/10.1177_03795721211006658FIG2.JPEG.\\n\\nMartin-Prevel, Yves, Mary Arimond, Pauline Allemand, Doris Wiesmann, Terri J. Ballard, Megan Deitchler, Marie Claude Dop, et al. 2017. “Development of a Dichotomous Indicator for Population-Level Assessment of Dietary Diversity in Women of Reproductive Age.” _Current Developments in Nutrition_ 1 (12): cdn.117.001701. https://doi.org/10.3945/CDN.117.001701.\\n\\nMiller, Victoria, Gitanjali M Singh, Jennifer Onopa, Julia Reedy, Peilin Shi, Jianyi Zhang, Adeem Tahira, et al. 2021. “Global Dietary Database 2017: Data Availability and Gaps on 54 Major Foods, Beverages and Nutrients among 5.6 Million Children and Adults from 1220 Surveys Worldwide.” _BMJ Global Health_ 6 (2): e003585. https://doi.org/10.1136/bmjgh-2020-003585.\\n\\nMiller, Victoria, Patrick Webb, Frederick Cudhea, Peilin Shi, Jianyi Zhang, Julia Reedy, Josh Erndt-Marino, et al. 2022. “Global Dietary Quality in 185 Countries from 1990 to 2018 Show Wide Differences by Nation, Age, Education, and Urbanicity.” _Nature Food_ 3 (9): 694–702. https://doi.org/10.1038/s43016-02200594-9.\\n\\nMiller, Victoria, Patrick Webb, Renata Micha, and Dariush Mozaffarian. 2020. “Defining Diet Quality: A Synthesis of Dietary Quality Metrics and Their Validity for the Double Burden of Malnutrition.” _The Lancet Planetary Health_ 4 (8): e352–70. https://doi.org/10.1016/S2542-5196(20)30162-5.\\n\\nMolani-Gol, Roghayeh, Sorayya Kheirouri, and Mohammad Alizadeh. 2023. “Does the High Dietary Diversity Score Predict Dietary Micronutrients Adequacy in Children under 5 Years Old? A Systematic Review.” _Journal of Health, Population, and Nutrition_ 42 (1). https://doi.org/10.1186/S41043-022-00337-3.\\n\\nMolla, Abebaw, Gudina Egata, Firehiwot Mesfin, Mikyas Arega, and Lemma Getacher. 2020. “Prevalence of Anemia and Associated Factors among Infants and Young Children Aged 6–23 Months in Debre Berhan Town, North Shewa, Ethiopia.” _Journal of Nutrition and Metabolism_ 2020. https://doi.org/10.1155/2020/2956129.\\n\\n20\", \"datasets\": [{\"mention_name\": {\"text\": \"Net State of Nutrition Index\", \"confidence\": 0.8108536005020142, \"start\": 110, \"end\": 138, \"clean_text\": \"Net State of Nutrition Index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 110, \"end\": 138, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 110, \"end\": 138, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 110, \"end\": 138}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 110, \"end\": 138, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9782419800758362, \"start\": 110, \"end\": 138}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8108536005020142, \"start\": 110, \"end\": 138}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.840355634689331, \"start\": 110, \"end\": 138}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.840355634689331, \"start\": 110, \"end\": 138}}, {\"mention_name\": {\"text\": \"Global Burden of Disease Study 2019\", \"confidence\": 0.7721362113952637, \"start\": 620, \"end\": 655, \"clean_text\": \"Global Burden of Disease Study 2019\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 620, \"end\": 655, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 620, \"end\": 655, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 620, \"end\": 655}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 620, \"end\": 655, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.4327617883682251, \"start\": 620, \"end\": 655}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7721362113952637, \"start\": 620, \"end\": 655}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8333907723426819, \"start\": 620, \"end\": 655}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8333907723426819, \"start\": 620, \"end\": 655}}, {\"mention_name\": {\"text\": \"Global Dietary Database 2017\", \"confidence\": 0.7565608024597168, \"start\": 1990, \"end\": 2018, \"clean_text\": \"Global Dietary Database 2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1990, \"end\": 2018, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1990, \"end\": 2018, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1990, \"end\": 2018}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1990, \"end\": 2018, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9977703094482422, \"start\": 249, \"end\": 277}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7565608024597168, \"start\": 249, \"end\": 277}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4295059144496918, \"start\": 249, \"end\": 277}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4295059144496918, \"start\": 249, \"end\": 277}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"Monteiro, Carlos A, Geoffrey Cannon, Renata B Levy, Jean-Claude Moubarac, Maria LC Louzada, Fernanda Rauber, Neha Khandpur, et al. 2019. “Ultra-Processed Foods: What They Are and How to Identify Them.” _Public Health Nutrition_ 22 (5): 936–41. https://doi.org/10.1017/S1368980018003762.\\n\\n- Monteiro, Carlos Augusto, Geoffrey Cannon, Mark Lawrence, Maria Laura Costa Louzada, and Pricilla Pereira Machado. 2019. _Ultra-Processed Foods, Diet Quality, and Health Using the NOVA Classification System_ . _Fao_ . Rome: FAO. https://www.fao.org/3/ca5644en/ca5644en.pdf.\\n\\nMoradi, Sajjad, Mohammad ali Hojjati Kermani, Reza Bagheri, Hamed Mohammadi, Ahmad Jayedi, Melissa M. Lane, Omid Asbaghi, Sanaz Mehrabani, and Katsuhiko Suzuki. 2021. “Ultra-Processed Food Consumption and Adult Diabetes Risk: A Systematic Review and Dose-Response Meta-Analysis.” _Nutrients_ 13 (12): 4410. https://doi.org/10.3390/nu13124410.\\n\\n- Murray, Christopher J L, Aleksandr Y Aravkin, Peng Zheng, Cristiana Abbafati, Kaja M Abbas, Mohsen Abbasi-Kangevari, Foad Abd-Allah, et al. 2020. “Global Burden of 87 Risk Factors in 204 Countries and Territories, 1990–2019: A Systematic Analysis for the Global Burden of Disease Study 2019.” _The Lancet_ 396 (10258): 1223–49. https://doi.org/10.1016/S0140-6736(20)30752-2.\\n\\n- Mwangome, Martha, and Andrew M. Prentice. 2019. “Tackling the Triple Threats of Childhood Malnutrition.” _BMC Medicine_ 17 (1): 210. https://doi.org/10.1186/s12916-019-1464-9.\\n\\n- NCD Child. 2023. NCD Global Burden of Disease Dashboard. <u>https://www.ncdchild.org/global-burdens-ofdisease-dashboard/</u>\\n\\n- Neufeld, Lynnette M, Sheryl Hendriks, and Marta Hugas. 2023. “Healthy Diet: A Definition for the United Nations Food Systems Summit 2021.” _Science and Innovations for Food Systems Transformation_ , 21–30. https://doi.org/10.1007/978-3-031-15703-5_3.\\n\\n- Ng, Marie, Tom Fleming, Margaret Robinson, Blake Thomson, Nicholas Graetz, Christopher Margono, Erin C Mullany, et al. 2014. “Global, Regional, and National Prevalence of Overweight and Obesity in Children and Adults during 1980–2013: A Systematic Analysis for the Global Burden of Disease Study 2013.” _The Lancet_ 384 (9945): 766–81. https://doi.org/10.1016/S0140-6736(14)60460-8.\\n\\n- Nguyen, Phuong Hong, Lan Mai Tran, Nga Thu Hoang, Megan Deitchler, Mourad Moursi, and Gilles Bergeron. 2023. “The Global Diet Quality Score Is Associated with Nutrient Adequacy and Depression among Vietnamese Youths.” _Annals of the New York Academy of Sciences_ , August. https://doi.org/10.1111/nyas.15053.\\n\\n- Nilson, Eduardo Augusto Fernandes, Adriana Blanco Metlzer, Marie Eve Labonté, and Patrícia Constante Jaime. 2020. “Modelling the Effect of Compliance with WHO Salt Recommendations on Cardiovascular Disease Mortality and Costs in Brazil.” _PloS One_ 15 (7): e0235514. https://doi.org/10.1371/journal.pone.0235514.\\n\\n- Nilson, Eduardo Augusto Fernandes, Everton N. da Silva, and Patrícia C. Jaime. 2020. “Developing and Applying a Costing Tool for Hypertension and Related Cardiovascular Disease: Attributable Costs to Salt/Sodium Consumption.” _The Journal of Clinical Hypertension_ 22 (4): 642. https://doi.org/10.1111/JCH.13836.\\n\\n- O’Hearn, Meghan, Laura Lara-Castor, Frederick Cudhea, Victoria Miller, Julia Reedy, Peilin Shi, Jianyi Zhang, et al. 2023. “Incident Type 2 Diabetes Attributable to Suboptimal Diet in 184 Countries.” _Nature Medicine_ 29 (4): 982–95. https://doi.org/10.1038/s41591-023-02278-8.\\n\\n21\", \"datasets\": [{\"mention_name\": {\"text\": \"Global Burden of Disease Study 2019\", \"confidence\": 0.47372475266456604, \"start\": 1166, \"end\": 1201, \"clean_text\": \"Global Burden of Disease Study 2019\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1166, \"end\": 1201, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1166, \"end\": 1201, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1166, \"end\": 1201}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1166, \"end\": 1201, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6720088124275208, \"start\": 351, \"end\": 386}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.47372475266456604, \"start\": 351, \"end\": 386}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7882534265518188, \"start\": 351, \"end\": 386}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7882534265518188, \"start\": 351, \"end\": 386}}, {\"mention_name\": {\"text\": \"NCD Global Burden of Disease Dashboard\", \"confidence\": 0.6209571957588196, \"start\": 1485, \"end\": 1523, \"clean_text\": \"NCD Global Burden of Disease Dashboard\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1485, \"end\": 1523, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1485, \"end\": 1523, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1485, \"end\": 1523}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1485, \"end\": 1523, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.7948309779167175, \"start\": 670, \"end\": 708}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6209571957588196, \"start\": 670, \"end\": 708}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6667948961257935, \"start\": 670, \"end\": 708}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6667948961257935, \"start\": 670, \"end\": 708}}, {\"mention_name\": {\"text\": \"Global Burden of Disease Study 2013\", \"confidence\": 0.7119287252426147, \"start\": 2115, \"end\": 2150, \"clean_text\": \"Global Burden of Disease Study 2013\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2115, \"end\": 2150, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2115, \"end\": 2150, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2115, \"end\": 2150}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2115, \"end\": 2150, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.3375507593154907, \"start\": 464, \"end\": 499}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7119287252426147, \"start\": 464, \"end\": 499}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7034019827842712, \"start\": 464, \"end\": 499}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7034019827842712, \"start\": 464, \"end\": 499}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"Okunogbe, Adeyemi, Rachel Nugent, Garrison Spencer, Jaynaide Powis, Johanna Ralston, and John Wilding. 2022. “Economic Impacts of Overweight and Obesity: Current and Future Estimates for 161 Countries.” _BMJ Global Health_ 7 (9): e009773. https://doi.org/10.1136/BMJGH-2022-009773.\\n\\nOlofin, Ibironke, Christine M. McDonald, Majid Ezzati, Seth Flaxman, Robert E. Black, Wafaie W. Fawzi, Laura E. Caulfield, and Goodarz Danaei. 2013. “Associations of Suboptimal Growth with All-Cause and Cause-Specific Mortality in Children under Five Years: A Pooled Analysis of Ten Prospective Studies.” _PLoS ONE_ 8 (5): e64636. https://doi.org/10.1371/journal.pone.0064636.\\n\\nParra, Kimberly L., Halimatou S. Alaofe, John E. Ehiri, Velia Leybas Nuño, Manolo Mazariegos, Brenda Garcia, Estefania Martinez, Anna Junkins, and Pauline Jolly. 2021. “Prevalence and Determinants of Underweight, Overweight, and Obesity: A Cross-Sectional Study of Sociodemographic, Dietary, and Lifestyle Factors Among Adolescent Girls in Jutiapa, Guatemala.” _Food and Nutrition Bulletin_ 42 (4): 502–19. https://doi.org/10.1177/03795721211019638.\\n\\n- Popkin, Barry M. 2004. “The Nutrition Transition: An Overview of World Patterns of Change.” _Nutrition Reviews_ 62 (July): S140–43. https://doi.org/10.1111/j.1753-4887.2004.tb00084.x.\\n\\n- Popkin, Barry M., Camila Corvalan, and Laurence M. Grummer-Strawn. 2020. “Dynamics of the Double Burden of Malnutrition and the Changing Nutrition Reality.” _The Lancet_ 395 (10217): 65–74. https://doi.org/10.1016/S0140-6736(19)32497-3.\\n\\nPries, Alissa M., Suzanne Filteau, and Elaine L. Ferguson. 2019. “Snack Food and Beverage Consumption and Young Child Nutrition in Low- and Middle-Income Countries: A Systematic Review.” _Maternal & Child Nutrition_ 15 (June): e12729. https://doi.org/10.1111/mcn.12729.\\n\\n- Pries, Alissa M., Andrea M. Rehman, Suzanne Filteau, Nisha Sharma, Atul Upadhyay, and Elaine L. Ferguson. 2019. “Unhealthy Snack Food and Beverage Consumption Is Associated with Lower Dietary Adequacy and Length-for-Age z-Scores among 12-23-Month-Olds in Kathmandu Valley, Nepal.” _Journal of Nutrition_ 149 (10): 1843–51. https://doi.org/10.1093/JN/NXZ140.\\n\\nQiao, Jie, Xiling Lin, Yiwen Wu, Xin Huang, Xiaowen Pan, Jingya Xu, Jun Yun Wu, Yuezhong Ren, and Peng Fei Shan. 2022. “Global Burden of Non-Communicable Diseases Attributable to Dietary Risks in 1990– 2019.” _Journal of Human Nutrition and Dietetics_ 35 (1): 202–13. https://doi.org/10.1111/jhn.12904.\\n\\nReardon, Thomas, David Tschirley, Lenis Saweda O. Liverpool-Tasie, Titus Awokuse, Jessica Fanzo, Bart Minten, Rob Vos, et al. 2021. “The Processed Food Revolution in African Food Systems and the Double Burden of Malnutrition.” _Global Food Security_ 28 (March). https://doi.org/10.1016/J.GFS.2020.100466.\\n\\n- Rocha, Carla Eduarda Faustino, Magda do Carmo Parajára, Ísis Eloah Machado, Aline Siqueira Fogal Vegi, Mariana Carvalho de Menezes, and Adriana Lúcia Meireles. 2023. “Chronic Diseases Attributable to a Diet Rich in Processed Meat in Brazil: Burden and Financial Impact on the Healthcare System.” _Frontiers in Nutrition_ 10 (March). https://doi.org/10.3389/FNUT.2023.1114766.\\n\\nRousham, Emily K, Sophie Goudet, Oonagh Markey, Paula Griffiths, Benjamin Boxer, Christopher Carroll, E. S. Petherick, and R. Pradeilles. 2022. “Unhealthy Food and Beverage Consumption in Children and Risk of Overweight and Obesity: A Systematic Review and Meta-Analysis.” _Advances in Nutrition_ 13 (5): 1669–96. https://doi.org/10.1093/ADVANCES/NMAC032.\\n\\nSalehi-Abargouei, Amin, Faezeh Akbari, Nick Bellissimo, and Leila Azadbakht. 2016. “Dietary Diversity Score and Obesity: A Systematic Review and Meta-Analysis of Observational Studies.” _European Journal of Clinical Nutrition_ . https://doi.org/10.1038/ejcn.2015.118.\\n\\n22\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"Schneider, Pia, Barry Popkin, Meera Shekar, Julia Dayton Eberwein, Charlotte Block, and Kyoko Shibata Okamura. 2020. “Health and Economic Impacts of Overweight/Obesity.” _Human Development Perspectives_ , February, 69–94. https://doi.org/10.1596/978-1-4648-1491-4_CH3.\\n\\nShapiro, Myra J, Shauna M Downs, Haley J Swartz, Megan Parker, Diana Quelhas, Katharine Kreis, Klaus Kraemer, Keith P West, and Jessica Fanzo. 2019. “A Systematic Review Investigating the Relation Between Animal-Source Food Consumption and Stunting in Children Aged 6–60 Months in Low and Middle-Income Countries.” _Advances in Nutrition_ , June. https://doi.org/10.1093/advances/nmz018.\\n\\nSiekmans, Kendra, Patrizia Fracassi, Tomoko Kato, Ti Kian Siow, Diana Carter, and Susan Horton. forthcoming. “Narrative Review on the Hidden Health and Nutrition Costs of Unhealthy Diets in LMIC (Draft).” Rome.\\n\\nStanton, Alice V, Frédéric Leroy, Christopher Elliott, Neil Mann, Patrick Wall, and Stefaan De Smet. 2022. “36-Fold Higher Estimate of Deaths Attributable to Red Meat Intake in GBD 2019: Is This Reliable?” _The Lancet_ 399 (10332): e23–26. https://doi.org/10.1016/S0140-6736(22)00311-7.\\n\\nStevens, Gretchen A., Ty Beal, Mduduzi N.N. Mbuya, Hanqi Luo, Lynnette M. Neufeld, O. Yaw Addo, Seth Adu-Afarwuah, et al. 2022. “Micronutrient Deficiencies among Preschool-Aged Children and Women of Reproductive Age Worldwide: A Pooled Analysis of Individual-Level Data from PopulationRepresentative Surveys.” _The Lancet. Global Health_ 10 (11): e1590–99. https://doi.org/10.1016/S2214109X(22)00367-9.\\n\\nThe Food and Land Use Coalition. 2019. “Growing Better: Ten Critical Transitions to Transform Food and Land Use.” https://www.foodandlandusecoalition.org/global-report/.\\n\\nUNICEF. 2022. “UNICEF Global Databases: Infant and Young Child Feeding.” New York. <u>https://data.unicef.org/topic/nutrition/infant-and-young-child-feeding/</u> Vadiveloo, Maya, L. Beth Dixon, and Niyati Parekh. 2013. “Associations between Dietary Variety and Measures of Body Adiposity: A Systematic Review of Epidemiological Studies.” _The British Journal of Nutrition_ . https://doi.org/10.1017/S0007114512006150.\\n\\nVega-Solano, Jaritza, Karol Madriz-Morales, Adriana Blanco-Metzler, and Eduardo Augusto Fernandes-Nilson. 2023. “Estimation of the Economic Benefits for the Public Health System Related to Salt Reduction in Costa Rica.” _PLOS ONE_ 18 (2): e0279732. https://doi.org/10.1371/JOURNAL.PONE.0279732.\\n\\n- Verger EO, Agnes Le Port, Augustin Borderon, Gabriel Bourbon, Mourad Moursi, Mathilde Savy, François Mariotti and Yves Martin-Prevel. 2021. Dietary Diversity Indicators and Their Associations with Dietary Adequacy and Health Outcomes: A Systematic Scoping Review. _Advances in Nutrition_ , 12(5): 16591672.\\n\\n- Verger EO, Savy M, Martin-Prével Y, Coates J, Frongillo E, Neufeld L et al. Healthy diet metrics: a suitability assessment of indicators for global and national monitoring purposes. Geneva: World Health Organization; 2023.\\n\\n- Walters, Dylan D, Linh T H Phan, and Roger Mathisen. 2019. “The Cost of Not Breastfeeding: Global Results from a New Tool.” _Health Policy and Planning_ 34 (6): 407–17. https://doi.org/10.1093/heapol/czz050.\\n\\nWang, Huan, Anna W. Herforth, Bo Xi, and Zhiyong Zou. 2022. “Validation of the Diet Quality Questionnaire in Chinese Children and Adolescents and Relationship with Pediatric Overweight and Obesity.” _Nutrients_ 14 (17): 3551. https://doi.org/10.3390/nu14173551.\\n\\n23\", \"datasets\": [{\"mention_name\": {\"text\": \"PopulationRepresentative Surveys\", \"confidence\": 0.5450934171676636, \"start\": 1434, \"end\": 1466, \"clean_text\": \"PopulationRepresentative Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1434, \"end\": 1466, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1434, \"end\": 1466, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1434, \"end\": 1466}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1434, \"end\": 1466, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8275119662284851, \"start\": 594, \"end\": 626}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5450934171676636, \"start\": 594, \"end\": 626}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.721217155456543, \"start\": 594, \"end\": 626}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.721217155456543, \"start\": 594, \"end\": 626}}, {\"mention_name\": {\"text\": \"UNICEF Global Databases\", \"confidence\": 0.7856153249740601, \"start\": 1749, \"end\": 1772, \"clean_text\": \"UNICEF Global Databases\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1749, \"end\": 1772, \"clean_text\": \"\"}, \"producer\": {\"text\": \"UNICEF\", \"confidence\": 0.3731739819049835, \"start\": 1734, \"end\": 1740, \"clean_text\": \"UNICEF\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1749, \"end\": 1772}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1749, \"end\": 1772, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9987020492553711, \"start\": 149, \"end\": 172}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7856153249740601, \"start\": 149, \"end\": 172}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4507099390029907, \"start\": 149, \"end\": 172}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4507099390029907, \"start\": 149, \"end\": 172}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"Wells, Jonathan C., Ana Lydia Sawaya, Rasmus Wibaek, Martha Mwangome, Marios S. Poullas, Chittaranjan S. Yajnik, and Alessandro Demaio. 2020. “The Double Burden of Malnutrition: Aetiological Pathways and Consequences for Health.” _The Lancet_ 395 (10217): 75–88. <u>https://doi.org/10.1016/S01406736(19)32472-9.</u>\\n\\n# WHO. 2019. “Healthy Diet.” Cairo.\\n\\nhttps://apps.who.int/iris/bitstream/handle/10665/325828/EMROPUB_2019_en_23536.pdf.\\n\\n- ———. 2020. “Healthy Diet Fact Sheet.” 2020. https://www.who.int/news-room/fact-sheets/detail/healthydiet.\\n\\n- WHO, and UNICEF. 2021. “Indicators for Assessing Infant and Young Child Feeding Practices: Definitions and Measurement Methods.” Geneva.\\n\\nhttps://apps.who.int/iris/bitstream/handle/10665/340706/9789240018389-eng.pdf?sequence=1.\\n\\n- WHO, UNICEF, and FAO. 2021. “Report of the Technical Consultation on Measuring Healthy Diets: Concepts, Methods and Metrics.” Geneva.\\n\\n- World Bank. 2020. _Obesity: Health and Economic Consequences of an Impending Global Challenge_ . _Obesity: Health and Economic Consequences of an Impending Global Challenge_ . https://doi.org/10.1596/978-14648-1491-4.\\n\\nWorld Cancer Research Fund, and American Institute for Cancer Research. 2018. “Continuous Update Project Expert Report 2018. Recommendations and Public Health and Policy Implications.” https://www.wcrf.org/wp-content/uploads/2021/01/Recommendations.pdf.\\n\\nZaharia, Sonia, Shibani Ghosh, Robin Shrestha, Swetha Manohar, Andrew L. Thorne-Lyman, Bernard Bashaasha, Nassul Kabunga, et al. 2021. “Sustained Intake of Animal-Sourced Foods Is Associated with Less Stunting in Young Children.” _Nature Food_ 2 (4): 246–54. https://doi.org/10.1038/s43016-02100259-z.\\n\\nZhang, Bei, Liyuan Pu, Tian Zhao, Li Wang, Chang Shu, Shan Xu, Jing Sun, Ruijie Zhang, and Liyuan Han. 2023. “Global Burden of Cardiovascular Disease from 1990 to 2019 Attributable to Dietary Factors.” _The Journal of Nutrition_ , March. https://doi.org/10.1016/j.tjnut.2023.03.031.\\n\\nZhuo, Ming, Ze Chen, Mao-Lin Zhong, Ye-Mao Liu, Fang Lei, Juan-Juan Qin, Tao Sun, et al. 2022. “The Global Disease Burden Attributable to a Diet Low in Fibre in 204 Countries and Territories from 1990 to 2019.” _Public Health Nutrition_ 26 (4): 1–12. https://doi.org/10.1017/S1368980022001987 24\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"<!-- Start of picture text --> OO) HEALTHY DIET<br>has greater diversity within food groups<br>% includes foods from several food groups and<br>meets required levels of all essential nutrients<br>a ENERGY SUFFICIENTNUTRIENTDIET ADEQUATE DIET<br>meets needs for short-term subsistence<br><!-- End of picture text --> <!-- Start of picture text --> EXPOSURE PROXY INDICATORS OUTCOMES OF INTEREST ECONOMIC COSTS<br>@ We. =<br>Infections S raicronutrientdeficiencies eSLowimpairment educational / n n n<br>i attainment<br>Wasti<br>we e Stunting/ Treatment cost<br>UNHEALTHYTM DIET R esnie al Infection(s),disease(s)<br>Does NOT meet principles Current productivity loss<br>of adequacy, diversity,<br>balance, moderation Overweight/<br>Obesity a9 communicableNon- Future productivity<br>ee loss<br>:<br>Obesogenic diseases<br>environment had Healtweig h ty Mortality’<br>T<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"<!-- Start of picture text --> I,<br>INF<br>ING<br>i ,c ;<br>/<br>A TM Jan Jon<br>vanA Nya fpr“So te f pr h<br>AN) NOAANY, rr e tin)<br>allli yy u eal<br>underweight i HealthyJNJweightIN ilOverweight/obeseTm.<br><!-- End of picture text --> A = underweight, responsive to infection control, etc. B = underweight, responsive to adequate energyand nutrient intake D = healthy weight, unhealthy food C= healthy weight, healthy food choices, choices with direct risks for NCDs E = overweight/obese, responsive to changes in obesogenic environment F = overweight/obese with risks for NCDs, responsive(especially tobalance, healthymoderation) food choices Diet-related; economic; costs would be estimatedfor B, D andF\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [27]}}, {\"page\": 28, \"chunk\": 28, \"input_text\": \"**Table 1: Selected manifestations and consequences of malnutrition in all its forms over the life course**\\n\\n|**Consequences**|**Manifestations**|**Lifecycle Phase**|**Manifestations**|**Consequences**|\\n|---|---|---|---|---|\\n|**Energy-insufficient, nutrient inade**|**quate diets**||**Energy-excess diets**||\\n|• Inadequate foetal growth &<br>development (intra uterine<br>growth retardation)<br>• Increased risk of maternal<br>mortality<br>• Preterm birth|• Underweight<br>• Inadequate GWG<br>• Nutrient<br>deficiencies|Pregnant<br>&<br>lactating women|• Overweight<br>• Obesity|• Gestational diabetes<br>• Epigenetic changes<br>• Increased risk of maternal mortality<br>• Pre-term birth|\\n|• Inadequate growth<br>• Inadequate brain development &<br>cognitive impairment<br>• Compromised function of the<br>immune system<br>• Increased risk for neonatal<br>mortality and morbidity|• Small for<br>gestational age<br>• Low birth weight<br>• Preterm birth|Newborns<br>&<br>Infants (U2)|• Large for gestational age|Increased risk for neonatal mortality<br>and morbidity<br>Overweight and obesity in adulthood<br>Type II diabetes in adulthood<br>Cardiovascular disease in adulthood|\\n|• Inadequate brain development &<br>cognitive impairment<br>• Compromised function of the<br>immune system|• Nutrient<br>deficiencies<br>• Underweight<br>• Stunting<br>• Wasting|Children<br>(up to 9 years)|• Overweight<br>• Obesity|Development of non-communicable<br>diseases<sup>3</sup>|\\n|• Cognitive impairment|• Underweight<br>• Nutrient<br>deficiencies|Adolescents<br>(10-19 years UN)|• Overweight<br>• Obesity|• Development of neurological<br>disorders<br>• Development of non-communicable<br>diseases|\\n|• Cognitive impairment<br>• Neurological disorders|• Underweight<br>• Nutrient<br>deficiencies|Adults (>19 to 64 years)|• Overweight<br>• Obesity|• Cardiovascular diseases, cancer,<br>type II diabetes|\\n|• Frailty<br>• Cognitive impairment<br>• Neurological disorders|• Nutrient<br>deficiencies<br>• Sudden weight<br>loss|Older adults<br>(65>)|• Overweight<br>• Obesity<br>• Sarcopenic obesity|• Cardiovascular diseases, cancer,<br>type II diabetes<br>• Risk for frailty<br>• Cognitive impairment<br>• Neurological disorders|\\n\\n> 3 NCD Child (2023).\\n\\n27\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [28]}}, {\"page\": 29, \"chunk\": 29, \"input_text\": \"**Table 2. Available global healthy diet metrics over the lifecycle**\\n\\n|**Level of diet**<br>**quality**|**Children 6-23 months**|**Adolescents (10-19 years)**|**Women of reproductive age**<br>**(15-49 years)**|**Adults (>15 years)**|\\n|---|---|---|---|---|\\n|Energy<br>(in)sufficiency<br>|Per capita calorie supply (compared t<br>|o age/sex specific requirements)<br>|||\\n|Micronutrient<br>(in)adequacy|**MDD**(minimum diet diversity,<br>previous day)<br>**EFF**(egg and/or flesh food<br>consumption, previous day)|NO VALIDATED INDICATOR<br>**MDD-W/Food Group**<br>**Diversity Score for girls/boys**<br>**15-19 years**<sup>**4**</sup><br>Prevalence of less than daily...<br>- fruit consumption<br>-vegetable consumption|**MDD-W**(minimum diet<br>diversity – women, previous<br>day)|**Food Group Diversity Score**<br>(previous day)|\\n|(Un)Healthy<br>diets|**ZVF**(zero vegetable or fruit<br>consumption, previous day)<br>**SwB**(sweet beverage<br>consumption, previous day)<br>**Unhealthy food consumption**<br>(sentinel unhealthy foods, previous<br>day)|NO VALIDATED METRIC<br>Average number of times per<br>day adolescents reported<br>consuming...<br>- carbonated soft drinks<br>- fruit<br>- vegetables<br>- fast food<br>Prevalence of at least weekly<br>fast food consumption|**Global Dietary Recommendat**<br>combined with MDD (for nutrie<br>2020<br>-<br>NCD-Protect score<br>-<br>NCD-Risk score<br>**Global Diet Quality Score (GD**<br>-<br>GDQS+<br>-<br>GDQS-<br>**Sentinel food group consumpt**<br>meat, processed meat, sweet foo<br>**-Sugar-sweetened drink consu**<br>**-Zero vegetable or fruit consu**|**ions Score (GDRS)**to be used<br>nt adequacy) per Herforth et al.<br>**QS)**<br>**ion**- whole grains, unprocessed red<br>ds<br>**mption**<br>**mption**|\\n\\n**Key:** light blue shading indicates measures particularly relevant for energy-(in)sufficiency; no shading indicates measures particularly relevant for micronutrient (in)adequacy; and orange shading indicates measures particularly relevant for (un)healthy diets\\n\\n**Note:** there are key gaps in measures for children 24-59 months and children in primary school grades (ages 5-9), as well as for older adults (age 50+ years) in LMIC. For children under the age of 6 months, the recommendation is that they should be exclusively breastfed. Data are available on prevalence of exclusive breastfeeding, some breastfeeding and no breastfeeding, but are not listed in the table.\\n\\n> 4 These metrics have not been validated specifically for this age group but they are captured as part of the survey sample. Sample sizes may not be sufficient for disaggregation of results for this age range (WHO and UNICEF 2021).\\n\\n28\", \"datasets\": [{\"mention_name\": {\"text\": \"Food Group Diversity Score\", \"confidence\": 0.5263811349868774, \"start\": 782, \"end\": 808, \"clean_text\": \"Food Group Diversity Score\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 782, \"end\": 808, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 782, \"end\": 808, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 782, \"end\": 808}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 782, \"end\": 808, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.5172513127326965, \"start\": 782, \"end\": 808}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5263811349868774, \"start\": 782, \"end\": 808}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.509128749370575, \"start\": 782, \"end\": 808}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.509128749370575, \"start\": 782, \"end\": 808}}, {\"mention_name\": {\"text\": \"Global Dietary Recommendat\", \"confidence\": 0.34539151191711426, \"start\": 1276, \"end\": 1302, \"clean_text\": \"Global Dietary Recommendat\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1276, \"end\": 1302, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1276, \"end\": 1302, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1276, \"end\": 1302}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1276, \"end\": 1302, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.5554667115211487, \"start\": 1276, \"end\": 1302}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.34539151191711426, \"start\": 1276, \"end\": 1302}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.7155420780181885, \"start\": 1276, \"end\": 1302}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7155420780181885, \"start\": 1276, \"end\": 1302}}, {\"mention_name\": {\"text\": \"NCD-Protect score\", \"confidence\": 0.45522579550743103, \"start\": 1354, \"end\": 1371, \"clean_text\": \"NCD-Protect score\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1354, \"end\": 1371, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1354, \"end\": 1371, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1354, \"end\": 1371}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1354, \"end\": 1371, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.5554667115211487, \"start\": 1354, \"end\": 1371}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.45522579550743103, \"start\": 1354, \"end\": 1371}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.7155420780181885, \"start\": 1354, \"end\": 1371}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7155420780181885, \"start\": 1354, \"end\": 1371}}, {\"mention_name\": {\"text\": \"NCD-Risk score\", \"confidence\": 0.36159318685531616, \"start\": 1380, \"end\": 1394, \"clean_text\": \"NCD-Risk score\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1380, \"end\": 1394, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1380, \"end\": 1394, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1380, \"end\": 1394}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1380, \"end\": 1394, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.5554667115211487, \"start\": 1380, \"end\": 1394}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.36159318685531616, \"start\": 1380, \"end\": 1394}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.7155420780181885, \"start\": 1380, \"end\": 1394}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7155420780181885, \"start\": 1380, \"end\": 1394}}, {\"mention_name\": {\"text\": \"Global Diet Quality Score\", \"confidence\": 0.5872371792793274, \"start\": 1400, \"end\": 1425, \"clean_text\": \"Global Diet Quality Score\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1400, \"end\": 1425, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1400, \"end\": 1425, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1400, \"end\": 1425}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1400, \"end\": 1425, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.5554667115211487, \"start\": 1400, \"end\": 1425}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5872371792793274, \"start\": 1400, \"end\": 1425}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.7155420780181885, \"start\": 1400, \"end\": 1425}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7155420780181885, \"start\": 1400, \"end\": 1425}}, {\"mention_name\": {\"text\": \"ions Score\", \"confidence\": 0.37802228331565857, \"start\": 1607, \"end\": 1617, \"clean_text\": \"ions Score\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1607, \"end\": 1617, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1607, \"end\": 1617, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1607, \"end\": 1617}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1607, \"end\": 1617, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.5554667115211487, \"start\": 1607, \"end\": 1617}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.37802228331565857, \"start\": 1607, \"end\": 1617}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.7155420780181885, \"start\": 1607, \"end\": 1617}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7155420780181885, \"start\": 1607, \"end\": 1617}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [29]}}, {\"page\": 30, \"chunk\": 30, \"input_text\": \"**Table 3: Summary of potential global healthy diet metrics for adults and their strengths and limitations for country study purposes**\\n\\n|**Metric**|**Description**|**Strengths**|**Limitations**|\\n|---|---|---|---|\\n|**Minimum Diet Diversity –**<br>**Women**|<br>• Gender-sensitive dichotomous indicator of<br>whether women 15-49 years have consumed<br>at least 5 out of 10 defined food groups the<br>previous day or night.<br>• Food group indicator|<br>• Valid indicator of better<br>micronutrient adequacy<br>• Available from Gallup World Poll<br>surveys for 55 LMIC plus<br>Palestine & DHS for 3 countries<br>(as of 20-Jun-2023)|• Does not include food groups<br>associated with greater NCD risk<br>• Not validated for men|\\n|**Food Group Diversity**<br>**Score**|• Measure of the number of food groups<br>consumed the previous day or night, of the<br>ten food groups used to construct the MDD-<br>W indicator<br>• Food group indicator|• Valid measure of greater nutrient<br>intakesfor men and women<br>• Available from Gallup World Poll<br>surveys for 55 LMIC plus<br>Palestine (as of 20-Jun-2023)|•<br>Does not include food<br>groups associated with<br>greater NCD risk|\\n|**Global Diet**<br>**Recommendation (GDR)**<br>**Score)**(Two sub-scores)|• Measure based on food consumption from<br>nine health-protective food groups and eight<br>food groups to avoid/limit the previous day<br>or night<br>• Validated against quantitative intakes<br>aligned with each of 11 WHO<br>recommendations on healthy diets<br>• DQ-Q asks yes/no questions about<br>consumption of a sentinel set of individual<br>foods<sup>5</sup>; no data on quantity of consumption|• Indicator (GDR Score >=10) is a<br>valid indicator of adhering to >=6<br>of current 11 global dietary<br>recommendations (and by proxy,<br>NCD risk) for men and women<br>• Food group questions adapted for<br>each country<br>• Available from Gallup World Poll<br>surveys for 55 LMIC plus<br>Palestine (as of 20-Jun-2023)|• Does not indicate nutrient<br>adequacy (unless paired with Food<br>Group Diversity Score or MDD-<br>W)<br>• Not validated against health<br>outcomes<br>• Using both sub-scores is more<br>informative than using combined<br>score|\\n|**Global Diet Quality Score**<br>**(GDQS)**(Two sub-scores)|• Food-based metric consisting of 25 food<br>groups classified as healthy, unhealthy, or<br>unhealthy when consumed in excessive<br>amounts<br>• Uses open recall of foods and beverages<br>consumed, requires some data on quantity<br>of consumption|• Valid quantitative measure of<br>nutrient adequacy and NCD risk<br>for men and women 15+ years<br>• Validated against health outcomes|• Limited availability of nationally<br>representative datasets (~7 LMIC,<br>as of 23-Jun-2023)<br>• Using both sub-scores is more<br>informative than using combined<br>score|\\n|**Global Dietary Database**<br>**(GDD) – 54 dietary risk**<br>**factors**|• Derived from modeling approach using<br>nationally or subnationally representative<br>surveys<br>• Model generates means for 54 dietary<br>factors for 185 countries|• Dietary intakes estimated for most<br>food groups, stratified by<br>demographic characteristics<br>(education, urban/rural residence)<br>• Validated with health outcomes|• For countries without 24-hour<br>dietary recall data, GDD estimates<br>are modeled based on existing data<br>sources<br>• Modeled data may be unclear to<br>non-experts|\\n\\n> 5 DQ-Q also allows for monitoring intake of ultra-processed foods and is tracked in a way that has some relevance for sustainable diets.\\n\\n29\", \"datasets\": [{\"mention_name\": {\"text\": \"Global Dietary Database\", \"confidence\": 0.6273102760314941, \"start\": 2825, \"end\": 2848, \"clean_text\": \"Global Dietary Database\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2825, \"end\": 2848, \"clean_text\": \"\"}, \"producer\": {\"text\": \"Gallup World Poll\", \"confidence\": 0.6426364183425903, \"start\": 526, \"end\": 543, \"clean_text\": \"Gallup World Poll\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2825, \"end\": 2848}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2825, \"end\": 2848, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.39426591992378235, \"start\": 2825, \"end\": 2848}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6273102760314941, \"start\": 2825, \"end\": 2848}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5563308000564575, \"start\": 2825, \"end\": 2848}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5563308000564575, \"start\": 2825, \"end\": 2848}}, {\"mention_name\": {\"text\": \"dietary recall data\", \"confidence\": 0.8453046679496765, \"start\": 3271, \"end\": 3290, \"clean_text\": \"dietary recall data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3271, \"end\": 3290, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3271, \"end\": 3290, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3271, \"end\": 3290}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3271, \"end\": 3290, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9822518825531006, \"start\": 71, \"end\": 90}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8453046679496765, \"start\": 71, \"end\": 90}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5303472280502319, \"start\": 71, \"end\": 90}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5303472280502319, \"start\": 71, \"end\": 90}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [30]}}, {\"page\": 31, \"chunk\": 31, \"input_text\": \"|**Metric**|**Description**|**Strengths**|**Limitations**|\\n|---|---|---|---|\\n||• Includes 14 foods, 7 beverages and 33<br>nutrients|• GDD data can be used to estimate<br>disease burdens attributable to poor<br>diet quality||\\n|**Global Burden of Disease**<br>**– 13 dietary risk factors**|• Derived from modelling approach using<br>multiple sources of dietary data (24-hr<br>recall, supply utilization, FFQ, sales data,<br>Household Budget Surveys)<br>• Dietary risk factors are those with<br>convincing or probable evidence of health<br>effect|• Validated with health outcomes<br>• GBD data can be used to estimate<br>disease burdens attributable to poor<br>diet quality|• For countries without 24-hour<br>dietary recall data, GBD estimates<br>are modeled based on existing data<br>sources (data on food availability<br>or food sales)<br>• Modeled data may be unclear to<br>non-experts|\\n\\nAdapted from WHO and UNICEF (2021) 30\", \"datasets\": [{\"mention_name\": {\"text\": \"GDD data\", \"confidence\": 0.5731877088546753, \"start\": 134, \"end\": 142, \"clean_text\": \"GDD data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.4081246852874756, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 134, \"end\": 142, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 134, \"end\": 142}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 134, \"end\": 142, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9667346477508545, \"start\": 134, \"end\": 142}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5731877088546753, \"start\": 134, \"end\": 142}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5753747820854187, \"start\": 134, \"end\": 142}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5753747820854187, \"start\": 134, \"end\": 142}}, {\"mention_name\": {\"text\": \"GBD data\", \"confidence\": 0.4325709939002991, \"start\": 582, \"end\": 590, \"clean_text\": \"GBD data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 582, \"end\": 590, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 582, \"end\": 590, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 582, \"end\": 590}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 582, \"end\": 590, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8082690834999084, \"start\": 582, \"end\": 590}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4325709939002991, \"start\": 582, \"end\": 590}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6188321113586426, \"start\": 582, \"end\": 590}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6188321113586426, \"start\": 582, \"end\": 590}}, {\"mention_name\": {\"text\": \"dietary recall data\", \"confidence\": 0.6369357109069824, \"start\": 706, \"end\": 725, \"clean_text\": \"dietary recall data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 706, \"end\": 725, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 706, \"end\": 725, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 706, \"end\": 725}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 706, \"end\": 725, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9296804666519165, \"start\": 11, \"end\": 30}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6369357109069824, \"start\": 11, \"end\": 30}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4413219094276428, \"start\": 11, \"end\": 30}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4413219094276428, \"start\": 11, \"end\": 30}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [31]}}, {\"page\": 32, \"chunk\": 32, \"input_text\": \"**Table 4. Summary of types of costs included in previous costing studies**\\n\\n|**Cost item, age-group to which item applies and summary of calculation method**|**Example**|\\n|---|---|\\n|**Undernutrition**||\\n|**Direct costs: treatment costs of acute malnutrition:**calculat_e from unit costs per child from existing_<br>_studies, multiplied by numbers of under-five children treated._|Isanaka et al., 2017|\\n|**Indirect costs: Present value of future productivity losses due to premature mortality:**_calculate_<br>_projected annual deaths in 0-59 months group from prevalence of stunting multiplied by_<br>_corresponding mortality risk. Project lifetime losses of productivity over the working life age 15-64,_<br>_proxied by current wage of those 15-64, adjusted by real growth of per capita GDP and discounted by_<br>_3%_|ECLAC and WFP, 2017; see<br>Bertram et al. 2017 for<br>discussion of discounting|\\n|**Indirect cost: Present value of future productivity losses due to cognitive losses which lead to**<br>**reduced educational attainment**:_calculate from prevalence of stunting in children 0-59 months,_<br>_multiplied by corresponding projected loss of wages associated with the associated underachievement_<br>_of education, adjusted by real growth of per capita GDPand discounted by 3%*_|ECLAC and WFP, 2017.|\\n|**Overweight and obesity**||\\n|**Direct costs: treatment costs for obesity-related illnesses:**_both medical treatment costs and patient_<br>_out-of-pocket costs e.g. travel and time for treatment_;|Okunogbe et al.,2022|\\n|**Indirect costs: costs of premature mortality****_:_**_lost years of life expectancy for those currently 20 and_<br>_older in the population_|ECLAC and WFP, 2017|\\n|**Indirect costs: costs of productivity losses due to morbidity**:_for those 20-64 due to absenteeism_<br>_(either of individual who is sick, or caregiver)**_|ECLAC and WFP, 2017|\\n\\n***** ECLAC and WFP (2017) also include additional costs of grade repetition to the education system: these costs are relatively small compared to future wage losses and may not be worth the effort to calculate\\n\\n**Note that some, e.g. Okunogbe et al. (2022) included losses due to presenteeism, i.e. lower work effort associated with overweight/obesity; others include lower employment rates or decreased full-time work by those with overweight/obesity; however it is less clear what the direction of causality is, hence whether or not this should be included.\\n\\n31\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [32]}}, {\"page\": 33, \"chunk\": 33, \"input_text\": \"**Table 5 Definition and source of variables for exploratory regression analyses.**\\n\\n|**Variable name**|**Definition**|**Source**|\\n|---|---|---|\\n|**Per capita calorie**supply<br>(kcal/capita/day)|Estimates of per capita food supplies available for human consumption during the reference period in terms<br>of quantity, caloric value, protein and fat content (2020). Calorie supplies are reported in kilocalories (1<br>calorie = 4.19 kJ. Availability does not imply actual consumption due to intervening factors such as food<br>waste, and likely exceeds values from dietary intake surveys|FAOSTAT|\\n|**MDD-W**|% of women who consumed at least five out of ten predefined food groups, 2021|Gallup World Poll (2021) /<br>Food systems dashboard;|\\n|**MDD**(6-23 months)|Percentage of children 6–23 months of age who consumed foods and beverages from**at least five** **out of**<br>**eight**defined food groups during the previous day. Various years, 2014-2019 (1 observation from 2011, 1<br>from 2008 and 1 from 2006)|UNICEF|\\n|**NCD-Protect**score among adults ≥<br>15|NCD-Protect scores are on a scale of 0 to 9; higher scores indicate greater protection against<br>noncommunicable diseases. Data are for individuals ≥15 years of age from the 2021 Gallup World Poll for<br>2022|World Gallup Poll<br>(GDQP) / Food systems<br>dashboard|\\n|**NCD-Risk**score among adults ≥ 15|NCD-Risk scores are on a scale of 0 to 9; higher scores indicate greater risk for noncommunicable diseases.<br>Data are for individuals ≥15 years of age from the 2021 Gallup World Poll for 2022|World Gallup Poll<br>(GDQP) / Food systems<br>dashboard|\\n|Prevalence of**overweight & obesity**<br>among women (20 - 49 years) %|Percentage of women 20–49 years of age with a BMI greater than or equal to 25 kg/m2 in 2016|UNICEF Global Database<br>(based on NCD-RisC)|\\n|Prevalence of**underweight**among<br>women (20 - 49 years) %|Percentage of women 20–49 years of age with a BMI less than 18.5 kg/m2 in 2016|UNICEF Global Database<br>(based on NCD-RisC)|\\n|Adult**diabetes**prevalence|Proportion of adults aged 18+ years with diabetes. Diabetes is defined as having a fasting glucose of 7.0<br>mmol/L or higher, being on medication for raised blood glucose, or having a past diagnosis of diabetes. Data<br>for 2014, except 1 country 2003, 1 country 2005, 2 countries 2010, 1 country 2012.|NCD-RisC / Food systems<br>dashboard|\\n|Adult**raised blood pressure**<br>prevalence|Percent of adult population (18+ years) with systolic blood pressure ≥ 140 mmHg or diastolic blood pressure<br>≥ 90 mmHg in 2015|NCD-RisC / Food systems<br>dashboard|\\n|**Gini**coefficient|Index of inequality, 0=perfect equality, 100=perfect inequality, latest year available (ranges by country from<br>2010 to 2021)|World Bank|\\n|**Region**= AFRO (omitted dummy)|Benin, Burkina Faso, Cameroon, Chad, Gabon, Ghana, Kenya, Malawi, Mali, Mozambique, Niger, Nigeria,<br>Senegal, Sierra Leone, Tunisia, Uganda, United Republic of Tanzania, Zimbabwe|World Health<br>Organization|\\n|**Region** =AMRO|Bolivia, Chile, Colombia, Ecuador, Honduras, Mexico, Nicaragua, USA,||\\n\\n32\", \"datasets\": [{\"mention_name\": {\"text\": \"Gallup World Poll\", \"confidence\": 0.5691540837287903, \"start\": 1242, \"end\": 1259, \"clean_text\": \"Gallup World Poll\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1242, \"end\": 1259, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1242, \"end\": 1259, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1242, \"end\": 1259}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1242, \"end\": 1259, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9646009802818298, \"start\": 1242, \"end\": 1259}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5691540837287903, \"start\": 1242, \"end\": 1259}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5436420440673828, \"start\": 1242, \"end\": 1259}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5436420440673828, \"start\": 1242, \"end\": 1259}}, {\"mention_name\": {\"text\": \"World Gallup Poll\", \"confidence\": 0.4299736022949219, \"start\": 1560, \"end\": 1577, \"clean_text\": \"World Gallup Poll\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1560, \"end\": 1577, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1560, \"end\": 1577, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1560, \"end\": 1577}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1560, \"end\": 1577, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9566460251808167, \"start\": 1560, \"end\": 1577}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4299736022949219, \"start\": 1560, \"end\": 1577}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5204474925994873, \"start\": 1560, \"end\": 1577}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5204474925994873, \"start\": 1560, \"end\": 1577}}, {\"mention_name\": {\"text\": \"UNICEF Global Database\", \"confidence\": 0.41907304525375366, \"start\": 1781, \"end\": 1803, \"clean_text\": \"UNICEF Global Database\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1781, \"end\": 1803, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1781, \"end\": 1803, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1781, \"end\": 1803}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1781, \"end\": 1803, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.395778626203537, \"start\": 1781, \"end\": 1803}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.41907304525375366, \"start\": 1781, \"end\": 1803}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7700690031051636, \"start\": 1781, \"end\": 1803}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7700690031051636, \"start\": 1781, \"end\": 1803}}, {\"mention_name\": {\"text\": \"NCD-RisC\", \"confidence\": 0.6048480272293091, \"start\": 2005, \"end\": 2013, \"clean_text\": \"NCD-RisC\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2005, \"end\": 2013, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2005, \"end\": 2013, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2005, \"end\": 2013}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2005, \"end\": 2013, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.5987444519996643, \"start\": 2005, \"end\": 2013}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6048480272293091, \"start\": 2005, \"end\": 2013}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.563143789768219, \"start\": 2005, \"end\": 2013}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.563143789768219, \"start\": 2005, \"end\": 2013}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [33]}}, {\"page\": 34, \"chunk\": 34, \"input_text\": \"|**Region** =EMRO|Afghanistan, Bahrain, Egypt, Jordan, Lebanon, Morocco, Pakistan, Yemen|\\n|---|---|\\n|**Region**= EURO|Albania, Armenia, Azerbaijan, Greece, Israel, Kazakhstan, Kyrgyzstan, Russian Federation, Tajikistan,<br>Türkiye, Uzbekistan,|\\n|**Region** =SEARO|Bangladesh, India, Nepal, Sri Lanka,|\\n|**Region**=WPRO|Cambodia, China, Indonesia, Lao (People’s Democratic Republic of), Philippines, Vietnam|\\n\\n33\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [34]}}, {\"page\": 35, \"chunk\": 35, \"input_text\": \"**Table 6. Summary of cross-country regression results, children 6-23 months**\\n\\n|**Independent variables**|**Dependent vari**<br>|**able**<br>||\\n|---|---|---|---|\\n||**Child**<br>**stunting**|**Child**<br>**wasting**|**Child**<br>**overweight**|\\n|MDD|-0.137<br>(0.096)|<br> <br>0.053<br>(0.041)|<br> <br>0.053<br>(0.041)|\\n|Gini index|0.184<br>(0.253)|0.154<br>(0.108)|0.154<br>(0.108)|\\n|Daily calorie supply<br>(Kcal/capita/day)|-0.013***<br>(0.004)|0.003<br>(0.002)|0.003<br>(0.002)|\\n|Region Americas (AMRO)|-3.853<br>(6.744)|-0.413 (2.882)|-0.413<br>(2.882)|\\n|Region Eastern Mediterranean<br>(EMRO)|8.593<br>(6.240)|8.245***<br>(2.667)|8.245***<br>(2.667)|\\n|Region Europe (EURO)|-6.553<br>(6.648)|-0.548<br>(2.841)|-0.548<br>(2.841)|\\n|Region Western Pacific (WPRO)|7.663<br>(4.545)|2.254<br>(1.942)|2.254<br>(1.942)|\\n|Region Southeast Asia<br>(SEARO)|9.795**<br>(4.523)|-2.137<br>(1.933)|-2.137<br>(1.933)|\\n|Constant|51.958***<br>(17.973)|-10.049<br>(7.681)|-10.049<br>(7.681)|\\n|Observations|40|40|40|\\n|R2|0.511|0.360|0.469|\\n|Adjusted R2|0.385|0.195|0.332|\\n|Residual Std. Error|8.635 (df = 31)|3.726 (df =<br>31)|3.690 (df = 31)|\\n|F Statistic|4.047*** (df =<br>8; 31)|2.179* (df = 8;<br>31)|3.424*** (df =<br>8; 31)|\\n|AIC Score|295.7882|227.777|227.777|\\n\\nNote: *p<0.1; **p<0.05; ***p<0.01; Omitted region dummy is Africa 34\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [35]}}, {\"page\": 36, \"chunk\": 36, \"input_text\": \"**Table 7. Summary of cross-country regression results, non-pregnant women 20-49 years**\\n\\n||**Dependent variable**<br>’|<br>’|\\n|---|---|---|\\n|Independent variable|Womens low BMI#|Womens overweight/obesity|\\n|Calories (Kcal/capita/day)|-0.000447<br>(0.000164)**|-0.0117138<br>(0.0298)|\\n|Calories squared|-|3.87e-6<br>(5.09e-6)|\\n|MDD-W|-0.00822<br>(0.00542)|-|\\n|GINI Index|-0.0222<br>(0.0109)**|-|\\n|NCDProtect|-|1.332622<br>(2.550)|\\n|NCDRisk|-|4.662309<br>(2.258)**|\\n|Region Americas|-0.948<br>(0.232)***|13.22596<br>(4.657)***|\\n|Region Eastern Mediterranean|-0.743<br>(0.223)***|15.75202<br>(3.799)***|\\n|Region European|-0.667<br>(0.259)**|1.158839<br>(4.771)|\\n|Region Western Pacific|0.587<br>(0.281)**|-18.46418<br>(4.849)***|\\n|Region Southeast Asia|0.664<br>(0.235)***|-13.44361<br>(4.467)***|\\n|Constant|4.679<br>(0.639)***|24.38325<br>(43.935)|\\n|Observations|53|55|\\n|R2|0.841|0.7670|\\n|Adjusted R2|0.7448|0.7204|\\n|F Statistic|19.97*** (df=8,44)|16.46*** (df=8,46)|\\n|AIC Score|63.77|395.7938|\\n\\nNote: # indicates that ln(underweight) was used, based on better fit; *p<0.1 **p<0.05; ***p<0.01; omitted region is Africa. .\\n\\n35\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [36]}}, {\"page\": 37, \"chunk\": 37, \"input_text\": \"**Table 8: Summary of cross-country regression results, adults**\\n\\n||**Dependent variable**||\\n|---|---|---|\\n|**Independent variable**|**Hypertension**|**Diabetes#**|\\n|Calories (Kcal/capita/day|0.0167183<br>(0.0107)|-0.0031879<br>(0.00948)|\\n|Calories squared|-3.41e-6<br>(1.83e-6)*|8.74e-7<br>(1.70e-6)|\\n|Overweight|Omitted (model fit worse)|Omitted (model fit<br>worse)|\\n|NCDProtect|-0.632<br>(0.884)|-0.3880131<br>(0.687)|\\n|NCDRisk|Omitted (model fit worse)|Omitted (model fit<br>worse)|\\n|Region Americas|-7.66973<br>(1.558)***|2.28057<br>(1.213)*|\\n|Region Eastern Mediterranean|-1.99171<br>(1.398)|7.087293<br>(1.024)***|\\n|Region European|-1.264047<br>(1.483)|1.840187<br>(1.204)|\\n|Region Western Pacific|-4.08404<br>(1.72)**|0.2327971<br>(1.271)|\\n|Region South East Asian|<br>-2.668754<br>(1.646)|2.412013<br>(1.845)**|\\n|Constant|10.14608<br>(15.828)|10.21116<br>(13.82)|\\n|Observations|55|41|\\n|R2|0.6048|0.6753|\\n|Adjusted R2|0.5361|0.5941|\\n|F Statistic|8.80*** (df=7,47)|8.32*** (df=7,33)|\\n|AIC Score|285.2234|176.2193|\\n\\n#Used square root of diabetes to normalize distribution\\n\\nNote: *p<0.1; **p<0.05; ***p<0.01; omitted region is Africa 36\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [37]}}, {\"page\": 38, \"chunk\": 38, \"input_text\": \"# **Appendix. Data sources – technical annex**\\n\\n# **<u>Measures of diet characteristics, by age group</u>**\\n\\nThis appendix provides additional detail and sources for the measures required to model the economic impact of unhealthy diets. Tables 1, 2 and 3 provide further detail to Table 2 of the main text, organized according to the same age groups (6-23 months; adolescents; women of reproductive age; adults). For each age group, there are dietary recommendations as to foods to include, and foods to avoid. Tables 4 and 5 provide details on data sources, and Table 6 summarizes the current WHO dietary guidelines.\\n\\n# **1. Exclusive breastfeeding under six months (EBF) infants < 6 months**\\n\\nThe WHO Global Strategy for Infant and Young Child Feeding recommends that infants be exclusively breastfed until they turn six months of age. Exclusive breastfeeding is the safest and healthiest option for children everywhere, guaranteeing infants a food source that is uniquely adapted to their needs while also being safe, clean, healthy and accessible.\\n\\n# **2. Infants and young children 6-23 months**\\n\\nTo measure diets of young children, a set of 17 indicators for assessing infant and young child feeding practices were updated and published in 2021 by WHO and UNICEF.<sup>6</sup> The recommended indicators are population-level indicators designed for data collection in large-scale surveys to support program assessment, planning and monitoring. UNICEF currently maintains a global database for these indicators, with the most recent update in October 2022.<sup>7</sup>\\n\\n# **3. Children 3-5 years, school-age children and adolescents**\\n\\nThere are no global guidelines on older children and adolescent diets. WHO (2018)<sup>8</sup> does have a guideline on promoting physical activity in this age group.\\n\\n# **4. Women of reproductive age (15-49 years)**\\n\\nFor women of reproductive age, various dietary measures have been developed to assess the level of nutrient adequacy (or risk of micronutrient inadequacy) as well as consumption of foods that are associated with lower or higher risk of overweight, obesity and related NCDs.\\n\\n**Minimum dietary diversity for women** is an indicator designed and validated as a proxy for one dimension of diet quality (micronutrient adequacy) for non-pregnant women 15-49 years of age. MDD-W is calculated based on reported intake of 10 food groups and is defined as consumption of at least five out of these 10 food groups.\\n\\nThere is currently no harmonized database for MDD-W globally, due in part to varying methods of data collection (open and sentinel list-based recalls, national and subnational surveys). The **Food Systems Dashboard** currently hosts MDD-W meta-data for 55 countries plus Palestine (54 LMIC), based on Gallup World Poll data that are not open-access and were collected using a telephone-based interview. In the future, additional MDD-W data are expected to be available from at least 21 nationally representative DHS surveys.\\n\\n> 6 Indicators for assessing infant and young child feeding practices: definitions and measurement methods ( World Health Organization and UNICEF 2021).\\n\\n> 7 UNICEF (2022). https://data.unicef.org/topic/nutrition/infant-and-young-child-feeding/\\n\\n> 8 WHO (2018). Guideline: implementing effective actions for improving adolescent nutrition. Geneva: World Health Organization. https://apps.who.int/iris/handle/10665/260297\\n\\n37\", \"datasets\": [{\"mention_name\": {\"text\": \"Gallup World Poll data\", \"confidence\": 0.8206144571304321, \"start\": 2766, \"end\": 2788, \"clean_text\": \"Gallup World Poll data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2766, \"end\": 2788, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2766, \"end\": 2788, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2766, \"end\": 2788}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2766, \"end\": 2788, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"microdata\", \"confidence\": 0.6768903136253357, \"start\": 494, \"end\": 516}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8206144571304321, \"start\": 494, \"end\": 516}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5767202973365784, \"start\": 494, \"end\": 516}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5767202973365784, \"start\": 494, \"end\": 516}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [38]}}, {\"page\": 39, \"chunk\": 39, \"input_text\": \"# **Appendix Table 1: Dietary metrics for infants and young children – definitions and data source characteristics**\\n\\n|**Dietary intake variables**|**Indicator definition**|**Data source**|**Data**<br>|**Notes**|\\n|---|---|---|---|---|\\n|**_Infants< 6 months_**<br>|||**availability**<br>||\\n|**Exclusive breastfeeding**<br>under six months (EBF)|percentage of infants 0–5 months<br>of age who were fed exclusively<br>with breast milk during the<br>previous day|UNICEF<br>Global<br>Database|N=136<br>countries;<br>most recent<br>survey year|EBF is defined as breastfeeding with no other food or<br>drink, not even water.|\\n|**_Infants and young children_**<br>|<br>**_6-23 months_**<br>||<br>||\\n|**Minimum Diet Diversity**<br>6-23 months (MDD)|percentage of children 6–23<br>months of age who consumed<br>foods and beverages fromat least<br>five out of eightdefined food<br>groups during the previous day|UNICEF<br>Global<br>Database|N=107<br>countries;<br>most recent<br>survey year|The eight food groups are: 1. breast milk; 2. grains,<br>roots, tubers and plantains; 3. pulses (beans, peas,<br>lentils), nuts and seeds; 4. dairy products (milk, infant<br>formula, yogurt, cheese); 5. flesh foods (meat, fish,<br>poultry, organ meats); 6. eggs; 7. vitamin-A rich fruits<br>and vegetables; and 8. other fruits and vegetables.<br>There is no minimum quantity of food or beverage from<br>a food group required to count.|\\n|**Egg and/or flesh food**<br>**consumption 6-23 months**<br>**(EFF)**|Percentage of children 6–23<br>months of age who consumed egg<br>and/or flesh food during the<br>previous day|UNICEF<br>Global<br>Database|N=106<br>countries,<br>most recent<br>survey year|Based on consumption of food groups 5 (flesh foods:<br>meat, fish, poultry, organ meats) and 6 (eggs). Children<br>are counted if either food group has been consumed.|\\n|**Zero vegetable or fruit**<br>**consumption 6-23 months**<br>**(ZVF)**|Percentage of children 6–23<br>months of age who did not<br>consume any vegetables or fruits<br>during the previous day|UNICEF<br>Global<br>Database|N=106<br>countries,<br>most recent<br>survey year|Based on consumption of food groups 7 (vitamin A-rich<br>fruits and vegetables) and 8 (other fruits and vegetables).<br>Plantains, starchy roots and tubers in food group 2 (such as<br>white potatoes, yams and cassava) do not count.|\\n|**Sweet beverage**<br>**consumption 6-23 months**<br>**(SwB)**|Percentage of children 6–23<br>months of age who consumed<br>a sweet beverage during the<br>previous day|https://data.uni<br>cef.org/wp-<br>content/uploads<br>/2021/09/UNIC<br>EF_Expanded_<br>Global_Databa<br>ses_Unhealthy_<br>practices_2022.<br>xlsx|Recently<br>added to<br>DHS and<br>MICS<br>surveys;<br>very few<br>country<br>data as yet|Sweet beverages include commercially produced and<br>packaged, sweetened beverages<br>such as soda pop, fruit-flavoured drinks, sports drinks,<br>chocolate and other flavoured milk drinks, malt drinks, etc;<br>100% fruit juice as well as fruit-flavoured drinks; home-<br>made drinks of any kind to which sweeteners have been<br>added.|\\n|**Unhealthy food**<br>**consumption 6-23 months**<br>**(UFC)**|Percentage of children 6–23<br>months of age who consumed<br>selected sentinel unhealthy foods<br>during the previous day|https://data.uni<br>cef.org/wp-<br>content/uploads<br>/2021/09/UNIC|Recently<br>added to<br>DHS and<br>MICS: only|“sentinel unhealthy foods” are foods that are likely to be<br>consumed by IYC and are high in sugar, salt and/or<br>unhealthy fats. These include four categories of unhealthy<br>foods: 1) candies, chocolates and other sugar confections;|\\n\\n38\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [39]}}, {\"page\": 40, \"chunk\": 40, \"input_text\": \"|**Dietary intake variables**|**Indicator definition**|**Data source**|**Data**<br>**availability**|**Notes**|\\n|---|---|---|---|---|\\n|||EF_Expanded_<br>Global_Databa<br>ses_Unhealthy_<br>practices_2022.<br>xlsx|few country<br>data as yet|2) frozen treats like ice cream, popsicles, etc.; 3) cakes,<br>pastries, sweet biscuits, etc.; 4) chips, crisps, cheese puffs,<br>French fries, instant noodles, etc.|\\n\\n39\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [40]}}, {\"page\": 41, \"chunk\": 41, \"input_text\": \"**Appendix Table 2: Dietary metrics for school-age and adolescent children – definitions and data source characteristics**\\n\\n|**Dietary intake variables**|**Definition**|**Data source**|**Year**|**Data**<br>**availability**|**Notes**|\\n|---|---|---|---|---|---|\\n|**Frequency of carbonated**<br>**soft drink consumption**|Average number of times per day<br>adolescents (12-17 years) reported<br>consuming carbonated soft drinks|Data collected from<br>Global School-based<br>Student Health|year of<br>survey<br>varies by|N=63<br>countries<br>with data|Assesses behavioural risk<br>factors and protective factors in<br>10 key areas using a self-|\\n|**Frequency of fruit**<br>**consumption**|Average number of times per day<br>adolescents (12-17 years) reported<br>consuming fruit|Surveys<br>https://www.foodsyst|country<br>(range 2009<br>to 2015)||administered questionnaire<br>administered in schools|\\n|**Frequency of vegetable**<br>**consumption**|<br>Average number of times per day<br>adolescents (12-17 years) reported<br>consuming vegetables|emsdashboard.org/|||‘_during the past 30 days, how_<br>_many times per day do you_<br>_usually drink/eat..._’|\\n|**Frequency of fast food**<br>**consumption**|Average number of times per week<br>adolescents (12-17 years) reported<br>consuming fast food||||Meta-analysis byBeal et al.<br>2019|\\n|**At least daily carbonated**<br>**soft drink consumption**|<br>Proportion of adolescents (12-17<br>years) who reported consuming<br>carbonated soft drinks at least 1<br>time per day over the last 30 days|||||\\n|**At least weekly fast food**<br>**consumption**|Proportion of adolescents (12-17<br>years) who reported consuming fast<br>food at least 1 time per week|||||\\n|**Less than daily fruit**<br>**consumption**|Proportion of adolescents (12-17<br>years) who reported consuming<br>fruit less than 1 time per day over<br>the last 30 days|||||\\n|**Less than daily vegetable**<br>**consumption**|Proportion of adolescents (12-17<br>years) who reported consuming<br>vegetables less than 1 time per day<br>overthelast 30 days|||||\\n\\n40\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [41]}}, {\"page\": 42, \"chunk\": 42, \"input_text\": \"# **Appendix Table 3: Dietary metrics for adults – definitions and data source characteristics**\\n\\n|**Dietary intake**<br>|**Indicator definition**|**Data source**|**Year**|**Data**<br><sup>**1**</sup>|**Notes**|\\n|---|---|---|---|---|---|\\n|**variables**||||**availability**||\\n|**Women of reproductive**<br>|**age**<br>|||||\\n|**Minimum Diet**<br>**Diversity for women**<br>**(MDD-W)**|Percentage of women (15-49 y) who<br>consumed at least 5 of 10 defined food<br>groups the previous day or night.|World Gallup Poll<br>(Global Diet Quality<br>Project)|2021-2022|N=55<br>countries|The food groups include: 1.<br>Grains, white roots and tubers<br>and plantains; 2. Pulses<br>(beans, peas and lentils); 3.<br>Nuts and seeds; 4. Milk and<br>milk products; 5. Meat,<br>poultry and fish; 6. Eggs; 7.<br>Dark green leafy vegetables;<br>8. Other vitamin A-rich fruits<br>and vegetables; 9. Other<br>vegetables; 10. Other fruits.|\\n|**Adults (age ≥ 15 years):**<br>|<br>||||<br>|\\n|**Food group diversity**<br>**score**|The number of food groups consumed<br>the previous day or night, of the ten food<br>groups used in the MDD-W indicator|World Gallup Poll<br>(Global Diet Quality<br>Project)|2021-2022|N=55<br>countries|The score ranges from 0 to 10<br>expressed as an average score<br>for the population with a<br>higher score indicating<br>inclusion of more food groups<br>in the diet.|\\n|**Global Dietary**<br>**Recommendations**<br>**(GDR) score**|Score based on food consumption from<br>nine health-protective food groups<br>(NCD-Protect) and eight food groups to<br>limit or avoid (NCD-Risk) during the<br>previous day or night, which are<br>associated with meeting WHO global<br>dietary recommendations.|World Gallup Poll<br>(Global Diet Quality<br>Project)|2021-2022|N=55countri<br>es|The score ranges from 0 to 18<br>expressed as an average score.|\\n|**NCD-Protect score**|Score based on consumption during the<br>previous day or night of nine food groups<br>that are associated with meeting WHO<br>recommendations on fruits, vegetables,<br>whole grains, pulses, nuts and seeds, and<br>fiber.|World Gallup Poll<br>(Global Diet Quality<br>Project)|2021-2022|N=55<br>countries|An indicator of dietary factors<br>protective against NCDs<br>The score ranges from zero to<br>nine expressed as an average<br>score.|\\n|**NCD-Risk score**|Score based on consumption during the<br>previous day or night of eight food<br>groups that are negatively associated with<br>meeting WHOrecommendations on free|World Gallup Poll<br>(Global Diet Quality<br>Project)|2021-2022|N=55<br>countries|An indicator of dietary risk<br>factors for NCDs|\\n\\n41\", \"datasets\": [{\"mention_name\": {\"text\": \"World Gallup Poll\", \"confidence\": 0.7443368434906006, \"start\": 2486, \"end\": 2503, \"clean_text\": \"World Gallup Poll\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2486, \"end\": 2503, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2486, \"end\": 2503, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"2021-2022\", \"confidence\": 0.5697476863861084, \"start\": 2540, \"end\": 2549}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2486, \"end\": 2503, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.49060720205307007, \"start\": 69, \"end\": 86}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7443368434906006, \"start\": 69, \"end\": 86}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7405011057853699, \"start\": 69, \"end\": 86}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7405011057853699, \"start\": 69, \"end\": 86}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [42]}}, {\"page\": 43, \"chunk\": 43, \"input_text\": \"|**Dietary intake**<br>**variables**|**Indicator definition**|**Data source**|**Year**|**Data**<br>**availability**<sup>**1**</sup>|**Notes**|\\n|---|---|---|---|---|---|\\n||sugar, salt, total and saturated fat, and red<br>and processed meat.||||The score ranges from zero to<br>nine expressed as an average<br>score.|\\n|**All-5**|% of adults ≥ 15 years who consumed all<br>five food groups typically recommended<br>for daily consumption in food-based<br>dietary guidelines around the world|World Gallup Poll<br>(Global Diet Quality<br>Project)|2021-2022|N=55<br>countries|Proportion of the population<br>who consumed all five food<br>groups [_as well as each one_<br>_individually_] at least one<br>**vegetable**, at least one**fruit**,<br>at least one**pulse, nut or**<br>**seed**, at least one**animal-**<br>**source food**, and at least one<br>**starchy staple**in the previous<br>day or night.|\\n|**Zero vegetable or fruit**<br>**consumption**|Proportion of the population who<br>consumed no vegetables or fruits during<br>the previous day or night.|World Gallup Poll<br>(Global Diet Quality<br>Project)|2021-2022|N=55<br>countries||\\n|**Soft drink**<br>**consumption**|Proportion of the population who<br>consumed a sugar-sweetened soft drink<br>during the previous day or night. Sugar-<br>sweetened soft drinks include soda,<br>energy drinks, and sports drinks.|World Gallup Poll<br>(Global Diet Quality<br>Project)|2021-2022|N=55<br>countries||\\n\\n1 Data are also available for Palestine 42\", \"datasets\": [{\"mention_name\": {\"text\": \"World Gallup Poll\", \"confidence\": 0.6096991300582886, \"start\": 487, \"end\": 504, \"clean_text\": \"World Gallup Poll\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 487, \"end\": 504, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 487, \"end\": 504, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 487, \"end\": 504}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 487, \"end\": 504, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9464566707611084, \"start\": 487, \"end\": 504}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.533452570438385, \"start\": 487, \"end\": 504}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7731593251228333, \"start\": 487, \"end\": 504}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7731593251228333, \"start\": 487, \"end\": 504}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [43]}}, {\"page\": 44, \"chunk\": 44, \"input_text\": \"# **Appendix Table 4. Nutrition/health metrics for children and adolescents**\\n\\n|**Outcome measure**|**Data source**|**Data**<br>**availability**|**Notes**|\\n|---|---|---|---|\\n|**Children under 5**||||\\n|**Neonatal mortality rate**and number of<br>neonatal deaths|UNICEF Global<br>Database||https://data.unicef.org/topic/child-survival/neonatal-mortality/|\\n|**U5 mortality rate**and number of deaths<br>of children U5|UNICEF Global<br>Database||https://data.unicef.org/topic/child-survival/under-five-mortality/|\\n|**Child stunting**(survey estimates,<br>updated Dec 2022)|UNICEF Global<br>Database|N=202<br>countries<br>(date depends<br>on survey<br>year)|https://data.unicef.org/wp-<br>content/uploads/2022/05/UNICEF_WHO_WB_Global_Expanded_Datab<br>ases_Stunting_Dec_2022.xlsx|\\n|**Child wasting**(survey estimates,<br>updated Dec 2022)|UNICEF Global<br>Database|N=202<br>countries<br>(date depends<br>on survey<br>year)|https://data.unicef.org/wp-<br>content/uploads/2022/05/UNICEF_WHO_WB_Global_Expanded_Datab<br>ases_Wasting_Dec_2022.xlsx|\\n|**Child overweight**(survey estimates,<br>updated Dec 2022)|UNICEF Global<br>Database|N=202<br>countries<br>(date depends<br>on survey<br>year)|https://data.unicef.org/wp-<br>content/uploads/2022/05/UNICEF_WHO_WB_Global_Expanded_Datab<br>ases_Overweight_Dec_2022.xlsx|\\n|**Child anaemia**: Percentage of children<br>aged 6−59 months with a Hb<br>concentration less than 110 g/L, adjusted<br>for altitude.|WHO Global Health<br>Observatory|n=124<br>countries,<br>2019|https://www.who.int/data/gho/data/indicators/indicator-<br>details/GHO/prevalence-of-anaemia-in-children-under-5-years-(-)|\\n|**Children and adolescents 5-19**||||\\n|**Overweight**|https://www.foodsystem<br>sdashboard.org/|2016|Proportion of 5-19 year olds with BMI-for-age > 1 SD above median of<br>WHO growth reference for this age group; Based on NCD-RisC<br>estimates|\\n|**Obesity**|As above||As above, but with BMI-for-age>2 SD above growth reference|\\n|**Underweight**|As above||As above, but with BMI-for-age<1 SD below growth reference|\\n\\n43\", \"datasets\": [{\"mention_name\": {\"text\": \"NCD-RisC\", \"confidence\": 0.48041868209838867, \"start\": 1854, \"end\": 1862, \"clean_text\": \"NCD-RisC\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1854, \"end\": 1862, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1854, \"end\": 1862, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1854, \"end\": 1862}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1854, \"end\": 1862, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.753801167011261, \"start\": 1854, \"end\": 1862}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.48041868209838867, \"start\": 1854, \"end\": 1862}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7286625504493713, \"start\": 1854, \"end\": 1862}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7286625504493713, \"start\": 1854, \"end\": 1862}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [44]}}, {\"page\": 45, \"chunk\": 45, \"input_text\": \"# **Appendix Table 5. Nutrition/health metrics for women of reproductive age**\\n\\n|**Nutrition/health outcome**<br>**variables**|Data source|Data availability|Notes|\\n|---|---|---|---|\\n|**Anaemia**pregnant women (15-<br>49 years) and non-pregnant<br>women (15-49 years)|UNICEF Global Database<br>(based on WHO GHO)|All countries,<br>Latest year: 2019||\\n|**Underweight**in WRA (20-49<br>years)|UNICEF Global Database<br>(based on NCD-RisC)|All countries, Latest<br>year: 2016|Percentage of women 20-49 years of age with a<br>BMI <18.5 kg/m2<br>Data also available for men from NCD-RisC.|\\n|**Overweight and obesity**in<br>WRA|UNICEF Global Database<br>(based on NCD-RisC)|All countries, Latest<br>year: 2016|Percentage of women 20–49 years of age with a<br>BMI ≥ 25 kg/m<sup>2</sup><br>Data also available for men from NCD-RisC.|\\n|**Low-birth weight**prevalence<br>and prevalence of newborns<br>without birthweight data|UNICEF Global Database|N=148 (54 no LBW<br>estimate)<br>Latest year:**2015**<br>(will be updated later<br>this year)||\\n|**Stillbirth**rate and number of<br>stillbirths|UNICEF Global Database|Latest year: 2021||\\n|**Preterm birth**|Modeled estimates by<br>Chawanpaiboon et al. 2019<sup>5</sup>|Latest year: 2014||\\n\\n44\", \"datasets\": [{\"mention_name\": {\"text\": \"WHO GHO\", \"confidence\": 0.4571962356567383, \"start\": 303, \"end\": 310, \"clean_text\": \"WHO GHO\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 303, \"end\": 310, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 303, \"end\": 310, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 303, \"end\": 310}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 303, \"end\": 310, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9654960036277771, \"start\": 303, \"end\": 310}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4571962356567383, \"start\": 303, \"end\": 310}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.645258903503418, \"start\": 303, \"end\": 310}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.645258903503418, \"start\": 303, \"end\": 310}}, {\"mention_name\": {\"text\": \"NCD-RisC\", \"confidence\": 0.6970696449279785, \"start\": 573, \"end\": 581, \"clean_text\": \"NCD-RisC\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 573, \"end\": 581, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 573, \"end\": 581, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 573, \"end\": 581}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 573, \"end\": 581, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.3866012990474701, \"start\": 573, \"end\": 581}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6970696449279785, \"start\": 573, \"end\": 581}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.6305593252182007, \"start\": 573, \"end\": 581}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6305593252182007, \"start\": 573, \"end\": 581}}, {\"mention_name\": {\"text\": \"UNICEF Global Database\", \"confidence\": 0.6481522917747498, \"start\": 1083, \"end\": 1105, \"clean_text\": \"UNICEF Global Database\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1083, \"end\": 1105, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1083, \"end\": 1105, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1083, \"end\": 1105}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1083, \"end\": 1105, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.971458375453949, \"start\": 18, \"end\": 40}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6481522917747498, \"start\": 18, \"end\": 40}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6273669600486755, \"start\": 18, \"end\": 40}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6273669600486755, \"start\": 18, \"end\": 40}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [45]}}, {\"page\": 46, \"chunk\": 46, \"input_text\": \"# **Appendix Table 6. World Health Organization guidelines for a healthy diet; World Cancer Research Fund recommendation regarding meat**\\n\\n## **For adults**\\n\\nA healthy diet includes the following:\\n\\n- Fruit, vegetables, legumes (e.g. lentils and beans), nuts and whole grains (e.g. unprocessed maize, millet, oats, wheat and brown rice).\\n\\n- At least 400 g (i.e. five portions) of fruit and vegetables per day _(2)_ , excluding potatoes, sweet potatoes, cassava and other starchy roots.\\n\\n- Less than 10% of total energy intake from free sugars _(2, 7)_ , which is equivalent to 50 g (or about 12 level teaspoons) for a person of healthy body weight consuming about 2000 calories per day, but ideally is less than 5% of total energy intake for additional health benefits _(7)_ . Free sugars are all sugars added to foods or drinks by the manufacturer, cook or consumer, as well as sugars naturally present in honey, syrups, fruit juices and fruit juice concentrates.\\n\\n- Less than 30% of total energy intake from fats _(1, 2, 3)_ . Unsaturated fats (found in fish, avocado and nuts, and in sunflower, soybean, canola and olive oils) are preferable to saturated fats (found in fatty meat, butter, palm and coconut oil, cream, cheese, ghee and lard) and _trans-_ fats of all kinds, including both industriallyproduced _trans-_ fats (found in baked and fried foods, and pre-packaged snacks and foods, such as frozen pizza, pies, cookies, biscuits, wafers, and cooking oils and spreads) and ruminant _trans-_ fats (found in meat and dairy foods from ruminant animals, such as cows, sheep, goats and camels). It is suggested that the intake of saturated fats be reduced to less than 10% of total energy intake and _trans-_ fats to less than 1% of total energy intake _(5)_ . In particular, industrially-produced _trans_ -fats are not part of a healthy diet and should be avoided _(4, 6)_ .\\n\\n- Less than 5 g of salt (equivalent to about one teaspoon) per day _(8)._ Salt should be iodized.\\n\\n## **For infants and young children**\\n\\nIn the first 2 years of a child’s life, optimal nutrition fosters healthy growth and improves cognitive development. It also reduces the risk of becoming overweight or obese and developing NCDs later in life.\\n\\nAdvice on a healthy diet for infants and children is similar to that for adults, but the following elements are also important:\\n\\n- Infants should be breastfed exclusively during the first 6 months of life.\\n\\n- Infants should be breastfed continuously until 2 years of age and beyond.\\n\\n- From 6 months of age, breast milk should be complemented with a variety of adequate, safe and nutrient-dense foods. Salt and sugars should not be added to complementary foods.\\n\\nIf you eat red meat, limit consumption to no more than about three portions per week. Three portions is equivalent to about 300 to 500 grams cooked weight of red meat. <u>Consume very little, if any, processed meat.</u> Source: WHO (2020); World Cancer Research Fund and American Institute for Cancer Research 45\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099839401162431298/pdf/IDU152ca6b541085314ca41a22a1bf67750bec89.pdf\", \"pages\": [46]}}]}"}}},{"rowIdx":78,"cells":{"doc_id":{"kind":"string","value":"34239758"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34239758\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"admreg\": \"Other,Other\", \"admreg_key\": \"912972,912972\", \"authors\": {\"0\": {\"author\": \"AlAzzawi,Shireen\"}, \"1\": {\"author\": \"Dang,Hai-Anh H.\"}, \"2\": {\"author\": \"Hlasny,Vladimir\"}, \"3\": {\"author\": \"Abanokova,Ksenia\"}, \"4\": {\"author\": \"Behrman,Jere R.\"}}, \"count\": \"Middle East and North Africa,World\", \"count_key\": \"517203,517191\", \"docna\": {\"0\": {\"docna\": \"Female Headship and Poverty in the Arab\\n Region: Analysis of Trends and Dynamics Based on a New Typology\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"Development Indicators and Data (DECID),Development Data Group (DECDG)\", \"projn\": \"1W-Better Data And Methods For Tracking\\n Global Poverty -- P175686\", \"trustfund\": \"TF0B4291-Outreach of Poverty Monitoring,TF0B4288-Cross-country Poverty Measurement,TF0B6571-Within-Country Poverty Measurement\", \"trustfund_key\": \"1971901,1971916,2076169\", \"prdln\": \"Research Activity\", \"prdln_key\": \"1582145\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34239758\"}, \"repnb\": \"WPS10672\", \"docdt\": \"2024-01-16T00:00:00Z\", \"datestored\": \"2024-01-16T07:16:03Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"Various challenges are thought to render\\n female-headed households (FHHs) vulnerable to poverty in the\\n Arab region. Yet, previous studies have had mixed results\\n and the absence of household panel survey data hinders\\n analysis of poverty dynamics. This paper addresses these\\n challenges by proposing a novel typology of FHHs and\\n analyzes synthetic panels constructed from 20 rounds of\\n repeated cross-sectional surveys spanning the past two\\n decades from the Arab Republic of Egypt, Iraq, Jordan,\\n Mauritania, the West Bank and Gaza, and Tunisia. The paper\\n finds that the definition of FHHs matters for measuring\\n poverty levels and dynamics. Most types of FHHs are less\\n poor than non–FHHs on average, but FHHs with a\\n major share of female adults are generally poorer. FHHs are\\n more likely to escape poverty than households on average,\\n but FHHs without children are the most likely to do so.\\n While more children are generally associated with more\\n poverty for FHHs, there is heterogeneity across countries in\\n addition to heterogeneity across measures of FHHs. The\\n findings provide useful inputs for social protection and\\n employment programs aiming at reducing gender inequalities\\n and poverty in the Arab region.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10672; PEOPLE\", \"display_title\": \"Female Headship and Poverty in the\\n Arab Region: Analysis of Trends and Dynamics Based on a New Typology\", \"disclosure_date\": \"2024-01-16T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099916001162436719/text/IDU11b4274c21f7ee1444b192b010c62ef384a79.txt\", \"listing_relative_url\": \"/research/2024/01/34239758/female-headship-poverty-arab-region-analysis-trends-dynamics-based-new-typology\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099916001162436719/IDU11b4274c21f7ee1444b192b010c62ef384a79\", \"new_url\": \"2024/01/34239758/IDU11b4274c21f7ee1444b192b010c62ef384a79\", \"ext_pub_date\": \"2024-01-16T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Poverty m1326687 695,Gender and Human Development m1326908 919,Poverty Measurement and Analysis m1326941 707,Inequality and Shared Prosperity m1327655 700,Household Surveys m1504071 2761\", \"docm_id\": \"IDU-1b4274c2-f7ee-444b-92b0-0c62ef384a79\", \"chronical_docm_id\": \"IDU-1b4274c2-f7ee-444b-92b0-0c62ef384a79\", \"publishtoextweb_dt\": \"2024-01-16T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10672\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"projectid\": \"P175686\", \"guid\": \"099916001162436719\", \"no_of_pages\": \"97\", \"author_keys\": {\"0\": {\"author_key\": \"NA\"}, \"1\": {\"author_key\": \"747967\"}, \"2\": {\"author_key\": \"1253459\"}, \"3\": {\"author_key\": \"1288641\"}, \"4\": {\"author_key\": \"86114\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099916001162436719/IDU11b4274c21f7ee1444b192b010c62ef384a79\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10672 Female Headship and Poverty in the Arab Region Analysis of Trends and Dynamics Based on a New Typology _Shireen AlAzzawi Hai-Anh Dang Vladimir Hlasny Kseniya Abanokova Jere Behrman_ Development Economics Development Data Group January 2024\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10672\\n\\n# **Abstract**\\n\\nVarious challenges are thought to render female-headed households (FHHs) vulnerable to poverty in the Arab region. Yet, previous studies have had mixed results and the absence of household panel survey data hinders analysis of poverty dynamics. This paper addresses these challenges by proposing a novel typology of FHHs and analyzes synthetic panels constructed from 20 rounds of repeated cross-sectional surveys spanning the past two decades from the Arab Republic of Egypt, Iraq, Jordan, Mauritania, the West Bank and Gaza, and Tunisia. The paper finds that the definition of FHHs matters for measuring poverty levels and dynamics. Most types of FHHs are less poor than non– FHHs on average, but FHHs with a major share of female adults are generally poorer. FHHs are more likely to escape poverty than households on average, but FHHs without children are the most likely to do so. While more children are generally associated with more poverty for FHHs, there is heterogeneity across countries in addition to heterogeneity across measures of FHHs. The findings provide useful inputs for social protection and employment programs aiming at reducing gender inequalities and poverty in the Arab region.\\n\\nThis paper is a product of the Development Data Group, Development Economics. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at hdang@worldbank.org.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [{\"mention_name\": {\"text\": \"household panel survey data\", \"confidence\": 0.9204021692276001, \"start\": 232, \"end\": 259, \"clean_text\": \"household panel survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 232, \"end\": 259, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 232, \"end\": 259, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 232, \"end\": 259}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 232, \"end\": 259, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8196063041687012, \"start\": 232, \"end\": 259}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9204021692276001, \"start\": 232, \"end\": 259}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6291412115097046, \"start\": 232, \"end\": 259}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6291412115097046, \"start\": 232, \"end\": 259}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# **Female Headship and Poverty in the Arab Region: Analysis of Trends and Dynamics Based on a New Typology**\\n\\nShireen AlAzzawi, Hai-Anh Dang, Vladimir Hlasny, Kseniya Abanokova, and Jere Behrman<sup>*</sup>\\n\\n**Keywords** : poverty, feminization, female-headedness typology, synthetic panels, Arab region, household surveys\\n\\n**JEL Codes** : I3, J16, N35, O1\\n\\n> * AlAzzawi (salazzawi@scu.edu) is lecturer, Santa Clara University, Santa Clara CA, USA; Dang (hdang@worldbank.org; corresponding author) is senior economist, Living Standards Measurement Unit, Development Data Group, World Bank, USA and is also affiliated with IZA, Indiana University, and London School of Economics and Political Science; Hlasny (vhlasny@gmail.com) is economic affairs officer, UN ESCWA, Beirut, Lebanon; Abanokova (kabanokova@worldbank.org) economist, Living Standards Measurement Unit, Development Data Group, World Bank, USA; Behrman (jbehrman@econ.upenn.edu) is WR Kenan Jr Professor of Economics & Sociology, University of Pennsylvania, USA. We would like to thank Kathleen Beegle, Nazmul Chaudhury, Ebad Ebadi, Daniel Halim, Emily Hannum, Dean Jolliffe, Michael Kevane, Mahdi Majbouri, and participants at the American Economic Association meeting (New Orleans) and the Central Banks of the Middle East and North Africa's Annual Conference on Development Economics (Rabat) for helpful comments on an earlier draft. We would also like to thank the Chief Economist’s Office for the Middle East and North Africa (MNACE) for guidance and financial support and the UK Foreign Commonwealth and Development Office (FCDO) for additional funding assistance through the Data and Evidence for Tackling Extreme Poverty (DEEP) Research Program.\", \"datasets\": [{\"mention_name\": {\"text\": \"household surveys\", \"confidence\": 0.5197855234146118, \"start\": 306, \"end\": 323, \"clean_text\": \"household surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 306, \"end\": 323, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 306, \"end\": 323, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 306, \"end\": 323}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 306, \"end\": 323, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9276911020278931, \"start\": 306, \"end\": 323}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5197855234146118, \"start\": 306, \"end\": 323}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6369195580482483, \"start\": 306, \"end\": 323}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6369195580482483, \"start\": 306, \"end\": 323}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **1. Introduction**\\n\\nAre female-headed households (FHHs) in the Arab region more likely to be poor, and increasingly so? Social and cultural barriers often hinder women’s economic participation in the region and several recent studies find that women are at an increasing disadvantage compared to men in labor markets (Amara and Jemmali, 2018; AlAzzawi and Hlasny, 2022). The COVID-19 pandemic further deepened gender inequality in many countries (Dang and Nguyen, 2021; Alon _et al._ , 2022). Yet, few studies have investigated the topic of poverty feminization in the region. These are important policy questions since many countries in the region specifically target FHHs based on the premise that they are more vulnerable to poverty and probably increasingly vulnerable to poverty, particularly during the pandemic.<sup>1</sup> Furthermore, the Arab region is home to countries of different income levels with diverse social and cultural circumstances, resulting in different context-specific factors contributing to poverty feminization.\\n\\nThat poverty is more prevalent among women than men is widely assumed, and various explanations are offered for it. These include lower school enrollment rates and less work experience (Grant and Behrman, 2010), limited access to income-generating assets such as land (Deere and Leon, 2003), credit and other financial services (Demirguc-Kunt _et al._ , 2013), physical and social capital, and technology (World Bank, 2011; Klasen _et al._ , 2015), and market discrimination (Buvinic and Gupta, 1997).\\n\\nThere is, however, far less agreement on the existence of “feminization of poverty” affecting FHHs (Chant, 2010; Duflo, 2012; Klasen _et al._ , 2015; Bradshaw _et al._ , 2017). Buvinic and Gupta (1997) observe that of 65 studies covering Africa, Asia, Latin America and the Caribbean, 38\\n\\n> 1 These include the Arab Republic of Egypt’s largest poverty-targeting cash transfer program, Takaful, and other subsidy programs in Jordan, Lebanon, and Tunisia (NAF, 2020; Nasri, 2020; ESCWA, 2021; World Bank, 2022).\\n\\n2\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"studies find that FHHs were overrepresented among the poor, 15 others found that poverty was associated with certain FHH types, and the remaining eight studies show no such relationship. While Quisumbing _et al._ (2001) and Medeiros and Costa (2008) find FHHs to be consistently poorer in only 10 developing countries in Africa, Asia and Latin America, Chant (2003) fails to obtain a similar finding in studies for the three continents <mark>. More recently,</mark> Milazzo and van de Walle (2017) even find that despite a growing population share of FHHs in Africa, FHHs saw faster poverty reduction than male-headed households (MHHs). Furthermore, one particular challenge in understanding the current literature on poverty feminization is variations in how FHHs are defined (see Appendix A, Table A.1; we return to this discussion in the next section).\\n\\nSeveral authors investigated the gender dimension of poverty in the Arab region before the onset of the Arab Spring uprisings in 2011 (Nassar, 1997; Datt _et al._ , 2001; El-Laithy, 2001). More recent post Arab-Spring studies examined poverty dynamics for the whole population and different population groups (e.g., Dang and Ianchovichina, 2018). Yet, these studies do not investigate the gender prism; just a few studies explicitly examine poverty feminization related to FHHs but only on a single-country basis (AlAzzawi, 2018; Amara and Jemmali, 2018; AbdelLatif _et al._ , 2019). Furthermore, these few studies stopped short of examining poverty dynamics for FHHs due to the lack of panel data.<sup>2</sup> These knowledge gaps prevent efficient and cost-effective policy interventions, since policies that address chronic poverty could be quite different from those that tackle transient poverty.<sup>3</sup>\\n\\n> 2 Mixed results exist regarding static poverty across countries. For example, comparison between MHHs and FHHs by self-reported headship revealed that for the Arab Republic of Egypt, FHHs are less likely to be poor (AlAzzawi, 2018; AbdelLatif _et al._ , 2019) while the opposite result holds for Tunisia (Amara and Jemmali, 2018).\\n\\n> 3 For example, while social safety-net programs better address transient poverty (e.g., as they help prevent the nonpoor but vulnerable households from falling into poverty), longer-term investments in human capital and infrastructure can tackle chronic poverty. See, e.g., Barret (2005) and Ravallion (2016) for further discussion on different policy interventions regarding chronic poverty versus transitory poverty.\\n\\n3\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"Our study makes several new contributions to the literature. First, we propose and evaluate a novel typology of FHHs consisting of four main types and several sub-types, which are based on self-reported responses, demographic characteristics, and socio-economic characteristics. This approach allows us to employ more nuanced headship definitions that reach beyond the traditional identification solely based on the household head’s gender to better include other aspects of household female composition. Our proposed typology also calls for more attention to the important role of children in defining FHHs, since FHHs with children could show remarkably different, static and dynamic, poverty outcomes from those of FHHs without any children or those of non-FHHs. Second, we study the trends in the FHH poverty–gender nexus and poverty dynamics, for six countries across the Arab region—namely the Arab Republic of Egypt, Iraq, Jordan, Mauritania, the West Bank and Gaza (Palestine), and Tunisia—for which little knowledge exists regarding poverty by FHH status.\\n\\nThird, despite the absence of actual panel data, we construct synthetic panels that allow us to examine FHH poverty dynamics for these countries. By conducting analyses on both poverty incidence and dynamics, we contribute to a better understanding of the dynamic economic wellbeing of FHHs over time. In fact, to our knowledge, we offer the first multi-country study that investigates to what extent FHH poverty exists across a number of countries in the Arab region, and whether FHHs, according to a variety of household types, are more likely to enter or escape poverty over time, using recent survey data. We also make a new data contribution by carefully assembling and harmonizing relevant, up-to-date surveys from multiple sources in a region that is well recognized for limited data access.\\n\\nWe find that the shares of FHHs widely vary, ranging from 10 percent to more than 40 percent depending on the countries and definitions. Compared with non-FHHs, most types of FHHs 4\", \"datasets\": [{\"mention_name\": {\"text\": \"recent survey data\", \"confidence\": 0.3479207754135132, \"start\": 1656, \"end\": 1674, \"clean_text\": \"recent survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1656, \"end\": 1674, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1656, \"end\": 1674, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1656, \"end\": 1674}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1656, \"end\": 1674, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9865128993988037, \"start\": 760, \"end\": 778}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3479207754135132, \"start\": 760, \"end\": 778}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8571534156799316, \"start\": 760, \"end\": 778}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8571534156799316, \"start\": 760, \"end\": 778}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"(including self-reported, potential, and most-educated-female-adult FHHs) are 1 percent to 4 percent less likely to be poor while majority-female-adult FHHs are 3 percent to 5 percent more likely to be poor.\\n\\nWe also find considerable mobility in and out of poverty over the past decade, with the average poor FHH having between 21 and 54 percent chance of escaping poverty, depending on the country. Yet, country heterogeneity exists, with Iraq, Jordan, and Mauritania having upward mobility rates of between 41 and 54 percent, and Egypt, the West Bank and Gaza and Tunisia having upward mobility rates between 21 and 31 percent. More children are generally associated with more poverty and lower chances of escaping poverty for FHHs. The upward mobility rates out of poverty for FHHs without children, FHHs with children, and non-FHHs across all countries are respectively 42 percent, 37 percent, and 36 percent. The corresponding figures for downward movement into poverty for these FHH types are respectively 14 percent, 17 percent, and 19 percent. Our results on mobility are robust to different definitions of FHHs, alternative estimation models, and sample sizes.\\n\\nThe rest of the paper is organized as follows. Section 2 discusses the various definitions of female headship in the existing literature before proposing our new typology of female-headed households (Section 2.1) and introduces the synthetic panel method that allows us to assess FHH poverty feminization dynamically without actual panel data (Section 2.2). Section 3 reviews the available data (Section 3.1), welfare aggregates and standardization measures (Section 3.2), and presents descriptive statistics (Section 3.3). Section 4 reports the main results for cross-sectional poverty (Section 4.1) and poverty dynamics (Section 4.2), and Section 5 concludes with key findings and policy implications. Appendixes A and B present additional estimation results and 5\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"descriptive statistics, Appendix C discusses the synthetic-panel method, and Appendix D provides further analyses with equivalence scales.\\n\\n# **2. Analytical Framework**\\n\\n# **2.1. Typology of female-headed households**\\n\\nHouseholds vary in their compositions and socioeconomic characteristics. In the countries in our sample, the majority of households are comprised of a married couple with one or two income earners, with or without children. Single-head households vary broadly: from widowed retirees who may have already worked for many years and are now living with older children who might be supporting them, to a middle-aged mother who got divorced or lost her husband and is struggling to meet ends by joining the labor market for the first time. Among this group, the presence of another adult male, whether an earner or not, is yet another confounding factor, as well as the presence of children.\\n\\nThe heterogeneous nature of FHHs and the need to separately study different FHH types have been discussed extensively in the literature on poverty feminization (e.g. Kabeer, 1997; Quisumbing, _et al._ , 2001; Klasen _et al._ , 2015; Beegle _et al._ , 2016; Munoz Boudet _et al._ , 2018). The variety of household-headship designations in existing studies has led to mixed results regarding poverty feminization and dynamic patterns. Our reading of some selected studies in the past two decades suggests that while FHHs are not observed to be poorer than non-FHHs in many cases, FHHs can be poorer or have lower consumption depending on the specific type and country context (Appendix A, Table A.1). Advancing an FHH typology can thus be critical for clearing the apparent inconsistencies and for re-classifying households with what may be considered “ _de facto_ female heads” (based on demographic or socioeconomic characteristics), as opposed to “ _de_ 6\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"_jure_ female heads” (based on official status or self-reported information). This also has important implications for poverty reduction efforts targeted at vulnerable population groups.\\n\\nFigure 1 presents our proposed typology of FHHs, which consists of several layers. For the first layer, existing studies can be broadly grouped under two categories: “ _de jure_ FHH” and “ _de facto_ FHH” (second row). For the second layer, we consider three approaches under these two groups: the self-reported approach (under “ _de jure_ FHH”), and the demographic approach and the socioeconomic approach (under “ _de facto_ FHH”) (third row). For the third layer, we consider four main types of FHHs under these three approaches: i) Type 1: self-reported FHHs, ii) Type 2: FHHs defined using the majority share of females among adults in the household (i.e., majority-femaleadult FHHs), iii) Type 3: potential FHHs, and iv) Type 4: FHHs defined as those where the mosteducated adult member is female and no working-age employed male is present (i.e., mosteducated-female-adult FHHs) (fourth row). Furthermore, under these four main types, we also consider five alternative sub-types, which include _de jure_ and married FHHs (under self-reported FHHs), FHHs defined using the majority share of employed females in the household (under majority-female-adult FHHs), and asset and core FHHs (under potential FHHs) (fifth row).<sup>4</sup> Finally, all these types of FHHs should be considered separately with or without any children (last row), since the presence of children plays a crucial role in FHHs’ poverty as discussed below.\\n\\nThe typology is motivated by both our review of the literature and our empirical analysis for each type of FHHs in the Arab region. Figure 1 briefly refers to some illustrative studies that employ these approaches and we elaborate below on this new typology.\\n\\n# **_“De jure” FHHs: Self-reported approach_**\\n\\n> 4 While we propose these three approaches and types for classification purposes, they are not mutually exclusive and existing studies have combined one or more in defining female headship.\\n\\n7\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"A natural departure point to analyzing FHHs is to adopt the self‐reported identification of the head by the survey respondent (our first type, self-reported FHHs), which falls under the _de jure_ FHHs group. Marital status is a key characteristic in this respect. A large share of FHHs are formed as the result of a major marital shock such as divorce or widowhood. If, prior to the shock, the husband was the primary income-earner, the newly-formed FHH may be more likely to fall into poverty (Brown and van de Walle, 2021). FHHs formed through widowhood, especially at a young age with children present, can exhibit both more poverty and higher persistence of poverty (Appleton, 1996; Dreze and Srinivasan, 1997; Horrell and Krishnan, 2007; van de Walle, 2013; Munoz Boudet _et al._ , 2018; Brown and van de Walle, 2021) than FHHs formed largely “by choice”, through divorce or migration of the male spouse (Quisumbing, _et al._ , 2001; Klasen _et al._ , 2015; Beegle _et al._ , 2016; Bradshaw _et al._ , 2017).\\n\\nFemales who never marry or who seek divorce might have chosen this status because they have strong prospects for supporting their newly-formed households on their own, such as higher personal incomes or enabling family-support systems. Ignoring such considerations could mask differences between self-reported FHHs that are financially secure and those that are economically vulnerable (Kabeer, 1997; van de Walle, 2013; Milazzo and van de Walle, 2017). Consequently, it may also be useful to consider alternative types of households based on their marital status— never married, married, divorced or separated, and widowed.\\n\\nIn our samples, most self-reported MHHs are married, and this group is the largest in the sample. By contrast, from 69 to 77 percent of self-reported FHHs in all six countries are widowed, except for Mauritania and the West Bank and Gaza, where 35 and 53 percent, respectively, of selfreported FHHs are widowed households when considering all years together (Appendix B, Tables B.1-B.7). The second-largest group of self-reported FHHs have married heads, but this share 8\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"typically remains about 20 percent or below in all countries and years, except for Mauritania where it rose rapidly to almost 40 percent after 2008.<sup>5</sup>\\n\\n# **“** **_De facto” FHHs: Demographic and socioeconomic approaches_**\\n\\nThe term “head” is a loaded term carrying strong connotations about household decisionmaking power that has traditionally been given to the oldest-male member (regardless of his breadwinner status). This is certainly an issue in the Arab region, where the traditional patriarchal system may preclude designating the female as “head” in the presence of a disabled adult male or a son (regardless of age), even if the woman is the main income-earner.\\n\\nA _de facto_ FHH can thus be defined as one where the male head is temporarily or regularly absent, or (if co-resident) is not the main breadwinner (Buvinic and Youssef, 1978; Buvinic _et al._ , 1983; Klasen _et al._ , 2015). _De facto_ headship accounts for the demographic composition of the household, as well as the socioeconomic circumstances determining the respective members’ relative contributions to household resources (Rosenhouse, 1989; Handa, 1994, 1996; Rogers, 1995; Varley, 1996; Buvinic and Gupta, 1997; Fuwa, 2000; Posel, 2001; Budlender, 2008; Grown and Valodia, 2010; Chant, 2010; Rogan, 2013; Klasen, 2015).\\n\\nA _de facto_ FHH may be more vulnerable to poverty for several reasons. In many societies, the absence of male connections to local economic and social institutions can be debilitating. _De facto_\\n\\n> 5 One complication in classifying married self-reported FHHs arises where one spouse works overseas and sends home remittances, which is common in the region. If the overseas spouse is the husband, the stay-behind spouse might or might not designate herself as the household head in his absence. This can underestimate poverty among “true” FHHs, where the female head does not rely on others for support, since some of the self-declared female heads or main income-earners are in fact temporary designees while the main income-contributing spouses are overseas. In the surveys for all years, remittances are major sources of income for self-reported FHHs, consisting for example of 68% of the consumption per capita in Egypt, and 37% of the consumption per capita in Jordan (Appendix B, Tables B.1 and B.4, all years columns). However, the surveys lump together domestic and overseas remittances and do not allow any further breakdown or provide information on the amount of overseas remittances. The surveys do not identify the relations between the remitters and the households, which complicates matters as such remittances might be alimony or in-kind support.\\n\\n9\", \"datasets\": [{\"mention_name\": {\"text\": \"surveys for all years\", \"confidence\": 0.3183646500110626, \"start\": 2083, \"end\": 2104, \"clean_text\": \"surveys for all years\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2083, \"end\": 2104, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2083, \"end\": 2104, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2083, \"end\": 2104}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2083, \"end\": 2104, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9893665909767151, \"start\": 637, \"end\": 658}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3183646500110626, \"start\": 637, \"end\": 658}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7303478121757507, \"start\": 637, \"end\": 658}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7303478121757507, \"start\": 637, \"end\": 658}}, {\"mention_name\": {\"text\": \"surveys\", \"confidence\": 0.38753002882003784, \"start\": 2515, \"end\": 2522, \"clean_text\": \"surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2515, \"end\": 2522, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2515, \"end\": 2522, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2515, \"end\": 2522}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2515, \"end\": 2522, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9653307199478149, \"start\": 217, \"end\": 224}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.38753002882003784, \"start\": 217, \"end\": 224}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.44756925106048584, \"start\": 217, \"end\": 224}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.44756925106048584, \"start\": 217, \"end\": 224}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"FHHs residing with the female heads’ fathers or older sons may still be better off than FHHs who do not have support of working-age males; for example, agricultural production may become especially harder due to fewer working-age household members working on the farm (Rogan, 2013; Brown and van de Walle, 2021). These FHHs may also have less access to productive assets such as livestock or extension services. Moreover, women in Arab labor markets have far fewer job opportunities compared to men. Their labor-force participation rates are currently among the lowest in the world; their unemployment rates are also four times the world average (UNDP, 2022). When they do work, they tend to face wage and occupational discrimination (Elhamidi and Said, 2008) or are overrepresented in the informal sector with low pay and no social insurance. Residing in a majority-female household, or in one where the majority of earners are female, affects the propensity of being poor (Rogan, 2013; Munoz Boudet _et al._ , 2018).\\n\\nIdeally, an objective criterion would be used to assign headship to the family member whose income or decision-making contributes most to maintaining the family. For instance, Gammage (1998) found that using the _maintenance_ criteria to define FHHs in El Salvador and Costa Rica results in a markedly higher percentage of such female-maintained households (FMHs) compared to the _de jure_ FHH group, and higher poverty incidence. Unfortunately, household surveys in the Arab region do not provide information about individual incomes or total earnings, only aggregates for the household.\\n\\nIn the absence of such information, based on our review of the literature (Table A.1 in Appendix A), we propose several alternative definitions of _de facto_ FHHs using demographic and socioeconomic criteria to provide a more multifaceted understanding of FHHs. Regarding household composition, we consider households where the _proportion of females among (workingage) adults exceeds 0.5_ (second type, majority-female-adult FHHs). We define _potential FHHs_ as 10\", \"datasets\": [{\"mention_name\": {\"text\": \"household surveys\", \"confidence\": 0.888959527015686, \"start\": 1466, \"end\": 1483, \"clean_text\": \"household surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1466, \"end\": 1483, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1466, \"end\": 1483, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1466, \"end\": 1483}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1466, \"end\": 1483, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9798505306243896, \"start\": 655, \"end\": 672}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.888959527015686, \"start\": 655, \"end\": 672}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5156755447387695, \"start\": 655, \"end\": 672}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5156755447387695, \"start\": 655, \"end\": 672}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"those households where there are no working-age males present (third type, potential FHHs). The final type of FHHs combines the demographic and socioeconomic criteria and consists of _households with no employed males, whose most-educated adult member is female_ (fourth type, most-educated-female-adult FHHs).\\n\\nNotably, these main types of FHHs can also include subcategories. For example, under the second-type majority-female-adult FHHs, we can consider those where the _proportion of employed females exceeds that of employed males (majority-employed-female-adult FHHs)._ Similarly, under the third-type potential FHH, we can consider a subcategory called _core FHH_ that encompasses only the potential FHHs with employed females, and another subcategory called _asset FHH_ encompassing only households with females who have ownership rights over the dwellings they reside in.\\n\\n# **_Key confounding factors: Presence of children_**\\n\\nTo account for additional household circumstances interacting with household poverty status, we should distinguish FHHs _with and without children_ . Access to childcare affects women’s labor force participation (LFP) in many countries around the world, rich and poor alike (Akgunduz and Plantega, 2018; Clark _et al._ , 2019). In European countries, childless women (with or without a partner) and single mothers have higher personal earnings than women whose family trajectories combined parenthood and partnership (Muller _et al._ , 2020). In Egypt, childcare similarly presents a considerable barrier to women’s employment (Caria _et al._ , 2022). Yet, only a handful of previous studies have examined how poverty differs with and without children, but mostly for self-reported FHHs (Medeiros and Costa, 2008; Liu _et al._ , 2017; AlAzzawi, 2018). Exceptions include Munoz Boudet _et al._ (2018) and Munoz Boudet _et al._ (2021), who look at household gender composition 11\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"with and without children. Importantly, the common finding in these few studies is that FHHs with children are generally poorer than FHHs without children.\\n\\nFurthermore, a related economics literature on equivalence scales suggests that scale adjustments (for different numbers of adults and children) could have substantial effects on poverty and profiles of the poor for various countries at different income levels (Lanjouw and Ravallion, 1995; Newhouse _et al._ , 2017; Abanokova _et al._ , 2022). This is especially relevant for FHHs; for example, FHHs tend to have lower numbers of household members, but higher dependency ratios (World Bank, 2011; Klasen _et al._ , 2015). In our sample (Appendix B, Table B.7), across all years and countries, the average size of FHHs is 5.7, while that of MHHs is 7.4. The average age of female heads is much higher than that of male heads (56 for FHHs vs. 48 for MHHs). Female heads are also mostly widowed (70.7%, compared to 1.3% for male heads).\\n\\nFor widowed FHHs, their offspring are typically already-grown, independent adults who might be contributing to household expenses from their own earnings. The current welfare of these female heads is likely a function of their accumulated earnings, or more likely those of their deceased or living spouses and adult children, and thus are not strictly comparable to (male or female) heads with young children who rely on current labor market earnings to support themselves and their families. This distinction is especially pertinent to dynamic analysis. If the full sample of female or male heads were treated as a single group, this would unduly bias the results in favor of the elderly, widowed female heads without children, and against the much younger working male heads with children. Consequently, it is important to examine poverty incidence and dynamics among FHHs with or without children.\\n\\nIn summary, our new typology consists of four main types of FHHs and their associated five sub-types (variants). We investigate poverty trends and dynamics of these four types of FHHs for 12\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"households with and without children, further differentiating between those with different numbers of children.\\n\\n# **2.2. Empirical framework**\\n\\nWe provide both static and dynamic analyses of FHH (headcount) poverty in the Arab region. For static analysis, we examine the differences in poverty between different types of FHHs and non-FHHs. Specifically, we estimate the following linear probability model i hF i hF i i i i h′ i i (1) iih iih iih iih iih ii ii iih where ppi is a binary variable representing the poverty status (i.e., equals 1 if poor and 0 = γγ FF + θθ FF ∗CChi + ββ ZZ + μμ + ττ + εε hll otherwise) for household pp _i_ , _i= 1,..., n_ in survey round _j_ , _j_ = 1 or 2, country _c, c_ = 1,..., 6, for FHH type _h_ (F i ). i i i iih iih<sup>is the number of young children age 0-14 in the household (who</sup> are generally not old enough to enter the labor force). FF CChi i is a vector of control variables, iih including household employment and demographic characteristics and residence area (i.e., urban/ ZZ rural residence). and are respectively the country and survey round (year) fixed effects that ii ii control for unobserved macro factors that can affect the whole country or outcomes in specific μμ ττ years, and i is the error term. iih In Equation (1), εε h and h are the coefficients of interest. Compared to non-FHHs, h presents the association between poverty and different types of FHHs without any children, γγ θθ γγ h h presents this association for FHHs with exactly one child, and so on. For easier interpretation, we γγ + θθ ���������������� can also fix the number of children at the mean ( i ı ) and consider the association iih between poverty and different types of FHHs with the average number of children as CChııi ii h + ���������������� h i ı . γγ iih θθ CChııi ii 13\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"It is useful to estimate and compare two different versions of Equation (1), one without the control variables i and one with these control variables. If the estimates for h considerably iih change (or weakens) if the control variables are added, this indicates that the specified FHH type’s ZZ γγ exposure to poverty is sensitive to these control variables. Put differently, this presents a test whether the specified FHH type can capture a relationship with poverty that is not explained by the control variables (i.e., how good the definition of the specified FHH type is). The findings based on our review of the literature suggest that FHHs’ exposure to poverty ( h and to some extent h) are likely sensitive to household composition and employment characteristics. γγ For the dynamic analysis, let θθ i represent type- _h_ FHHs’ household consumption (or income) iih per capita, and i h be the poverty line in period yy _j_ for country _c._ We are interested in knowing the unconditional measures of upward poverty mobility such as zz ) (2) ii1iih 1ii ii2iill 2ii which represents the percentage of type-PP(yy < zz _h_ FHHs that are poor in the first survey round (year) but aai yy > zz nonpoor in the second survey round, or the conditional upward mobility measures such as ) (3) ii2iih 2ii ii1iih 1ii which represents the percentage of poor households in the first round that escape poverty in the PP(yy > zz | yy < zz\\n\\nsecond round.\\n\\nIf true panel data were available, we could straightforwardly estimate the quantities in (2) and (3); but in the absence of such data, we can use synthetic panels to study mobility. We employ recent advances with synthetic panel data methods (Dang and Lanjouw, 2023) to construct synthetic panel data and provide more insights into the dynamics of poverty for FHHs over time.<sup>6</sup>\\n\\n> 6 Recent validations and applications of the synthetic-panel methods by various researchers for different country contexts in Africa, Latin America, the Middle East, and Europe have been encouraging regarding accurate projections\\n\\n14\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"Different from traditional pseudo-panel methods that require multiple rounds of cross-sectional surveys to study poverty mobility at the cohort level, the method that we apply works with as few as two survey rounds and provides poverty estimates at the more disaggregated household level. This method essentially exploits the time-invariant variables across the cross-sectional surveys to link different cohorts, in combination with additional cohort-based assumptions about the error terms, to construct the synthetic panels. Further discussion of this method and detailed estimates are provided in Appendix C.\\n\\n# **3. Data and Descriptive Statistics**\\n\\n# **3.1. Data**\\n\\nWe analyze 20 survey rounds from six countries: Egypt, Iraq, Jordan, Mauritania, the West Bank and Gaza, and Tunisia. For Egypt, we use the Household Income, Expenditure and Consumption Surveys (HIECs) for 2012-2013, 2015, 2017-2018, and 2019-2020; for Iraq, the Household Socio-Economic Survey (IHSESs) for 2007 and 2012; for Jordan the Household Expenditure and Income Surveys (HIESs) for 2010-2011 and 2013-2014; for Mauritania, the Permanent Survey of Living Conditions of Households (EPCVs) for 2004, 2008, 2014, and 2019; for the West Bank and Gaza, the Expenditure and Consumption Survey (PECSs) for 2007, 2009, 2011, and 2016-2017; and for Tunisia, the National Survey on Household Budget, Consumption and Standard of Living (NSHBCs) for 2005, 2010, 2015, and 2021. These surveys provide rich information on household expenditures and various household and individual characteristics for the different household types.\\n\\nof economic status (Ferreira _et al._ , 2012; Beegle _et al._ , 2016; UNDP, 2016; OECD, 2018; Salvuci and Tarp, 2021; Ghomi, 2022).\\n\\n15\", \"datasets\": [{\"mention_name\": {\"text\": \"cross-sectional surveys\", \"confidence\": 0.6640803813934326, \"start\": 80, \"end\": 103, \"clean_text\": \"cross-sectional surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 80, \"end\": 103, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 80, \"end\": 103, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 80, \"end\": 103}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 80, \"end\": 103, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.7397081255912781, \"start\": 80, \"end\": 103}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6640803813934326, \"start\": 80, \"end\": 103}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7530847191810608, \"start\": 80, \"end\": 103}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7530847191810608, \"start\": 80, \"end\": 103}}, {\"mention_name\": {\"text\": \"Household Income, Expenditure and Consumption Surveys\", \"confidence\": 0.8255940079689026, \"start\": 812, \"end\": 865, \"clean_text\": \"Household Income, Expenditure and Consumption Surveys\"}, \"acronym\": {\"text\": \"HIECs\", \"confidence\": 0.49630871415138245, \"start\": 867, \"end\": 872, \"clean_text\": \"HIECs\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 812, \"end\": 865, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 812, \"end\": 865}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 812, \"end\": 865, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9992302656173706, \"start\": 812, \"end\": 865}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8255940079689026, \"start\": 812, \"end\": 865}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9454901814460754, \"start\": 812, \"end\": 865}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9454901814460754, \"start\": 812, \"end\": 865}}, {\"mention_name\": {\"text\": \"Household Socio-Economic Survey\", \"confidence\": 0.7391727566719055, \"start\": 935, \"end\": 966, \"clean_text\": \"Household Socio-Economic Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.3608025312423706, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 935, \"end\": 966, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 935, \"end\": 966}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 935, \"end\": 966, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9992302656173706, \"start\": 935, \"end\": 966}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7391727566719055, \"start\": 935, \"end\": 966}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9454901814460754, \"start\": 935, \"end\": 966}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9454901814460754, \"start\": 935, \"end\": 966}}, {\"mention_name\": {\"text\": \"Household Expenditure and Income Surveys\", \"confidence\": 0.7777807712554932, \"start\": 1010, \"end\": 1050, \"clean_text\": \"Household Expenditure and Income Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1010, \"end\": 1050, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1010, \"end\": 1050, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1010, \"end\": 1050}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1010, \"end\": 1050, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9992302656173706, \"start\": 1010, \"end\": 1050}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7777807712554932, \"start\": 1010, \"end\": 1050}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9454901814460754, \"start\": 1010, \"end\": 1050}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9454901814460754, \"start\": 1010, \"end\": 1050}}, {\"mention_name\": {\"text\": \"Permanent Survey of Living Conditions of Households\", \"confidence\": 0.7697837352752686, \"start\": 1108, \"end\": 1159, \"clean_text\": \"Permanent Survey of Living Conditions of Households\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1108, \"end\": 1159, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1108, \"end\": 1159, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1108, \"end\": 1159}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1108, \"end\": 1159, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9965062141418457, \"start\": 262, \"end\": 313}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7697837352752686, \"start\": 262, \"end\": 313}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7478246688842773, \"start\": 262, \"end\": 313}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7478246688842773, \"start\": 262, \"end\": 313}}, {\"mention_name\": {\"text\": \"Expenditure and Consumption Survey\", \"confidence\": 0.7654843926429749, \"start\": 1232, \"end\": 1266, \"clean_text\": \"Expenditure and Consumption Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1232, \"end\": 1266, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1232, \"end\": 1266, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1232, \"end\": 1266}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1232, \"end\": 1266, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9965062141418457, \"start\": 386, \"end\": 420}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7654843926429749, \"start\": 386, \"end\": 420}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7478246688842773, \"start\": 386, \"end\": 420}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7478246688842773, \"start\": 386, \"end\": 420}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"Several of these surveys were harmonized by the Economic Research Forum (Egypt’s 20122013, 2015, 2017-2018 HIECSs; Iraq’s 2007 and 2012 IHSESs; Jordan’s 2010-2011, and 20132014 HIESs; the West Bank and Gaza’s 2009 and 2011 PECSs; and Tunisia’s 2005 and 2010 NSHBCs. The most recent surveys for Egypt (2019-2020), the West Bank and Gaza (2016-2017), Tunisia (2015 and 2021), and the Mauritanian EPCVs were obtained from national statistical agencies CAPMAS, PCBS, INS and ONS, respectively. We implemented careful harmonization of these surveys with the previous survey years and translated the variables from Arabic or French to English.\\n\\nWe present the poverty lines for the six countries in Tables A.2-A.7 (Appendix A), compiling them from official sources and World Bank publications. We used region-specific poverty lines within each country to account for spatial differences in consumption (expenditure) patterns and price levels.<sup>7</sup> Since our focus is on poverty analysis, we used consumption values and national poverty lines in local currency units and in survey-year prices to sidestep conversion issues (e.g., with the PPP or market exchange rates) and adjustment for inflation.\\n\\n# **3.2. Living-standards indicator**\\n\\nExpenditure is widely regarded as a better indicator of permanent income when households, particularly in poorer countries, exercise consumption smoothing and use savings to augment unstable incomes due to seasonal or informal employment or unexpected shocks (Deaton, 1997; Deaton and Zaidi, 2002; Mancini and Vecchi, 2022). We use household consumption expenditures\\n\\n> 7 We were able to do this for all countries in our sample except for Jordan. According to DOS reports, Jordan’s Department of Statistics (DOS) did not publish region-specific poverty lines and used a single poverty line for all of Jordan in 2010 and 2013. Jordan’s DOS does not publish region-specific Consumer Price Indices so we were unable to take spatial price differences into consideration.\\n\\n16\", \"datasets\": [{\"mention_name\": {\"text\": \"HIECSs\", \"confidence\": 0.4404199421405792, \"start\": 107, \"end\": 113, \"clean_text\": \"HIECSs\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 107, \"end\": 113, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 107, \"end\": 113, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 107, \"end\": 113}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 107, \"end\": 113, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9943522214889526, \"start\": 107, \"end\": 113}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4404199421405792, \"start\": 107, \"end\": 113}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4220644235610962, \"start\": 107, \"end\": 113}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4220644235610962, \"start\": 107, \"end\": 113}}, {\"mention_name\": {\"text\": \"NSHBCs\", \"confidence\": 0.4564729928970337, \"start\": 258, \"end\": 264, \"clean_text\": \"NSHBCs\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 258, \"end\": 264, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 258, \"end\": 264, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 258, \"end\": 264}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 258, \"end\": 264, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9943522214889526, \"start\": 258, \"end\": 264}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4564729928970337, \"start\": 258, \"end\": 264}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4220644235610962, \"start\": 258, \"end\": 264}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4220644235610962, \"start\": 258, \"end\": 264}}, {\"mention_name\": {\"text\": \"Mauritanian EPCVs\", \"confidence\": 0.3264755606651306, \"start\": 382, \"end\": 399, \"clean_text\": \"Mauritanian EPCVs\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 382, \"end\": 399, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 382, \"end\": 399, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 382, \"end\": 399}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 382, \"end\": 399, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9907678961753845, \"start\": 382, \"end\": 399}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3264755606651306, \"start\": 382, \"end\": 399}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8084661364555359, \"start\": 382, \"end\": 399}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8084661364555359, \"start\": 382, \"end\": 399}}, {\"mention_name\": {\"text\": \"region-specific poverty lines\", \"confidence\": 0.5714824199676514, \"start\": 796, \"end\": 825, \"clean_text\": \"region-specific poverty lines\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 796, \"end\": 825, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 796, \"end\": 825, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 796, \"end\": 825}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 796, \"end\": 825, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.47212544083595276, \"start\": 74, \"end\": 103}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.5714824199676514, \"start\": 74, \"end\": 103}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9015558958053589, \"start\": 74, \"end\": 103}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9015558958053589, \"start\": 74, \"end\": 103}}, {\"mention_name\": {\"text\": \"household consumption expenditures\", \"confidence\": 0.4764317274093628, \"start\": 1571, \"end\": 1605, \"clean_text\": \"household consumption expenditures\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1571, \"end\": 1605, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1571, \"end\": 1605, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1571, \"end\": 1605}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1571, \"end\": 1605, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.536160409450531, \"start\": 849, \"end\": 883}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4764317274093628, \"start\": 849, \"end\": 883}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9746451377868652, \"start\": 849, \"end\": 883}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9746451377868652, \"start\": 849, \"end\": 883}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"per capita as the welfare measure underlying poverty analysis.<sup>8</sup> This includes all monetary expenditures on consumer goods and non-monetary consumption, such as imputed rents, use-value of durables, own production and in-kind transfers (i.e., gifts) received by households. Food consumption includes food that the household has purchased, grown and received from other sources. Non-food consumption is the sum of expenditure on all non-food items, including expenditure on fuel, clothing, schooling, health and miscellaneous items, and in-kind transfers.\\n\\nIt can be useful to ensure comparability of household expenditures across different contexts to account for potential differences in households’ age and size compositions, as well as economies of scale in consumption. Studies have examined individual-level, rather than household-level, consumption to better disaggregate expenditures by gender (Dunbar _et al._ , 2013; De Vreyer and Lambert, 2021). Unfortunately, the available surveys provide data on household consumption aggregates rather than individual-level consumption; therefore this approach cannot be applied to the available data.\\n\\nAnother approach is to calculate the Adult Equivalent Expenditure (AEE) (or income) for each household, which gives smaller weight to children than adults and takes economies of scale into consideration. For example, Deaton and Paxson (1998) suggest using a parametric form of an equivalence scale, where a child is assumed to require a fraction of what an adult needs, and where the elasticity of needs with respect to adjusted household size is a constant αα . This gives rise to the following formula δδ\\n\\n> 8 This is also the most common approach employed in recent studies of poverty in countries in the Middle East (Marotta _et al._ , 2011; CAPMAS, 2013).\\n\\n17\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"where i∗ is the AEE for household _i_ in survey _j_ , which is an adjusted version of household expenditure conditional on the number of adults yy i and children i (we suppress the country and FHH type indexes for less cluttered notation). The smaller aa is, the smaller the relative weight of kk children; the higher is , the smaller the degree of economies of scale assumed.αα<sup>9</sup> We construct several different AEE levels for each household based on this method, using different values for δδ the weight of children ( ) and degree of economies of scale ( ) and show the results in Figure D.1, Appendix D. The relationship between household size and poverty dynamics reveals varying αα δδ scenarios for FHHs, with FHHs generally having a higher likelihood of escaping poverty than nonFHHs when assessing consumption on a per capita basis.<sup>10</sup>\\n\\n# **3.3. Descriptive statistics**\\n\\nTable 1 presents some key sample statistics on the prevalence of FHHs defined according to our proposed typology (Section 2.1) across the six Arab countries and different (survey) years. The four main types of FHHs are shown in bold while the alternative sub-types are shown in regular font. The shares of self-reported FHHs remain relatively stable over time in most countries, except for Mauritania. In recent years, this share hovers from around 10 percent (Iraq, the West Bank and Gaza) to 13 percent (Jordan) and 18 percent (Egypt, Tunisia).<sup>11</sup> Mauritania has the largest share of self-reported FHHs, which has almost doubled from 18.9 percent in 2004 to 36.6 percent in 2019. The shares of majority-female-adult FHHs are significantly higher in all countries,\\n\\n> 9 When = 1 and = 1, we have per capita expenditure, which assumes no economies of scale and an equal weight for children and adults in the household. 10 These results are consistent with Abanokova δδ αα _et al.’_ s (2022) finding regarding the sensitivity of income dynamics to scale parameters, showing persistent upward mobility when income is evaluated on a per capita basis for the Russian Federation.\\n\\n> 11 These shares are lower than the corresponding figure of 26 percent for African households observed in Milanzo and van de Walle (2017).\\n\\n18\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"ranging from 21 percent (Egypt) to 44 percent (Mauritania) in the most recent years. Potential FHHs are as prevalent as those identified by self-reporting in all the countries except Iraq, where they are half as prevalent. Finally, most-educated-female-adult FHHs have relatively low prevalence rates, ranging from around 6 percent in Iraq to 25 percent in Mauritania. There is a weak-to-medium correlation among the four FHH types (i.e., ranging from 0.27 to 0.51; Appendix A, Table A.8), suggesting that each of the proposed FHH types captures different information about female headship.\\n\\nCompared with the main four types, the alternative subtypes all provide lower-to-almostnegligible prevalence of FHHs. For example, under the self-reported FHH type, while _de jure_ FHHs account for between 8 percent and 17 percent of households for all countries and years, the corresponding figures for married FHHs are between 1 percent and 4 percent for all the countries, except for Mauritania in 2008 and later years. Under the majority-female-adults FHH type, the sub-type majority-employed-female-adult FHHs, however, yield a much smaller group of FHHs (ranging from around one-half to two-thirds as small). This is expected given the very low female LFP rates in the region, especially in such countries as Iraq and Jordan where they are among the lowest globally.\\n\\nFigure 2 illustrates the trends in poverty headcount ratios by country for the four main types of FHHs against those of the whole population for each country. This figure shows that different FHH types display clear differences regarding poverty levels and trends. Specifically, while potential FHHs (purple line) show faster poverty decreases in Iraq, Jordan, the West Bank and Gaza, and Tunisia, most-educated-female-adult FHHs (pink line) show slightly opposite trends from those of the whole population for Iraq. This contrasts with self-reported FHHs (green lines) 19\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"and potential FHHs, which predominantly have less poverty than the whole population for almost all the country-year observations.<sup>12</sup> For each country, Figure 3 and Figure 4 present the poverty differences between FHHs and non-FHHs for the four main FHH types respectively by year and by the number of children (age 0-14). Figure 3 indicates that the self-reported and potential FHH types typically have lower poverty ratios than non-FHH households across most years and countries. However, majorityfemale-adult FHHs tend to be poorer than the respective non-FHHs in most years and countries, except in Egypt 2017–2020. Most-educated-female-adult households have systematically more poverty prevalence than the corresponding non-FHHs in Iraq, Jordan, the West Bank and Gaza and Tunisia, but less poverty prevalence in Egypt and Mauritania.<sup>13</sup> Figure 4 shows that the presence of children is positively associated with poverty prevalence among FHHs for most of the countries, except for Egypt and Mauritania.\\n\\n# **4. Estimation Results**\\n\\n# **4.1. Cross-sectional poverty**\\n\\nTable 2 provides the estimation results for the associations between four main FHH types and poverty ( h in Equation (1)), without and with the household employment, demographic γγ\\n\\n> 12 Pooling data for all years and countries, we further show the FHH–non-FHH poverty differences for all FHH types and by the number of children in households in Figures A.2 and A.3 (Appendix A). These figures indicate that the self-reported and potential FHH types tend to have less poverty than non-FHH households across most years and countries, but the relationship between the number of children and poverty varies across countries.\\n\\n13 Table A.9 (Appendix A) provides cross-sectional poverty rates for different household types over time in six countries. Panels B and C additionally report these poverty rates for rural and urban subgroups, and Panels D and E report the poverty rates for households with children under 14 and without children under 14. Poverty rates are typically higher in rural area than in urban areas, except for the West Bank and Gaza, and higher for households with younger children. Given the consistently high poverty rates among FHHs defined by the share of women among adults, we also assess the poverty rates among self-reported MHHs according to the number of female adults present in Figure A.1 (Appendix A). Poverty almost always increases with the number of females in all six countries, validating the central finding from Figure 3.\\n\\n20\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"characteristics and residence-area variables shown respectively in the first four columns and the second four columns (Appendix A, Table A.10 offers the full results). Several interesting results stand out. First, the estimated h is negative and strongly statistically significant for three FHH types: self-reported, potential, and most-educated-female-adult FHHs, suggesting that these three FHH γγ� types are associated with less poverty prevalence. Majority-female-adult FHHs, in contrast, are associated with more poverty. This is generally consistent with our earlier discussion for Figure 3, indicating that these types of FHHs can serve as useful definitions. Second, the absolute magnitude of h increases for self-reported and most-educated-femaleadult FHHs but decreases for majority-female-adult and potential FHHs when the control variables γγ� are added. The t-tests for these changes are statistically significant. This suggests that, consistent with our earlier discussions of the literature (Sections 2.1 and 2.2), FHHs’ exposure to poverty is also affected by the control variables, including household employment, demographic characteristics, and residence area variables. Indeed, prior research for various countries suggests that FHHs are not systematically poorer or more vulnerable (Fuwa, 2000; Klasen _et al._ , 2015; Munoz Boudet _et al._ , 2018; Brown and Van de Walle, 2021). Liu _et al._ (2017) find that in eight of 14 Latin American countries, FHHs more likely live in poorer conditions, but these gaps either disappear or reverse when controlling for other household and demographic characteristics.\\n\\nTable 2 shows that self-reported FHHs are about 1 percent (without control variables) to 4 percent (with control variables) less likely to be poor than non-FHHs if there are no children in the household. The corresponding changes are about 3 percent (without control variables) to 2 percent (with control variables) for potential FHHs, and 1 percent (with control variables) for mosteducated-female-adult FHHs. However, majority-female-adult FHHs are 5 percent (without 21\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"control variables) to 3 percent (with control variables) more likely to be poor than non-FHHs if there are no children. Finally, the estimated interaction term between FHH types and the number of children ( ) is ll positive for three of the four FHH types (self-reported, potential, and most-educated-female-adult θθ<sup>�</sup> FHHs), but negative for majority-female-adult FHHs. While the absolute magnitudes of h are small, around 1 percent (i.e., one more child is associated with a 1-percent change in the probability θθ<sup>�</sup> of the household being poor), it is strongly statistically significant. Furthermore, when we fix the number of children at the mean of the estimation sample (i.e., 1.81 children), self-reported FHHs become 1 percent more likely to be poor (without control variables) and 3 percent less likely to be poor (with control variables). The corresponding probabilities, without and with control variables, become 2.5-2.8 percent more likely to be poor for most-educated-female-adult FHHs and 4-0.7 percent more likely to be poor for majority-female-adult FHHs. However, potential FHHs are 0.2 percent (without control variables) less likely to be poor and are 0.3 percent (with control variables) more likely to be poor. In addition, having more children (or larger household sizes) is associated with greater poverty risks (Appendix A, Table A.10). This result concurs with the finding by Munoz Boudet _et al._ (2018) and Munoz Boudet _et al._ (2021) that adult couple households with children are the largest and overrepresented group among poor households. This provides supportive evidence for our proposed typology that considers children when defining FHH types.\\n\\nThe five remaining FHH subtypes offer qualitatively similar results, showing that most FHH types are associated with less poverty, except for majority-employed-female-adult FHHs where the opposite result holds (Appendix A, Table A.11). This table also shows the interaction terms between FHH types and the number of children, which are mostly statistically significant. The results using the alternative logit model are qualitatively similar, albeit somewhat weaker for the 22\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"most-educated-female-adult FHHs (Appendix A, Tables A.12 and A.13).<sup>14</sup> We further consider the overlaps of three main FHH types (self-reported, potential, and most-educated-female-adult FHHs) and all four main FHH types and show the estimation results in Appendix A, Table A.14, which remain qualitatively similar.\\n\\n# **4.2. FHH poverty dynamics based on synthetic panels**\\n\\nWe turn next to discussing the results on poverty dynamics based on synthetic panels. For each country, Figure 5 reports the conditional upward mobility rates in the second survey year for the four main FHH types, considered separately with and without any children (Equation (3)). Figure 5 shows considerable (conditional) upward mobility at the national average level (dashed line) for some countries. In particular, the upward mobility rate is 45 percent in Iraq during 2007-2012 (i.e., 45 percent of the initial poor in the first year escape poverty in the second year), 54 percent in Jordan during 2010-2013, and 41 percent in Mauritania during 2014-2019. Still, a significant degree of immobility exists in Egypt, the West Bank and Gaza, and Tunisia, where most of the population remained poor in both years and only about one-third (or less) of the poor escaped poverty in the most recent year: 29 percent for Egypt during 2017–2020, 31 percent in the West Bank and Gaza during 2011–2017, and 21 percent in Tunisia during 2015–2021.<sup>15</sup> Unsurprisingly, non-FHHs have upward mobility rates that are almost the same as the national averages, given their large shares in the population (Table 1). But interestingly, FHHs without\\n\\n> 14 The estimated marginal effects for the interaction terms with children are qualitatively similar (using the Stata command “ginteff” (Radean, 2023)).\\n\\n> 15 The survey period lengths generally differ for the six countries so the estimated mobility rates are not exactly comparable across countries or to those in other studies. For a rough reference, Dang and Ianchovichina (2018) obtain a regional upward mobility rate around 52 percent in the early 2000s and 2010s. However, if we assume a similar rate of change for mobility across the years for all countries, we can obtain the average mobility per year for each country (Appendix A, Table A.15).\\n\\n23\", \"datasets\": [{\"mention_name\": {\"text\": \"synthetic panels\", \"confidence\": 0.39782387018203735, \"start\": 453, \"end\": 469, \"clean_text\": \"synthetic panels\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 453, \"end\": 469, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 453, \"end\": 469, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 453, \"end\": 469}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 453, \"end\": 469, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.3526242673397064, \"start\": 453, \"end\": 469}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.39782387018203735, \"start\": 453, \"end\": 469}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8160843849182129, \"start\": 453, \"end\": 469}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8160843849182129, \"start\": 453, \"end\": 469}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"children are most likely to experience upward mobility. Out of 24 FHH types across six countries, the probabilities of FHHs without any children escaping poverty are higher than the national averages in 22 cases. The exceptions are self-reported and potential FHHs in Jordan during 20102013, which have similar upward mobility rates as the national average. However, FHHs with children have upward mobility rates that are clearly higher than the national averages in five cases (self-reported, majority-female-adult, and most-educated-female-adult FHHs in Egypt during 2017-2020, and self-reported FHHs in Iraq during 2007-2012 and in Mauritania during 2014-2019) and clearly lower than the national averages in six cases (self-reported, majority-female-adult, and most-educated-female-adult FHHs in Jordan 2010-2013 and Tunisia 2015-2021). FHHs with children have similar upward mobility as the national averages for the remaining cases. Overall, across all countries and four main FHH types, the upward mobility rates for FHHs without children, FHHs with children, and non-FHHs are respectively 42 percent, 37 percent, and 36 percent.\\n\\nFigure 6 plots the conditional downward mobility (i.e., falling into poverty in the second year when being initial non-poor in the first year). The results are consistent with those shown in Figure 5, with FHHs without children experiencing the least downward mobility, followed by FHHs with children and non-FHHs. Overall, the downward mobility rates for FHHs without children, FHHs with children, and non-FHHs are respectively 14 percent, 17 percent, and 19 percent.\\n\\nAs an alternative to Figure 5, we plot the results of locally weighted regressions of upward mobility on the number of children (Appendix A, Figure A.3). This figure also shows that the number of children is negatively associated with upward mobility for most countries and FHH types. The results for other sub-types of FHHs are qualitatively similar, with FHHs without children having the most upward mobility (Appendix A, Figure A.4). The results for preceding 24\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"years are, however, somewhat mixed. FHHs without children had the strongest upward mobility for Egypt and Tunisia, but had similar upward mobility as FHHs with children for Mauritania and the West Bank and Gaza (Appendix A, Figures A.5-A.8). Finally, we plot the results for upward and downward mobility for FHH types, with and without children considered together, in Figures A.9 and A.10 (Appendix A). These figures show that FHHs have higher upward mobility and lower downward mobility than MHHs across all FHH types and countries, except for Jordan.<sup>16</sup>\\n\\n# **5. Conclusions and Policy Implications**\\n\\nThe dramatic events of the Arab Spring and the following decade of structural reforms and sectoral developments _inter alia_ have brought to the fore the importance of better understanding gender inequalities. We offer new analysis on the feminization of poverty as related to FHHs, using 20 survey rounds spanning the past two decades for six countries across the Arab region— namely Egypt, Iraq, Jordan, Mauritania, the West Bank and Gaza, and Tunisia—an understudied set of countries. We propose and evaluate a new typology of FHHs consisting of four main types (and several sub-types) of FHHs with a new focus on the presence of children, which offers policyrelevant insights regarding the trends and dynamics of poverty feminization. We assemble and harmonize the available cross-sectional data and construct synthetic panels to address the lack of actual panels.\\n\\nWe find that different FHH types display clear differences regarding poverty levels and trends. In particular, self-reported FHHs, potential FHHs, and most-educated-female-adult FHHs are less likely to be poor than non-FHHs for the six countries, while the opposite holds for majority-female-\\n\\n> 16 While Milazzo and Van de Walle (2017) find self-reported FHHs to be generally poorer, they also find these households to contribute more to the overall decline in poverty in Africa.\\n\\n25\", \"datasets\": [{\"mention_name\": {\"text\": \"20 survey rounds\", \"confidence\": 0.31895071268081665, \"start\": 903, \"end\": 919, \"clean_text\": \"20 survey rounds\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 903, \"end\": 919, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 903, \"end\": 919, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 903, \"end\": 919}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 903, \"end\": 919, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9989151954650879, \"start\": 903, \"end\": 919}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.31895071268081665, \"start\": 903, \"end\": 919}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8874151110649109, \"start\": 903, \"end\": 919}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8874151110649109, \"start\": 903, \"end\": 919}}, {\"mention_name\": {\"text\": \"cross-sectional data\", \"confidence\": 0.8878581523895264, \"start\": 1393, \"end\": 1413, \"clean_text\": \"cross-sectional data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1393, \"end\": 1413, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1393, \"end\": 1413, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1393, \"end\": 1413}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1393, \"end\": 1413, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.5427574515342712, \"start\": 654, \"end\": 674}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8878581523895264, \"start\": 654, \"end\": 674}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.42352136969566345, \"start\": 654, \"end\": 674}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.42352136969566345, \"start\": 654, \"end\": 674}}, {\"mention_name\": {\"text\": \"self-reported FHHs\", \"confidence\": 0.3472796380519867, \"start\": 1595, \"end\": 1613, \"clean_text\": \"self-reported FHHs\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1595, \"end\": 1613, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1595, \"end\": 1613, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1595, \"end\": 1613}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1595, \"end\": 1613, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.356119304895401, \"start\": 0, \"end\": 18}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3472796380519867, \"start\": 0, \"end\": 18}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7882933616638184, \"start\": 0, \"end\": 18}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7882933616638184, \"start\": 0, \"end\": 18}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"adult FHHs. Yet, more children are associated with more poverty for the former three types of FHHs and less poverty for the last type of FHHs.\\n\\nWe also find considerable (conditional) upward mobility, ranging between 21 and 54 percent of the initially poor in a country. But country heterogeneity exists, with Iraq, Jordan, and Mauritania having relatively more upward mobility, while Egypt, the West Bank and Gaza, and Tunisia have relatively less upward mobility. While most types of FHHs more likely experience upward mobility out of poverty (or less likely fall into poverty), FHHs without children have the strongest upward mobility (or the least downward mobility), followed by FHHs with children, and non-FHHs.\\n\\nOur proposed typology aligns with recent calls to go beyond identifying headship based on the gender of the head alone. For example, Beegle and van de Walle (2019) argue that since many women live in MHHs, especially in Sub-Saharan Africa, if resources are unequally shared among household members, simply comparing FHHs and MHHs based on heads’ gender can provide biased results. Summarizing opinions from experts on gender issues and survey design on the topic, Buvinic and van de Walle (2019) similarly call for other definitions based on other household characteristics including demographic and gender composition. Other concerns were also raised about practical survey challenges with headship (e.g., when the male head temporarily lives away from the household).\\n\\nThese discussions do not just serve academic purposes but have practical policy implications. Governments in the region strive to identify various vulnerable FHHs for effective socialprotection interventions aimed at targeting vulnerable groups and reducing gender inequalities. For example, Egypt’s largest poverty-targeting cash-transfer program, Takaful, uses proxy means testing to target households and the criteria include a much lower threshold for FHHs. In 2017, the 26\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [27]}}, {\"page\": 28, \"chunk\": 28, \"input_text\": \"poverty threshold used to determine eligibility was raised considerably for MHHs while it was kept constant for FHHs (ESCWA, 2021), resulting in the share of beneficiaries who were FHHs almost doubling from 48% to 92%. In Jordan, the National Aid Fund targets several categories of “vulnerable” FHHs such as widows with children, those without “support”, and divorced female heads, not just poor FHHs, while its poverty-reduction program directly targets poor FHHs (NAF, 2020; ESCWA, 2021; World Bank, 2022). In Lebanon, and Tunisia, FHHs, especially widows, are also prioritized (Nasri, 2020; ESCWA, 2021).\\n\\nAgainst this background, our findings offer highly relevant policy inputs and run against the conventional wisdom that FHHs are typically poorer than non-FHHs, which appears to be the implicit assumption underlying many targeting programs in the region and elsewhere. In contrast, we find majority-female-adult households or households with more children more vulnerable to (remaining in) poverty. While these results suggest that female headship definition using gender composition can offer an alternative approach—and potentially help identify a more vulnerable group—for poverty targeting, they also highlight the need for a more nuanced understanding of how female headship can be defined, especially in the presence of children. Furthermore, we also need to better understand the extent to which the different types of FHHs’ exposure to poverty can change, depending on various other factors such as whether we examine households’ static or dynamic poverty status, whether other household demographic and employment characteristics are considered, and last but not least, the country-specific contexts.\\n\\n27\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [28]}}, {\"page\": 29, \"chunk\": 29, \"input_text\": \"# **References**\\n\\n- Abanokova, K., Dang, H.A.H. and Lokshin, M., (2022). “Do adjustments for equivalence scales affect poverty dynamics? Evidence from the Russian federation during 1994–2017”. _Review of Income and Wealth_ , 68, pp.S167-S192.\\n\\n- AbdelLatif, Lobna M., Mohamed Ramadan, and Sarah A. Elbakry. (2019). “How gender biased are female-headed household transfers in Egypt?” _Middle East Development Journal_ , Vol. 11(2), pp.165-180.\\n\\n- Akgunduz, Y. E., & Plantenga, J. (2018). Childcare prices and maternal employment: A meta‐ analysis. _Journal of Economic Surveys_ , 32(1), 118-133.\\n\\n- AlAzzawi, Shireen. (2018). “Do Endowments Matter? Exploring the Gender Dimensions of Poverty in Egypt.” _Review of Income and Wealth_ , Vol. 64(S1), pp. S189-S224.\\n\\n- AlAzzawi, Shireen and Vladimir Hlasny. (2022). “Youth Labor Market Vulnerabilities: Evidence from Egypt, Jordan and Tunisia.” _International Journal of Manpower_ , Vol. 43(7), pp. 16701699.\\n\\n- <mark>Alon, Titan, Sena Coskun, Matthias Doepke, David Koll, and Michèle Tertilt. (2022). “From mancession to shecession: Women’s employment in regular and pandemic recessions.”</mark> _<mark>NBER Macroeconomics Annual</mark>_ <mark>, Vol. 36(1), pp. 83-151.</mark>\\n\\n- Amara, Mohamed, and Hatem Jemmali. (2018). “Household and Contextual Indicators of Poverty in Tunisia: A Multilevel Analysis.” _Social Indicators Research_ , Vol. 137, pp. 113-138.\\n\\n- <mark>Barrett, C. B. (2005). Rural poverty dynamics: development policy implications.</mark> _<mark>Agricultural Economics</mark>_ <mark>,</mark> _<mark>32</mark>_ <mark>, 45-60.</mark>\\n\\n- <mark>Beegle, K., & van de Walle, D. (2019). What can female headship tell us about women’s wellbeing? Probably not much.</mark> _<u><mark>https://blogs.worldbank.org/impactevaluations/what-can-femaleheadship-tell-us-about-womens-well-being-probably-not-much</mark></u>_\\n\\n- Beegle, Kathleen, Luc Christiaensen, Andrew Dabalen, and Isis Gaddis. (2016). _Poverty in a Rising Africa_ . Washington, DC: The World Bank.\\n\\n- Bibi, Sami and Rim Chatti. (2010). “Gender Poverty in Tunisia: Is there a Feminization Issue?” _Middle East Development Journal_ , Vol. 2(2), pp. 283-307.\\n\\n- Brown, Caitlin, and van de Walle, Dominique. (2021). “Headship and Poverty in Africa.” _World Bank Economic Review_ , Vol. 35(4), pp. 1038-1056.\\n\\n- Buvinic, Mayra and Geeta Rao Gupta. (1997). “Female-Headed Households and FemaleMaintained Families: Are They Worth Targeting to Reduce Poverty in Developing Countries?” _Economic Development and Cultural Change_ , Vol. 45(2), pp. 259-280.\\n\\n28\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [29]}}, {\"page\": 30, \"chunk\": 30, \"input_text\": \"- Buvinic, M., & van de Walle, D. (2019). The Debate about Headship in Poverty and Gender Studies. https://www.cgdev.org/blog/debate-about-headship-poverty-and-gender-studies\\n\\n- <mark>Caria, S., Crepon, B., ElBehairy, H., Fadlalmawla, N., Krafft, C., Nagy, A., ... & El Assiouty, S. (2022).</mark> _<mark>Child Care Subsidies, Employment Services and Women's Labor Market Outcomes in Egypt: First Midline Results</mark>_ <mark>. World Bank.</mark>\\n\\n- Chant, Sylvia. (2003). “ _Female Household Headship and the Feminization of Poverty: Facts, Fictions and Forward Strategies”,_ LSE Gender Institute, New Working Paper Series, Issue 9 (London: London School of Economics).\\n\\n- CAPMAS. (2014). “Poverty Indicators according to the Household Income, Expenditure and Consumption Survey 2012/2013”. <u>http://capmas.gov.eg/pepo/a.pdf.</u>\\n\\n- <mark>Clark, S., Kabiru, C. W., Laszlo, S., & Muthuri, S. (2019). The impact of childcare on poor urban women’s economic empowerment in Africa.</mark> _<mark>Demography</mark>_ <mark>,</mark> _<mark>56</mark>_ <mark>(4), 1247-1272.</mark>\\n\\n-\\n\\n- Dang, Hai Anh H., and Elena Ianchovichina. (2018). “Welfare dynamics with synthetic panels: the case of the Arab world in transition.” _Review of Income and Wealth_ , Vol. 64, pp. S114S144.\\n\\n- Dang, Hai-Anh and Peter Lanjouw. (2023). “Measuring poverty dynamics with synthetic panels based on cross-sections.” _Oxford Bulletin of Economics and Statistics_ , 85(3): 599-622.\\n\\n- Dang, Hai-Anh and <mark>Cuong Viet N</mark> guyen. (2021). “Gender Inequality during the COVID-19 Pandemic: Income, Expenditure, Savings, and Job Loss.” _World Development_ , Vol. 140, 105296.\\n\\n- Datt, G., D. Jolliffe, and M. Sharma. (2001). “A profile of poverty in Egypt.” _African Development Review_ , Vol. 13(2), pp. 202-237.\\n\\n- Deaton, A., & Paxson, C. (1998). Economies of scale, household size, and the demand for food. _Journal of Political Economy_ , 106(5), 897-930.\\n\\n- DeGraff, Deborah S., and Richard E. Bilsborrow. (1993). “Female-headed households and family welfare in rural Ecuador.” _Journal of Population Economics_ , Vol. 6, pp. 317-336.\\n\\n- <mark>De Vreyer, P., & Lambert, S. (2021). Inequality, poverty, and the intra-household allocation of consumption in Senegal.</mark> _<mark>World Bank Economic Review</mark>_ <mark>,</mark> _<mark>35</mark>_ <mark>(2), 414-435.</mark>\\n\\n- <mark>Deere, C. D., & León, M. (2003). The gender asset gap: Land in Latin America.</mark> _<mark>World Development</mark>_ <mark>,</mark> _<mark>31</mark>_ <mark>(6), 925-947.</mark>\\n\\n- <mark>Deere, C. D., Alvarado, G. E., & Twyman, J. (2012). Gender inequality in asset ownership in Latin America: Female owners vs household heads.</mark> _<mark>Development and Change</mark>_ <mark>,</mark> _<mark>43</mark>_ <mark>(2), 505-530.</mark>\\n\\n29\", \"datasets\": [{\"mention_name\": {\"text\": \"Household Income, Expenditure and Consumption Survey\", \"confidence\": 0.5557495355606079, \"start\": 728, \"end\": 780, \"clean_text\": \"Household Income, Expenditure and Consumption Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 728, \"end\": 780, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 728, \"end\": 780, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 728, \"end\": 780}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 728, \"end\": 780, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.569399893283844, \"start\": 728, \"end\": 780}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5557495355606079, \"start\": 728, \"end\": 780}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9274669289588928, \"start\": 728, \"end\": 780}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9274669289588928, \"start\": 728, \"end\": 780}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [30]}}, {\"page\": 31, \"chunk\": 31, \"input_text\": \"- <mark>Dreze, J., & Srinivasan, P. V. (1997). Widowhood and poverty in rural India: Some inferences from household survey data.</mark> _<mark>Journal of Development Economics</mark>_ <mark>,</mark> _<mark>54</mark>_ <mark>(2), 217-234.</mark>\\n\\n- <mark>Dunbar, G. R., Lewbel, A., & Pendakur, K. (2013). Children's resources in collective households: identification, estimation, and an application to child poverty in Malawi.</mark> _<mark>American Economic Review</mark>_ <mark>,</mark> _<mark>103</mark>_ <mark>(1), 438-471.</mark>\\n\\n- El-Laithy, Heba. (2001). “The Gender Dimensions of Poverty in Egypt.” ERF Working Paper 127.\\n\\n- ESCWA, 2021 “Targeted social protection in Arab countries before and during the Covid-19 crisis”. Economic and Social Commission for Western Asia.\\n\\n- Ferreira, Francisco H. G., Julian Messina, Jamele Rigolini, Luis-Felipe López-Calva, and Renos Vakis. (2012). _Economic Mobility and the Rise of the Latin American Middle Class._ Washington DC: World Bank.\\n\\n- Fuwa, Nobuhiko. (2000). “The poverty and heterogeneity among female-headed households revisited: the case of Panama.” _World Development_ , Vol. 28(8), pp. 1515-1542.\\n\\n- Gammage, Sarah. (1998). “The Gender Dimension of Household Poverty: Is Headship Still a Useful Concept?” International Center for Research on Women (Washington DC)\\n\\n- <mark>Ghomi, M. (2022). Who is afraid of sanctions? The macroeconomic and distributional effects of the sanctions against Iran.</mark> _<mark>Economics & Politics</mark>_ <mark>,</mark> _<mark>34</mark>_ <mark>(3), 395-428.</mark>\\n\\n- Grant, Monica J., and Jere R. Behrman. (2010). “Gender Gaps in Educational Attainment in Less Developed Countries.” _Population and Development Review_ , Vol. 36(1), pp. 71-89.\\n\\n- Horrell, Sara, and Pramila Krishnan. (2007). “Poverty and productivity in female-headed households in Zimbabwe.” _Journal of Development Studies_ , Vol. 43(8), pp. 1351-1380.\\n\\n- IMF. (2011). Islamic Republic of Mauritania: Poverty Reduction Strategy Paper, IMF Country Report No. 11/252, August, www.imf.org/external/pubs/ft/scr/2011/cr11252.pdf.\\n\\n- Klasen, Stephan, Tobias Lechtenfeld, and Felix Povel. (2015). “A feminization of vulnerability? Female headship, poverty, and vulnerability in Thailand and Vietnam.” _World Development_ , Vol. 71, pp. 36-53.\\n\\n- <mark>Lanjouw, P., & Ravallion, M. (1995). Poverty and household size. E</mark> _<mark>conomic Journal</mark>_ <mark>,</mark> _<mark>105</mark>_ <mark>(433), 1415-1434.</mark>\\n\\n- Liu, Chia, Albert Esteve, and Rocío Treviño. (2017). “Female-headed households and living conditions in Latin America.” _World Development_ , Vol. 90, pp. 311-328.\\n\\n30\", \"datasets\": [{\"mention_name\": {\"text\": \"household survey data\", \"confidence\": 0.9281083345413208, \"start\": 106, \"end\": 127, \"clean_text\": \"household survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 106, \"end\": 127, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 106, \"end\": 127, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 106, \"end\": 127}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 106, \"end\": 127, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9168452024459839, \"start\": 106, \"end\": 127}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9281083345413208, \"start\": 106, \"end\": 127}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5488042235374451, \"start\": 106, \"end\": 127}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5488042235374451, \"start\": 106, \"end\": 127}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [31]}}, {\"page\": 32, \"chunk\": 32, \"input_text\": \"- Mancini, Giulia, and Vecchi, Giovanni. (2022). _On the Construction of a Consumption Aggregate for Inequality and Poverty Analysis_ . Washington, D.C.: World Bank Group.\\n\\n- <mark>Marotta, D., Yemtsov, R., El-Laithy, H., Abou-Ali, H., & Al-Shawarby, S. (2011). Was growth in Egypt between 2005 and 2008 pro-poor? From static to dynamic poverty profile.</mark> _<mark>World Bank Policy Research Working Paper</mark>_ <mark>5589.</mark>\\n\\n- Medeiros, Marcelo, and Joana Costa. (2008). “Is there a feminization of poverty in Latin America?” _World Development_ , Vol. 36(1), pp. 115-127.\\n\\n- Milazzo, Annamaria, and Dominique van de Walle. (2017). “Women left behind? Poverty and headship in Africa.” _Demography_ , Vol 54(3), pp. 1119-1145.\\n\\n- <mark>Muller, J. S., Hiekel, N., & Liefbroer, A. C. (2020). The long-term costs of family trajectories: Women’s later-life employment and earnings across Europe.</mark> _<mark>Demography</mark>_ <mark>,</mark> _<mark>57</mark>_ <mark>(3), 1007-1034.</mark>\\n\\n- <mark>Munoz Boudet, A. M., Buitrago, P., Leroy De La Briere, B., Newhouse, D. L., Rubiano Matulevich, E. C., Scott, K., & Suarez-Becerra, P. (</mark> 2018). _Gender Differences in Poverty and Household Composition through the Life-Cycle: A Global Perspective_ . Policy Research Working Papers 8360. Washington, D.C: The World Bank.\\n\\n- Munoz Boudet, A. M., Bhatt, A., Azcona, G., Yoo, J., & Beegle, K. (2021). _A global view of poverty, gender, and household composition_ . Policy Research Working Papers 9553. Washington, D.C: The World Bank.\\n\\n- Nasri, Khaled (2020). Social Safety Nets in Tunisia: Do Benefits Reach the Poor and Vulnerable households at the Regional Level? GLO Discussion Paper, No. 440, Global Labor Organization (GLO), Essen\\n\\n- NAF 2020. “Assessing Gender Equity in National Aid Fund Programs”. Jordan National Aid Fund <u>https://naf.gov.jo/EN/List/Statistics_and_Data</u>\\n\\n- Nassar, Heba. (1997). “Feminization of Poverty.” The Egyptian Human Development Report; Research Papers Series, UNDP.\\n\\n- <mark>Newhouse, D., Becerra, P. S., & Evans, M. (2017). New global estimates of child poverty and their sensitivity to alternative equivalence scales.</mark> _<mark>Economics Letters</mark>_ <mark>,</mark> _<mark>157</mark>_ <mark>, 125-128.</mark>\\n\\n- OECD. (2018). A Broken Social Elevator? How to Promote Social Mobility. Paris: OECD Publishing. https://doi.org/10.1787/9789264301085-en\\n\\n- <mark>Palestinian Central Bureau of Statistics (PCBS) and World Bank. (2018). “Measuring poverty in West Bank and Gaza: Methodology review using PECS 2016.” Technical report.</mark>\\n\\n31\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS 2016\", \"confidence\": 0.37247034907341003, \"start\": 2556, \"end\": 2565, \"clean_text\": \"PECS 2016\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2556, \"end\": 2565, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2556, \"end\": 2565, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2556, \"end\": 2565}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2556, \"end\": 2565, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.5558839440345764, \"start\": 307, \"end\": 316}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.37247034907341003, \"start\": 307, \"end\": 316}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9736399054527283, \"start\": 307, \"end\": 316}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9736399054527283, \"start\": 307, \"end\": 316}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [32]}}, {\"page\": 33, \"chunk\": 33, \"input_text\": \"- <mark>Quisumbing, Agnes R., Lawrence Haddad, and Christine Peña. (2001). “Are women overrepresented among the poor? An analysis of poverty in 10 developing countries.”</mark> _<mark>Journal of Development Economics</mark>_ <mark>, Vol. 66(1), pp. 225-269.</mark>\\n\\n- <mark>Radean, M. (2023). ginteff: A generalized command for computing interaction effects.</mark> _<mark>Stata Journal</mark>_ <mark>,</mark> _<mark>23</mark>_ <mark>(2), 301-335.</mark>\\n\\n- <mark>Ravallion, M. (2016).</mark> _<mark>The economics of poverty: History, measurement, and policy</mark>_ <mark>. Oxford University Press.</mark>\\n\\n- <mark>Rogan, Michael. (2013). “Poverty and headship in post-apartheid South Africa, 1997–2006.”</mark> _<mark>Social Indicators Research</mark>_ <mark>, Vol. 113, pp. 491-511.</mark>\\n\\n- <mark>Rogers, Beatrice Lorge. (1995). “Alternative definitions of female headship in the Dominican Republic.”</mark> _<mark>World Development</mark>_ <mark>, Vol. 23(12), pp. 2033-203</mark>\\n\\n- <mark>Salvucci, Vincenzo, and Finn Tarp. (2021). “Poverty and vulnerability in Mozambique: An analysis of dynamics and correlates in light of the Covid‐19 crisis using synthetic panels.”</mark> _<mark>Review of Development Economics</mark>_ <mark>, Vol. 25(4), pp. 1895-1918.</mark>\\n\\n- United Nations Development Programme (UNDP). (2016). _Multidimensional Progress: Wellbeing beyond Income_ . New York: United Nations Development Programme.\\n\\n- Van de Walle, Dominique. (2013). “Lasting Welfare Effects of Widowhood in Mali.” _World Development_ , Vol. 51, pp. 1-19.\\n\\n- World Bank. (2011). _World development report 2012: Gender equality and development_ . World Bank: Washington, DC.\\n\\n- ---. (2022). “Jordan Emergency Cash Transfer COVID19 Response Project (P173974) Project Information Document.”\\n\\n32\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [33]}}, {\"page\": 34, \"chunk\": 34, \"input_text\": \"# **Table 1. Share of Female-Headed Households in Six Arab Countries (percentages)**\\n\\n|||**Egypt, **|**Arab Rep.**|||**Iraq**|**Jo**|**rdan**||**Mau **|**ritania**|||**West Ban**|**k and Gaza**|||**T**|**unisia**||\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n||2012|2015|2017|2020|2007|2012|2010|2013|2004|2008|2014|2019|2007|2009|2011|2017|2005|2010|2015|2021|\\n|**Te 1 Self-reorted FHH**|**17.80**|**17.86**|**18.43**|**17.59**|**11.30**|**9.63**|**13.88**|**13.24**|**18.92**|**31.30**|**30.19**|**36.58**|**9.12**|**10.00**|**11.07**|**10.05**|**17.01**|**14.85**|**16.24**|**18.52**|\\n|**yp . p**|_1,346_|_2,104_|_2,265_|_1,967_|_1,906_|_2,531_|_402_|_669_|_1,860_|_4,273_|_3,033_|_3,654_|_108_|_388_|_505_|_397_|_2,128_|_1,704_|_4,088_|_3,153_|\\n|Offiil FHH|13.98|14.96|15.92|14.03|10.43|8.22|11.29|11.05|16.89|18.90|17.23|16.98|8.26|8.79|9.93|8.04|13.73|11.95|14.12|16.62|\\n|ca|_1,058_|_1,741_|_1,956_|_1,599_|_1,721_|_2,152_|_335_|_592_|_1,674_|_2,688_|_1,839_|_1,834_|_97_|_341_|_447_|_308_|_1,671_|_1,347_|_3,462_|_2,788_|\\n|N id|0.48|0.37|0.48|0.59|0.46|0.20|0.89|0.89|0.44|0.45|0.50|0.60|0.59|1.07|1.09|0|1.01|0.56|1.29|1.51|\\n|ever marre|_35_|_43_|_57_|_67_|_78_|_101_|_17_|_43_|_44_|_67_|_48_|_72_|_7_|_45_|_56_|_0_|_113_|_65_|_282_|_261_|\\n|Did/ td|0.95|1.33|1.55|1.80|0.78|0.62|0.51|0.66|6.15|6.98|6.22|6.64|0.99|0.65|1.14|0.82|1.28|1.18|1.53|1.90|\\n|vorce separae|_74_|_153_|_194_|_210_|_120_|_175_|_19_|_33_|_602_|_974_|_660_|_690_|_11_|_25_|_43_|_29_|_156_|_131_|_360_|_291_|\\n|Wid l|12.56|13.26|13.89|11.64|9.19|7.40|9.90|9.50|10.29|11.47|10.52|9.74|6.68|7.07|7.70|7.22|11.44|10.21|11.29|13.21|\\n|ow ony|_949_|_1,545_|_1,705_|_1,322_|_1,523_|_1,876_|_299_|_516_|_1,028_|_1,647_|_1,131_|_1,072_|_79_|_271_|_348_|_279_|_1,402_|_1,151_|_2,820_|_2,236_|\\n|Mid FHH|3.80|2.89|2.52|3.53|0.87|1.41|2.58|2.19|1.92|12.20|12.96|18.45|0.87|1.22|1.14|2.01|3.28|2.90|2.12|1.91|\\n|arre|_287_|_363_|_309_|_365_|_185_|_379_|_67_|_77_|_176_|_1,558_|_1,194_|_1,714_|_11_|_47_|_58_|_89_|_457_|_357_|_624_|_365_|\\n|**T 2 Sh f fl dlt>05**|**22.96**|**22.29**|**22.46**|**20.87**|**24.02**|**28.92**|**25.80**|**22.80**|**31.50**|**40.40**|**39.98**|**43.96**|**21.52**|**19.53**|**20.97**|**24.28**|**29.20**|**28.98**|**26.17**|**27.21**|\\n|**ype . are o emae aus.**|1,712|2,659|2,797|2,352|4,547|6,525|728|1,105|_3,075_|_5,606_|_3,897_|_4,393_|265|751|919|955|3,710|3,363|_6,907_|_4,740_|\\n|Share of employed females> share of|**5.42**|**6.22**|**6.12**|**6.02**|**4.81**|**3.81**|**6.02**|**7.56**|**11.40**|**12.90**|**14.30**|**14.68**|**7.65**|**6.72**|**6.54**|**4.59**|**10.94**|**9.01**|**10.17**|**11.04**|\\n|employed males|_406_|_733_|_757_|_660_|_916_|_1,070_|_178_|_291_|_1,056_|_1,888_|_1,422_|_1,454_|_88_|_269_|_300_|_176_|_1,313_|_987_|_2,382_|_1,854_|\\n|**T 3 Pttil FHH**|**16.68**|**16.49**|**19.37**|**18.86**|**5.73**|**3.59**|**13.25**|**12.78**|**17.01**|**27.09**|**25.02**|**27.88**|**11.09**|**10.35**|**10.55**|**9.88**|**18.30**|**17.54**|**18.13**|**26.75**|\\n|**ype . oena**|_1,244_|_1,935_|_2,344_|_2,087_|_1,007_|_1,621_|_359_|_576_|_1,643_|_3,694_|_2,345_|_2,824_|_127_|_398_|_482_|_397_|_2,280_|_1,974_|_4,609_|_4,618_|\\n||3.06|3.17|3.49|3.75|1.65|0.86|2.52|3.52|7.07|8.07|9.68|9.12|3.04|2.34|2.84|1.82|5.07|4.11|4.27|5.67|\\n|Core FHH|_229_|_364_|_419_|_407_|_283_|_388_|_77_|_129_|_635_|_1,160_|_939_|_912_|_34_|_97_|_130_|_74_|_620_|_457_|_1,022_|_947_|\\n||11.77|11.70|14.05|13.10|4.00|2.61|12.02|10.90|14.85|24.30|22.63|24.67|9.24|8.76|8.52|N.A.|15.56|15.55|15.22|N.A.|\\n|Asset FHH|_881_|_1,459_|_1,765_|_1,489_|_783_|_1,258_|_321_|_510_|_1,417_|_3,287_|_2,095_|_2,523_|_106_|_338_|_402_|_N.A._|_1,967_|_1,769_|_4,037_|_N.A._|\\n|**Type 4. Most educated adult member**|**12.92**|**13.88**|**14.30**|**13.98**|**8.54**|**5.87**|**18.49**|**18.48**|**8.76**|**16.47**|**13.83**|**24.67**|**16.87**|**15.25**|**14.82**|**11.52**|**15.42**|**17.75**|**17.84**|**20.11**|\\n|**is female & no employed males**|_951_|_1,610_|_1,749_|_1,554_|_1,485_|_1,990_|_499_|_932_|_969_|_2,384_|_1,378_|_2,490_|_201_|_595_|_643_|_419_|_1,936_|_2,024_|_4,479_|_3,520_|\\n\\nNote: The main definitions of female-headed households are in bold, and the variant definitions are in regular font. The numbers in bold refer to the percent of the cross-sectional sample for each period. The numbers in _italics_ refer to the sample size of each group. Type 1 _self-reported FHHs_ are obtained from self-reporting information in the survey questionnaires. Type 2 _majority-female-adult FHHs_ are defined as households where the proportion of females among (working age) adults exceeds 0.5. Type 3 _potential FHHs_ are those households where there are no working-age males present. Type 4 _most-educated-female-adult FHHs_ consist of households with no employed males, whose most educated adult member is female. Under Type 1 FHHs, the different sub-types are defined as in the survey questionnaires. Under Type 3 FHHs, sub-type 3 _core FHH_ encompasses only the potential FHHs with employed females, and sub-type 3 _asset FHH_ encompasses only households with females who have ownership rights over the dwellings they reside in.\\n\\n33\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [34]}}, {\"page\": 35, \"chunk\": 35, \"input_text\": \"# **Table 2. Probabilities of Being Poor**\\n\\n|||**Specific**|**ation 1**|||**Specificatio**|**n 2**||\\n|---|---|---|---|---|---|---|---|---|\\n|||**FHH Type 2**||**FHH Type 4**||||**FHH Type 4**|\\n||**FHH Type 1**<br>**Self-reported**|<br>**Majority-female-**<br>**adult**|**FHH Type 3**<br>**Potential**|<br>**Most-educated-**<br>**female-adult**|**FHH Type 1**<br>**Self-reported**|**FHH Type 2**<br>**Majority-female-adult**|**FHH Type 3**<br>**Potential**|<br> <br>**Most-educated-**<br>**female-adult**|\\n|Self-reported FHH|-0.007***<br>(0.00)||||-0.044***<br>(0.00)||||\\n|Self-reported FHH # Number of children|0.008***<br>(0.00)||||0.006***<br>(0.00)||||\\n|Share of female adults>0.5||0.052***<br>(0.00)||||0.032***<br>(0.00)|||\\n|Share of female adults>0.5# Number of||<br>-0.007***||||<br>-0.014***|||\\n|children||(0.00)||||(0.00)|||\\n||||-0.027***|||-0.017***|||\\n|Potential FHH|||(0.00)|||(0.00)|||\\n||||0.014***|||0.011***|||\\n|Potential FHH# Number of children|||(0.00)|||(0.00)|||\\n|Educated females||||0.002<br>(0.00)||||-0.006**<br>(0.00)|\\n|Educated females# Number of children||||<br>0.013***<br>(0.00)||||<br>0.013***<br>(0.00)|\\n|Household head`s|||||||||\\n|characteristics|N|N|N|N|Y|Y|Y|Y|\\n|Household characteristics|N|N|N|N|Y|Y|Y|Y|\\n|Country FE|Y|Y|Y|Y|Y|Y|Y|Y|\\n|<br>Survey year FE|Y|Y|Y|Y|Y|Y|Y|Y|\\n|N|214931|214931|214931|214931|211069|211069|211069|211069|\\n\\nNote: ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Robust standard errors are in parentheses. The full regression results using the linear probability model are provided in Appendix A, Table A.10.\\n\\n34\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [35]}}, {\"page\": 36, \"chunk\": 36, \"input_text\": \"<!-- Start of picture text --> Female-headed household Typology<br>“De jure” FHH “De facto” FHH<br>Self-reported approach Demographic approach Socio-economic approach<br>Mei Type 3: Potential FHH | Type4: Mosteducated member<br>Type 1: Self-reported Type 2: Major share of (no working-age male) wanetip i the household is female,<br>an females in HH Fuwa (2000) , ~ where no working-age<br>DeGraff and Bilsborrow(1993) Milazzo and van de Walle (2017)... employed male ispresent<br>Quisumbing et al. (2001) a ? \\\\<br>Core FAH<br>mart seus (or Married FHH employedNumber/ femalesShai scot FHH (femae' is engoged<br>reason for spouse's Liu et al. (2017) (earners) Deere et ol generating<br>absence) Brown and Van de Bibi and Chatti (2012) activities)<br>Buvinic and Gupta Walle (2021) M Goto eral Fuwa {2000)<br>wtasenat (2025) (ois) Rogan (2013)<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [36]}}, {\"page\": 37, \"chunk\": 37, \"input_text\": \"# **Figure 2. Cross-sectional Headcount Poverty Rate (percentage), by Household Type, Regional Poverty Lines**\\n\\n<!-- Start of picture text --> Egypt Iraq<br>2012 2015 2017 2020 2007 2012<br>Jordan Mauritania<br>2010 2013 2004 2008 2014 2019<br>Palestine Tunisia<br>2007 2009 2011 2017 2005 2010 2015 2021<br>Whole population Self-reported FHH Share of females>0.5<br>Potential FHH Educated females<br>35 30<br>30<br>25<br>25<br>20 20<br>15<br>Poverty rate (%) 15 Poverty rate (%)<br>30 60<br>25<br>50<br>20<br>40<br>15<br>10 30<br>Poverty rate (%)5 20 Poverty rate (%)<br>100 30<br>80 25<br>60 20<br>40 15<br>10<br>Poverty rate (%)20 Poverty rate (%)<br><!-- End of picture text --> Note: Population sampling weights are applied.\\n\\n36\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [37]}}, {\"page\": 38, \"chunk\": 38, \"input_text\": \"**Figure 3. FHH–non-FHH Differences in Headcount Poverty Rate (percentage points)**\\n\\nPanel A: Self-reported FHH Panel B: Share of females>0.5 <!-- Start of picture text --> 2012201520172020 ****** *** *** *** *** *** 2012201520172020<br>20072012 *** * ** 20072012<br>20102013 * ** *** 20102013<br>2004200820142019 *** *** *** ** ********* 2004200820142019<br>2007 2007<br>200920172011 *** ** ** *** 200920112017<br>2005201020152021 ******** *** * * * ** ****** 2005201020152021<br>-5 0 5 10<br>-15 -10 -5 0 5<br>Panel C: Potential FHH Panel D: Educated females<br>2012201520172020 *** *** *** ** * ****** *** ** * 2012201520172020<br>20072012 * * ** 20072012<br>20102013 ** * * * * ** *** 20102013<br>2004200820142019 **** ** * ********* ** * 2004200820142019<br>2007200920172011 *** ****** *** ** *** *** 2007200920112017<br>2005201020152021 ** * ** * * * *** *** 2005201020152021<br>-40 -30 -20 -10 0 -20 -10 0 10<br>Egypt Egypt<br>Iraq Iraq<br>Jordan Jordan<br>Mauritania Mauritania<br>Palestine Palestine<br>Tunisia Tunisia<br>Egypt Egypt<br>Iraq Iraq<br>Jordan Jordan<br>Mauritania Mauritania<br>Palestine Palestine<br>Tunisia Tunisia<br><!-- End of picture text -->\\n\\n**Note:** Headcount poverty rates are estimated using per capita household expenditures. Stars indicate significantly higher headcount poverty ratio between FHHs and non-FHHs in each category. ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Negative difference means FHHs are less likely to be poor.\\n\\n37\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [38]}}, {\"page\": 39, \"chunk\": 39, \"input_text\": \"# **Figure 4. Differences in Headcount Poverty Rate for FHHs vs. non-FHHs by Number of Children under 14 (percentage points)**\\n\\n<!-- Start of picture text --> Panel A: Self-reported FHH Panel B: Share of females>0.5<br>Egypt Egypt<br>Iraq Iraq<br>Jordan Jordan<br>Mauritania Mauritania<br>Palestine Palestine<br>Tunisia Tunisia<br>-10 0 10 20 30 40 -10 0 10 20 30<br>Panel C: Potential FHH Panel D: Educated females<br>Egypt Egypt<br>Iraq Iraq<br>Jordan Jordan<br>Mauritania Mauritania<br>Palestine Palestine<br>Tunisia Tunisia<br>-20 0 20 40 -20 0 20 40<br>0 1 2 3 4 or more<br><!-- End of picture text -->\\n\\n**Note:** Authors’ calculation based on pooled cross section. Headcount poverty rates are estimated using per capita household expenditures. The number of children are shown for 0, 1, 2, 3, and 4 or more children. The headcount poverty rates are shown on the x-axis, with positive numbers indicating more poverty for FHHs. The error bars are the 95% CIs.\\n\\n38\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [39]}}, {\"page\": 40, \"chunk\": 40, \"input_text\": \"# **Figure 5. Probabilities of Female-Headed Households Escaping Poverty in Second Year Conditional on Being Poor in First Year (percentage)**\\n\\n<!-- Start of picture text --> Egypt 2017-2020 Iraq 2007-2012<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>Jordan 2010-2013 Mauritania 2014-2019<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>Palestine 2011-2017 Tunisia 2015-2021<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>FHH wo/children FHH w/children Non-FHH<br>40 60<br>55<br>35<br>50<br>30<br>45<br>Percentage (%) Percentage (%)<br>25 40<br>60 50<br>55 45<br>50<br>40<br>Percentage (%) 45 Percentage (%)<br>35<br>30<br>45<br>40 25<br>35<br>20<br>30<br>Percentage (%) Percentage (%)<br>25 15<br><!-- End of picture text --> Note: Estimates are obtained with synthetic panel data and weighted with population weights where the second survey round is used as the base year. The figure shows the percentage of the population that moves out of poverty in the second year. FHH characteristics are measured in second period. Dashed red lines represent the national average for each period. Household heads' ages are restricted to 25-55 for the first survey round and adjusted accordingly for the second survey round. Standard errors are estimated with 1,000 bootstraps.\\n\\n39\", \"datasets\": [{\"mention_name\": {\"text\": \"synthetic panel data\", \"confidence\": 0.8500471711158752, \"start\": 948, \"end\": 968, \"clean_text\": \"synthetic panel data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 948, \"end\": 968, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 948, \"end\": 968, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 948, \"end\": 968}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 948, \"end\": 968, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9220925569534302, \"start\": 294, \"end\": 314}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8500471711158752, \"start\": 294, \"end\": 314}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5956490635871887, \"start\": 294, \"end\": 314}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5956490635871887, \"start\": 294, \"end\": 314}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [40]}}, {\"page\": 41, \"chunk\": 41, \"input_text\": \"# **Figure 6. Probabilities of Female-Headed Households Falling in Poverty in Second Year Conditional on Being Non-poor in First Year (percentage)**\\n\\n<!-- Start of picture text --> Egypt 2017-2020 Iraq 2007-2012<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>Jordan 2010-2013 Mauritania 2014-2019<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>Palestine 2011-2017 Tunisia 2015-2021<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>FHH wo/children FHH w/children Non-FHH<br>14<br>16<br>14 12<br>12 10<br>10 8<br>Percentage (%) 8 6 Percentage (%)<br>14 16<br>12 14<br>10 12<br>8 10<br>Percentage (%) 6 8 Percentage (%)<br>55 10<br>50 9<br>45 8<br>40 7<br>35 6<br>Percentage (%) 30 5 Percentage (%)<br><!-- End of picture text --> Note: Estimates are obtained with synthetic panel data and weighted with population weights where the second survey round is used as the base year. The figure shows the percentage of the population that enters poverty in the second year. FHH characteristics are measured in second period. Dashed red lines represent the national average for each period. Household heads' ages are restricted to 25-55 for the first survey round and adjusted accordingly for the second survey round. Standard errors are estimated with 1,000 bootstraps.\\n\\n40\", \"datasets\": [{\"mention_name\": {\"text\": \"synthetic panel data\", \"confidence\": 0.8394618630409241, \"start\": 946, \"end\": 966, \"clean_text\": \"synthetic panel data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 946, \"end\": 966, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 946, \"end\": 966, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 946, \"end\": 966}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 946, \"end\": 966, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9220925569534302, \"start\": 289, \"end\": 309}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8394618630409241, \"start\": 289, \"end\": 309}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5956490635871887, \"start\": 289, \"end\": 309}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5956490635871887, \"start\": 289, \"end\": 309}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [41]}}, {\"page\": 42, \"chunk\": 42, \"input_text\": \"# **Appendix A: Additional Tables and Figures**\\n\\n# **Table A.1. Overview of the key studies**\\n\\n||**Studies**|**Country**|**FHH**<br>**definition**|**Reference**<br>**group**|**Conclusions**|\\n|---|---|---|---|---|---|\\n||||Self-<br>reported<br>FHH|Self-<br>reported<br>MHH<br>|FHHs have lower income, land ownership, and average level of education than<br>MHHs and are less likely to be employed (with fewer hours if employed). Children<br>of FHHs are significantly less likely to be enrolled in school than children of MHHs.|\\n|1|DeGraff and<br>Bilsborrow (1993)|Ecuador|Self-<br>reported<br>widowed<br>FHH|Self-<br>reported<br>MHH, other<br>self-reported<br>FHH|Widowed or divorced FHHs have higher income and amount of land owned than<br>MHHs but lower children's school enrollment and are less likely to be in school<br>than are children of MHHs and children of other FHHs.|\\n||||Self-<br>reported<br>married<br>FHH|Self-<br>reported<br>MHH, other<br>self-reported<br>FHH|Married FHHs have lower income and amount of land owned than MHHs and<br>enrollment rate similar to MHH, but children of married FHHs are significantly less<br>likely to be enrolled in school than are children of MHHs but more likely than<br>widowed FHHs.|\\n|2|Buvinic and Gupta<br>(1997)|65 studies<br>on<br>developing<br>countries|Self-<br>reported<br>FHH, de<br>facto, de<br>jure FHH|Self-<br>reported<br>MHH|Thirty-eight studies - FHHs are poorer than MHHs when poverty is measured by<br>(total/per capita/per equivalent) household income and consumption expenditures,<br>access to services, and ownership of land and assets. Fifteen studies - certain types<br>of FHHs are more vulnerable to poverty than others. Eight studies – poverty in<br>FHHs is not higher than in MHH.|\\n||||Self-<br>reported<br>FHH|Self-<br>reported<br>MHH|FHHs are similar to MHHs when poverty is measured by per capita expenditure.<br>The difference between FHHs versus non-FHHs does not change if using different<br>poverty indicators.|\\n||||Self-<br>reported de<br>jure, de<br>facto FHH|Non-FHH|Widows/divorced/separated FHHs have significantly higher headcount poverty in<br>indigenous areas when poverty is measured by per capita expenditure. FHHs have<br>lower education than non-FHHs|\\n|3|Fuwa (2000)|Panama|Self-<br>reported<br>married<br>FHH, FHH<br>with<br>unmarried<br>partners|Non-FHH|FHHs with unmarried partners have higher headcount poverty ratios in urban and<br>indigenous areas when poverty is measured by per capita expenditure. The result is<br>robust to applying an equivalence scale using alternative poverty measures and<br>poverty lines.|\\n||||<br>Potential<br>FHH|Non-FHH|FHHs are not poorer than non-FHHs when poverty is measured by per capita<br>expenditure.|\\n||||“Working”<br>FHHs|Non-FHH|FHHs are similar to non-FHHs when poverty is measured by per capita expenditure.<br>FHHs have higher education endowments than non-FHHs, except in indigenous<br>areas.|\\n\\n41\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [42]}}, {\"page\": 43, \"chunk\": 43, \"input_text\": \"||||core FHH|Non-FHH|FHHs are less poor than non-FHHs when poverty is measured by per capita<br>expenditure.|\\n|---|---|---|---|---|---|\\n|4|Quisumbing_et al._<br>(2001)|10<br>developing<br>countries|Self-<br>reported<br>FHH,<br>females|Self-<br>reported<br>MHH, males|FHHs and individual females contribute disproportionately to overall poverty in 25-<br>50% of the dataset when headcount poverty is measured by (total/per capita/per<br>equivalent) household income and consumption expenditures and are insensitive to<br>the poverty line. FHHs and individual females are similar to MHHs or males when<br>using stochastic dominance criteria, but they are constantly worse off in Ghana and<br>Bangladesh.|\\n|5|Horrell and Krishnan<br>(2007)|Zimbabwe|Self-<br>reported<br>widowed<br>FHH|Self-<br>reported|The income per capita/ adult equivalent is lower in widowed FHHs than in the<br>MHHs.|\\n||||Self-<br>reported de-<br>facto FHH|MHH|The income per capita/ adult equivalent is higher in the de facto FHHs than in the<br>MHHs.|\\n|6|Medeiros and Costa<br>(2007)|8 Latin<br>American<br>ti|FHH,<br>females|MHH, males|Poverty is higher among FHHs, but there is no clear evidence of a recent and<br>widespread feminization of poverty in Latin America. Differences in poverty<br>among FHHs and MHHs increased in Argentina and Mexico, showing specific<br>types of feminization of poverty. The results are robust to different values of<br>poverty lines, the use of equivalence scales, and the distribution of household<br>income.|\\n|||counres|Self-<br>reported<br>FHH w/o<br>children|Couple HH<br>w/o children|The insignificant increase in poverty indices when comparing FHHs without<br>children to couple-headed HH without children in Bolivia. The rise in poverty<br>indices is significant at 5% when comparing FHHs with children to MHHs with<br>children in Costa Rica.|\\n|7|Deere_et al._(2012)|Latin<br>American<br>countries|HHs where<br>women have<br>ownership<br>rights|Self-<br>reported<br>FHHs who<br>have<br>ownership<br>rights|The gender of the household head is a poor substitute for a gendered analysis of<br>asset ownership within and among households since an analysis based on headship<br>tends to underestimate women’s ownership of assets.|\\n|8|Van de Walle (2013)|Mali|Self-<br>reported<br>widowed<br>FHH<br>Self-<br>reported<br>widowed<br>MHH|Self-<br>reported<br>married<br>FHH<br>Self-<br>reported<br>married<br>MHH|Widowed FHHs have significantly lower consumption per capita than married<br>FHHs, while MHHs do not have any significant differences in per capita<br>consumption between widowed MHHs and married MHHs.|\\n||||Self-<br>reported<br>widowed<br>FHH rural|Self-<br>reported<br>widowed<br>FHH urban|Widowed FHH living in rural areas have lower per capita consumption than all<br>other households living in rural areas. The gap between widowed FHHs and other<br>HHs is lower for HHs residing in urban areas.|\\n\\n42\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [43]}}, {\"page\": 44, \"chunk\": 44, \"input_text\": \"|||Self-<br>reported<br>widowed<br>FHH rural|Self-<br>reported<br>non-<br>widowed<br>FHH rural|Per capita consumption of widowed FHHs is around 12% lower than that of all rural<br>households. The results are robust to using an equivalence scale in measuring<br>consumption.|\\n|---|---|---|---|---|\\n|||Self-<br>reported<br>widowed<br>FHH urban|Self-<br>reported<br>non-<br>widowed<br>FHH urban|Per capita consumption of widowed FHHs is around 6% lower than that of all other<br>urban households. The results are robust to using an equivalence scale in measuring<br>consumption.|\\n|9<br>Rogan (2013)|South<br>Afi|Self-<br>reported de<br>jure FHH,<br>de facto<br>FHH, co-<br>resident<br>FHH|Self-<br>reported<br>MHH|Poverty rates are higher in FHHs than in MHHs, irrespective of how headship is<br>defined.|\\n||rca|Self-<br>reported de<br>jure FHH,<br>de facto<br>FHH|co-resident<br>FHH|Co-resident FHHs are less poor than other types of FHHs.|\\n|||core FHH|non-FHH|FHH has the lowest risk of poverty|\\n|||Self-<br>reported<br>FHH|Self-<br>reported<br>MHH|No significant differences between FHHs and MHHs were found regarding<br>consumption, the probability of shock exposure, or vulnerability to poverty in<br>Thailand or Viet Nam.|\\n|||Self-<br>reported de-<br>jure FHH|Self-<br>reported<br>MHH|De jure FHHs have lower consumption than MHHs in Viet Nam. There are no<br>significant differences between de-jure FHHs and MHHs regarding the probability<br>of shock exposure or vulnerability to poverty in Thailand or Viet Nam.|\\n|10<br>Klasen_et al._(2015)|Thailand,<br>Viet Nam|Self-<br>reported de-<br>facto FHH|Self-<br>reported<br>MHH|De facto FHHs have higher consumption than MHHs in Thailand. There are no<br>significant differences between de-facto FHHs and MHHs regarding the probability<br>of shock exposure or vulnerability to poverty in Thailand or Viet Nam.|\\n|||Self-<br>reported<br>single,<br>widowed<br>FHH|Self-<br>reported<br>MHH|FHHs with an absent spouse have higher consumption levels than MHHs in<br>Thailand. Single FHH has a lower consumption level than Viet Nam. There are no<br>significant differences between FHHs and MHHs regarding the probability of any<br>shock exposure in Thailand or Viet Nam. Single FHHs are less vulnerable to<br>poverty in Thailand but more vulnerable to poverty in Viet Nam.|\\n|11<br>Liu_et al._(2017)|14 Latin<br>American<br>countries|Self-<br>reported<br>married w/o<br>spouse,<br>single,<br>separated,|Self-<br>reported<br>married<br>FHH with<br>spouse|In eight of the 14 countries, FHHs are more likely to live in poor conditions.<br>However, MHHs are in more impoverished conditions than FHHs when married<br>status, urban or rural setting, ownership, and the presence of children are controlled<br>in the regression. Generally, married FHHs with the spouse present are better off<br>than any other category. The worst living conditions are associated with single,<br>separated, divorced, or widowed FHHs.|\\n\\n43\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [44]}}, {\"page\": 45, \"chunk\": 45, \"input_text\": \"||||widowed<br>FHH|||\\n|---|---|---|---|---|---|\\n|||20<br>countries in|Self-<br>reported<br>FHH|Self-<br>reported<br>MHH|While the share of FHHs in the population is growing during 1990-2012, poverty<br>has been falling faster among FHHs. FHHs contributed more to the overall decline<br>in poverty despite their smaller overall population share.|\\n|12|Milazzo and van de<br>Walle (2017)|<br>Sub-<br>Saharan<br>Africa|Self-<br>reported<br>FHH w/o a<br>resident<br>adult male|Self-<br>reported<br>MHH|The poverty trends of the various types of FHHs followed different paths across<br>countries and periods, with no one type consistently outperforming the others.|\\n|13|Alazzawi (2018)|Egypt,|Self-<br>reported<br>urban FHH<br>with<br>children|Self-<br>reported<br>urban MHH<br>with<br>children|FHHs have a higher predicted poverty rate than MHHs in urban areas. The factors<br>contributing to the poverty differential between FHH and MHH households are<br>education, employment status, occupation, sector, and region of residence.|\\n|||Arab Rep.|Self-<br>reported<br>rural FHH<br>with<br>children|Self-<br>reported<br>rural MHH<br>with<br>children|FHHs have a higher predicted poverty rate than MHHs in rural areas. Education,<br>employment status, occupation, number of rooms per capita, and region of<br>residence are factors that contribute to the poverty differential between FHHs and<br>MHHs|\\n|14|Munoz Boudet_et al._<br>(2018)|71<br>developing|couple/singl<br>e females<br>w/o children|other HH|Adult couple households with children, children, and other adults (extended<br>family) are the most frequent among poor households. Poor and non‐poor women<br>concentrate in the adult couple household with children. One adult female<br>household with children is more prevalent among the poor in Latin America, the<br>Caribbean, and Sub‐Saharan Africa.|\\n|||countries|Male/female<br>earner with<br>and w/o<br>children||Poor women live in households with children and with children and earner<br>dependents, where the earner is a single male or a head couple. Single female-<br>earner households comprise the largest percentage of poor households in Latin<br>America, the Caribbean, and Sub‐Saharan Africa.|\\n||B d V d|43 Afi|Self-<br>reported<br>FHH|Self-|FHHs have lower poverty rates than MHHs when using per capita welfare<br>measures. FHHs are significantly worse than MHH when poverty is measured<br>using consumption adjusted for economies of scale.|\\n|15|rown an an e<br>Walle (2021)|rcan<br>countries|Self-<br>reported<br>married<br>FHH|reported<br>MHH|MHHs are poorer than married FHHs|\\n\\n44\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [45]}}, {\"page\": 46, \"chunk\": 46, \"input_text\": \"# **<u>Table A.2. Poverty Line, Arab Republic of Egypt by Region, in LCU, Per Capita Annual Consumption in Survey Year Prices</u>**\\n\\n|||**Food**<br>|**poverty line**<br>|||**Po**<br>|**verty line**<br>||\\n|---|---|---|---|---|---|---|---|---|\\n|**Region**|**2012/2013**|**2015**|**2017/2018**|**2019/2020**|**2012/2013**|**2015**|**2017/2018**|**2019/2020**|\\n|**Urban governorates**|2748|4318|6065.3|7071|4320|6141|9280.1|11285|\\n|**Urban lower Egypt**|2484|3835|5667.6|6304|3840|5631|8536.9|9755|\\n|**Rural lower Egypt**|2568|3854|5901.7|6570|3852|5675|8673|10108|\\n|**Urban upper Egypt**|2568|3968|5752.1|6553|3972|5823|8728.5|10225|\\n|**Rural upper Egypt**|2496|3760|5896.5|6484|3756|5694|8865.6|10068|\\n|**Urban frontier**|2736|3990|5924.3|6696|3996|6247|8568.7|10409|\\n|**Rural frontier**|2688|3979|6304.7|7074|3984|5788|8979.3|10788|\\n|**Total**|2568|3921|5889.6|6604|3924|6141|8827|10279|\\n\\nSource: Compiled from various CAPMAS Poverty assessment updates.\\n\\n45\", \"datasets\": [{\"mention_name\": {\"text\": \"CAPMAS Poverty assessment updates\", \"confidence\": 0.7562760710716248, \"start\": 918, \"end\": 951, \"clean_text\": \"CAPMAS Poverty assessment updates\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 918, \"end\": 951, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 918, \"end\": 951, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 918, \"end\": 951}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 918, \"end\": 951, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.30848395824432373, \"start\": 149, \"end\": 182}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7562760710716248, \"start\": 149, \"end\": 182}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7038023471832275, \"start\": 149, \"end\": 182}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7038023471832275, \"start\": 149, \"end\": 182}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [46]}}, {\"page\": 47, \"chunk\": 47, \"input_text\": \"# **<u>Table A.3. Poverty Lines, Jordan, in LCU, Per Capita Annual Consumption in Survey Year Prices</u>**\\n\\n||**Food poverty line**|**Poverty li**|**ne**|\\n|---|---|---|---|\\n||2010<br>2013|2010|2013|\\n|**Jordan**|336<br>383|814|929|\\n\\nSource: Jordan Department of Statistics: DOS https://jorinfo.dos.gov.jo/Databank/pxweb/en/Poverty/Poverty__Poverty-Indicators/ 46\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [47]}}, {\"page\": 48, \"chunk\": 48, \"input_text\": \"# **<u>Table A.4. Poverty Lines, Iraq, in LCU, Per Capita Annual Consumption in Survey Year Prices</u>**\\n\\n||2007|2012|\\n|---|---|---|\\n|**Kurdistan**|1212|1709|\\n|**Baghdad**|987|1391|\\n|**Rest of Iraq**|865|1220|\\n|**Total**|1073|1266|\\n\\nSource: World Bank “Poverty Estimates and Trends in Iraq” <u>https://microdata.worldbank.org/index.php/catalog/2334/download/34771</u> 47\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [48]}}, {\"page\": 49, \"chunk\": 49, \"input_text\": \"# **<u>Table A.5. Poverty Lines, the West Bank and Gaza, in LCU, Per Capita Annual Consumption in Survey Year Prices</u>**\\n\\n||**Food**|**poverty line**||**Po**|**verty line**||\\n|---|---|---|---|---|---|---|\\n||2009|2011|2017|2009|2011|2017|\\n|**Gaza**|567|570|567|712|714|710|\\n|**West Bank**|609|632|710|765|792|889|\\n|**Total**|603|620|668|757|776|836|\\n\\nSource: Compiled from various PCBS poverty reports. Spatial deflator provided by PCBS was used to calculate regional poverty lines for Gaza and the West Bank relative to the national poverty line available from PCBS publications.\\n\\n48\", \"datasets\": [{\"mention_name\": {\"text\": \"PCBS poverty reports\", \"confidence\": 0.5995646119117737, \"start\": 382, \"end\": 402, \"clean_text\": \"PCBS poverty reports\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.5501880049705505, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"PCBS\", \"confidence\": 0.5762877464294434, \"start\": 382, \"end\": 386, \"clean_text\": \"PCBS\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 382, \"end\": 402}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 382, \"end\": 402, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"report\", \"confidence\": 0.9972706437110901, \"start\": 66, \"end\": 86}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5995646119117737, \"start\": 66, \"end\": 86}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.737628698348999, \"start\": 66, \"end\": 86}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.737628698348999, \"start\": 66, \"end\": 86}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [49]}}, {\"page\": 50, \"chunk\": 50, \"input_text\": \"|**Table A.6. Poverty Lines, Tunisi**|**a, in LCU, Per C**<br>**Food**<br>|**apita Annua**<br>**poverty line**<br>|**l Consumptio**<br>|**n in Survey Y**<br>|**ear Prices**<br>**P**<br>|**overty line**<br>|||\\n|---|---|---|---|---|---|---|---|---|\\n||2005|2010|2015|2021|2005|2010|2015|2021|\\n|**Cities (metropolitan)**|615|757|1085|1346.526|902|1038|1878|2682.997|\\n|**Small & medium towns (urban)**|596|733|1050.154|--|818|941|1702.871|--|\\n|**Noncommunal (rural)**|466|571|951.668|1529.233|581|669|1500.530|2223.527|\\n\\nSource: World Bank (2016). “Tunisia Poverty Assessment 2015”. Table A1.3.\\n\\n49\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [50]}}, {\"page\": 51, \"chunk\": 51, \"input_text\": \"# **<u>Table A.7. Poverty Lines, Mauritania, in LCU, Per Capita Annual Consumption in Survey Year Prices</u>**\\n\\n||**Extreme poverty line**|||**Pove**|**rty line**||\\n|---|---|---|---|---|---|---|\\n||**2004**|**2008**|**2004**|**2008**|**2014**|**2019**|\\n|**Total**|70400|96000|94650|129000|169445|191000|\\n\\nSource: IMF (2011) Table 1.1.\\n\\n50\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [51]}}, {\"page\": 52, \"chunk\": 52, \"input_text\": \"# **<u>Table A.8. Correlation between Main Types of Female-Headed Households</u>**\\n\\n||**FHH Type 1**<br>**Self-reported**|**FHH Type 2**<br>**Majority-female-adult**|**FHH Type 3**<br>**Potential**|**FHH Type 4**<br>**Most-educated-female-adult**|\\n|---|---|---|---|---|\\n|**FHH Type 1**|1.000|0.265***<br>(0.000)|0.415***<br>(0.000)|0.298***<br>(0.000)|\\n|**FHH Type 2**||1.000|0.319***<br>(0.000)|0.304***<br>(0.000)|\\n|**FHH Type 3**|||1.000|0.510***<br>(0.000)|\\n|**FHH Type 4**||||1.000|\\n\\nNote: ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels.\\n\\n51\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [52]}}, {\"page\": 53, \"chunk\": 53, \"input_text\": \"**-** **<u>Table A.9. Headcount Poverty Rate of Female and Non female Headed Households by Headship Definition (percent)</u>**\\n\\n||Egypt,|Arab Rep.|||||Mauritani|a (2004-2019)|West Ban|k and Gaza|||\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n||(2012|-2020)|Iraq (2|007-2012)|Jordan (2|010-2013)|||(2007|-2017)|Tunisia (|2005-2021)|\\n||FHH|<br>non-FHH|<br>FHH|<br>non-FHH|<br>FHH|<br>non-FHH|FHH|non-FHH|FHH|<br>non-FHH|<br>FHH|<br>non-FHH|\\n|**_Panel A: All_**|||||||||||||\\n|Self-Reported FHH|23.32|32.01<sup>*</sup>|20.22|24.79<sup>*</sup>|14.46|14.94|29.52|38.03<sup>*</sup>|50.63|59.70<sup>*</sup>|16.04|17.34<sup>*</sup>|\\n|Reported de jure FHH|21.51|31.92<sup>*</sup>|19.24|24.80<sup>*</sup>|16.08|14.80|29.66|36.80<sup>*</sup>|48.17|59.73<sup>*</sup>|13.46|17.57<sup>*</sup>|\\n|Reported married FHH|29.15|31.01|25.96|24.36|7.76|15.03<sup>*</sup>|28.35|36.73<sup>*</sup>|62.71|59.10|26.55<sup>*</sup>|16.96|\\n|<br>Potential FHH|17.41|32.41<sup>*</sup>|22.28|24.43|8.42|15.35<sup>*</sup>|31.43|36.80<sup>*</sup>|35.30|60.18<sup>*</sup>|13.46|17.69<sup>*</sup>|\\n|Core FHH|19.53|31.25<sup>*</sup>|15.41|24.43<sup>*</sup>|4.16|15.11<sup>*</sup>|30.47|36.17<sup>*</sup>|34.38|59.47<sup>*</sup>|13.41|17.32<sup>*</sup>|\\n|Asset FHH|16.33|32.05<sup>*</sup>|21.76|24.42|8.38|15.30<sup>*</sup>|32.88|36.37<sup>*</sup>|39.06|59.69<sup>*</sup>|15.78|17.28<sup>*</sup>|\\n|Share of female adults>0.5|29.62|31.32<sup>*</sup>|25.94<sup>*</sup>|23.64|17.15<sup>*</sup>|14.19|37.36<sup>*</sup>|34.61|62.04<sup>*</sup>|58.33|19.10<sup>*</sup>|16.42|\\n|Share of employed<br>females>employed males|22.3|31.44<sup>*</sup>|22.11|24.48<sup>*</sup>|7.48|15.37<sup>*</sup>|33.11|36.18<sup>*</sup>|54.89|59.38<sup>*</sup>|16.68|17.24|\\n|Most educated member is female<br>adult & no employed males|20.43|32.11<sup>*</sup>|29.12<sup>*</sup>|24.17|21.87|13.61|27.55|37.22<sup>*</sup>|64.52<sup>*</sup>|58.46|18.15<sup>*</sup>|17.01|\\n|<br>**_Panel B: Rural_**|||||||||||||\\n|Self-Reported FHH|28.16|37.9<sup>*</sup>|36.29|35.34|13.20|17.70|32.42|44.98<sup>*</sup>|45.81|56.02<sup>*</sup>|23.95|25.50|\\n|Reported de jure FHH|25.89|37.69<sup>*</sup>|34.15|35.45|13.55|17.61|33.94|43.01<sup>*</sup>|42.56|56.07<sup>*</sup>|18.79|25.85<sup>*</sup>|\\n|Reported married FHH|32.87|36.92<sup>*</sup>|43.82<sup>*</sup>|35.28|11.18|17.39|30.16|43.30<sup>*</sup>|61.12|55.38|33.52<sup>*</sup>|24.98|\\n|Potential FHH|22.05|38.24<sup>*</sup>|38.52|35.33|15.50|17.41|33.14|44.18<sup>*</sup>|37.66|56.25<sup>*</sup>|22.28|25.77<sup>*</sup>|\\n|Core FHH|25.67|37.00<sup>*</sup>|28.11|35.42|1.93|17.50<sup>*</sup>|34.21|42.46<sup>*</sup>|37.72|55.73<sup>*</sup>|22.96|25.41|\\n|Asset FHH|20.52|38.07<sup>*</sup>|38.85|35.34|12.51|17.54|33.88|43.83<sup>*</sup>|40.33|55.89<sup>*</sup>|24.28|25.42|\\n|Share of female adults>0.5|36.19|36.9|38.08<sup>*</sup>|33.89|21.37<sup>*</sup>|16.00|42.93<sup>*</sup>|40.94|60.23<sup>*</sup>|54.05|27.51<sup>*</sup>|24.34|\\n|Share of employed<br>females>employed males|27.46|37.17<sup>*</sup>|34.11|35.46|13.03|17.51|39.58|42.13<sup>*</sup>|59.83|55.11|25.30|25.32|\\n|Most educated member is female<br>adult & no employed males|24.25|37.90<sup>*</sup>|51.45<sup>*</sup>|34.76|23.55<sup>*</sup>|16.19|33.68|43.27<sup>*</sup>|60.50|54.97|28.20<sup>*</sup>|24.82|\\n|<br>**_Panel C: Urban_**|||||||||||||\\n|Self-Reported FHH|16.96|23.08<sup>*</sup>|15.19|18.52<sup>*</sup>|14.70|14.36|26.68|30.24<sup>*</sup>|50.17|58.89<sup>*</sup>|12.29|13.27<sup>*</sup>|\\n|<br>Reported de jure FHH|16.96|22.97<sup>*</sup>|15.08|18.48<sup>*</sup>|16.61|14.21|25.75|29.90<sup>*</sup>|47.92|58.92<sup>*</sup>|11.59|13.33<sup>*</sup>|\\n|Reported married FHH|16.81|22.39<sup>*</sup><br>|15.97|18.20<br>|7.29|14.53<sup>*</sup><br>|26.42|29.62<sup>*</sup>|61.23|58.33<br>|17.25<sup>*</sup>|13.09<br>|\\n|Potential FHH|11.5|23.56<sup>*</sup>|14.49|18.25<sup>*</sup>|7.33|14.91<sup>*</sup>|29.05|29.28|32.54|59.48<sup>*</sup>|8.69|13.75<sup>*</sup>|\\n|<br>Core FHH|<br>13.47|22.58<sup>*</sup>|<br>9.54|18.21<sup>*</sup>|<br>4.43|14.60<sup>*</sup>|<br>25.24|<br>29.51<sup>*</sup>|<br>31.97|58.70<sup>*</sup>|<br>8.68|13.32<sup>*</sup>|\\n|Asset FHH|9.11|23.18<sup>*</sup>|12.56|18.25<sup>*</sup>|7.72|14.82<sup>*</sup>|31.32<sup>*</sup>|28.90|36.09|58.94<sup>*</sup>|9.64|13.37<sup>*</sup>|\\n|Share of female adults>0.5|21.27|22.61|17.92|18.28|16.24<sup>*</sup>|13.81|31.12<sup>*</sup>|27.94|60.55<sup>*</sup>|57.77|14.40<sup>*</sup>|12.68|\\n|Share of employed<br>females>employed males|17.58|22.65<sup>*</sup>|11.31|18.40<sup>*</sup>|6.65|14.91<sup>*</sup>|26.20|29.73<sup>*</sup>|52.13|58.69<sup>*</sup>|12.50|13.23|\\n|Most educated member is female<br>adult & no employed males|16.40|23.10<sup>*</sup>|18.57|18.15|21.52|13.06|20.58|30.74<sup>*</sup>|63.05<sup>*</sup>|57.76|13.38|13.11|\\n|<br>**_Panel D: Have children under 14_**|||||||||||||\\n|Self-Reported FHH|35.2|39.33<sup>*</sup>|22.15|25.99<sup>*</sup>|29.00<sup>*</sup>|18.18|31.91|40.75<sup>*</sup>|67.33|64.40|28.70<sup>*</sup>|22.48|\\n|<br>Reported de jure FHH|37.62|39.05|21.27|26.01<sup>*</sup>|34.42<sup>*</sup>|18.11|33.39|39.21<sup>*</sup>|65.85|64.47|25.84<sup>*</sup>|22.81|\\n|Reported married FHH|31.51|39.24<sup>*</sup>|26.85|25.66|13.71|18.82|29.33|39.71<sup>*</sup>|72.35|64.45|33.15<sup>*</sup>|22.64|\\n|<br>Potential FHH|32.83|39.34<sup>*</sup>|27.84|25.64|25.95<sup>*</sup>|18.55|33.94|39.50<sup>*</sup>|59.40|64.60|29.98<sup>*</sup>|22.55|\\n|Core FHH|30.60|39.13<sup>*</sup>|23.68|25.68|21.68|18.73|32.23|38.93<sup>*</sup>|48.35|64.63<sup>*</sup>|25.61|22.89|\\n|Asset FHH|32.24|39.24<sup>*</sup>|28.13|25.65|27.28<sup>*</sup>|18.55|35.51|39.04<sup>*</sup>|64.80|64.51|33.36<sup>*</sup>|22.60|\\n|Share of female adults>0.5|40.1|38.72|27.32|24.90|26.82<sup>*</sup>|16.80|39.83<sup>*</sup>|37.41|70.08<sup>*</sup>|63.15|27.90<sup>*</sup>|21.48|\\n|Share of emloed|||||||||||||\\n|py<br>|35.69|39.1<sup>*</sup>|25.57|25.68|17.43|18.79|35.55|38.89<sup>*</sup>|67.74|64.38|26.60<sup>*</sup>|22.68|\\n|females>employed males|||||||||||||\\n|<br>Most educated member is female||<sup>*</sup>|<sup>*</sup>||<sup>*</sup>|||<sup>*</sup>|<sup>*</sup>||<sup>*</sup>||\\n|adult & no employed males|32.09|39.48|33.62|25.37|34.47|16.46|30.61|39.74|77.00|63.23|29.63|22.22|\\n|<br>**_Panel E: No children under 14_**|||||||||||||\\n|Self-Reported FHH|10.53|10.52|6.79|6.15|3.88|4.13|12.21|13.63|30.83|31.39|8.71|9.28|\\n||||||52||||||||\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [53]}}, {\"page\": 54, \"chunk\": 54, \"input_text\": \"|Reported de jure FHH|10.39|10.55|6.26<br>|6.26|4.54|3.98|12.76|13.34<br>|30.16|31.50|8.69|9.28|\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n|Reported married FHH|12.7|10.49|13.20<sup>*</sup>|6.17|0.00|4.20|9.25|13.42<sup>*</sup>|36.90|31.19|8.97|9.18|\\n|Potential FHH|6.24|11.58<sup>*</sup><br>|5.05|6.36<br>|0.81|4.77<sup>*</sup><br>|11.70|13.58<sup>*</sup>|23.57|32.92<sup>*</sup><br>|7.74|9.57<sup>*</sup><br>|\\n|Core FHH|7.59|10.65<sup>*</sup><br>|1.33|6.39<sup>*</sup>|0.00|4.33<sup>*</sup><br>|11.32|13.31|23.75|31.66<sup>*</sup><br>|7.38|9.29<sup>*</sup>|\\n|Asset FHH|6.14<br>|11.29<sup>*</sup>|4.76<br>|6.36|0.68|4.72<sup>*</sup><br>|12.19<br>|13.42|25.22<br>|32.03<sup>*</sup>|8.97<br>|9.21|\\n|Share of female adults>0.5|12.16<sup>*</sup>|9.89|8.69<sup>*</sup>|4.83|2.74|4.86<sup>*</sup>|15.30<sup>*</sup>|11.80|33.93<sup>*</sup>|29.72|11.40<sup>*</sup>|7.91|\\n|Share of employed<br>females>employed males|8.79|10.7<sup>*</sup>|3.09|6.57<sup>*</sup>|0.47|4.64<sup>*</sup>|10.29|13.61<sup>*</sup>|31.32|31.29|10.27<sup>*</sup>|8.99|\\n|Most educated member is female<br>adult & no employed males|8.74|10.90<sup>*</sup>|9.36<sup>*</sup>|5.84|3.02|4.41|10.04|14.05<sup>*</sup>|35.36<sup>*</sup>|30.23|11.40|8.51|\\n\\nNote: The data are pooled across all available years for each country. Headcount poverty rates are estimated using per capita household expenditures. Stars indicate statistically significant difference in headcount poverty between FHHs and non-FHHs in each category at the 5% or lower level. Population sampling weights are applied.\\n\\n53\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [54]}}, {\"page\": 55, \"chunk\": 55, \"input_text\": \"# **<u>Table A.10. Probabilities of Being Poor, Linear Probability Model (Main FHH Types)</u>**\\n\\n|||**Specif**|**ication 1**|||**Spec**|**ification 2**||\\n|---|---|---|---|---|---|---|---|---|\\n||**FHH Type 1**<br>**Self-reported**|**FHH Type 2**<br>**Majority-female-adult**|**FHH Type 3**<br>**Potential**|**FHH Type 4**<br>**Most-educated-female-adult**|**FHH Type 1**<br>**Self-reported**|**FHH Type 2**<br>**Majority-female-adult**|**FHH Type 3**<br>**Potential**|**FHH Type 4**<br>**Most-educated-female-adult**|\\n||-0.007***||||-0.044***||||\\n|Self-reported FHH|<br>(0.00)||||<br>(0.00)||||\\n|Self-reported FHH # Number of children age|0.008***||||0.006***||||\\n|0-14|(0.00)||||(0.00)||||\\n|||0.052***||||0.032***|||\\n|Share of female adults>0.5||(0.00)||||(0.00)|||\\n|Share of female adults>0.5# Number of||-0.007***||||-0.014***|||\\n|children age 0-14||(0.00)||||(0.00)|||\\n|il|||-0.027***||||-0.017***||\\n|Potenta FHH|||(0.00)||||(0.00)||\\n|Potential FHH# Number of children age 0-14|||0.014***<br>||||0.011***<br>||\\n||||(0.00)||||(0.00)||\\n|Edtd fl||||0.002||||-0.006**|\\n|ucae emaes||||(0.00)||||(0.00)|\\n|Educated females# Number of children age 0-||||<br>0.013***||||<br>0.013***|\\n|14||||(0.00)||||(0.00)|\\n|_Household head`s characteristics_|||||||||\\n|Hd`|||||-0.001***|-0.001***|-0.001***|-0.001***|\\n|eas age|||||(0.00)|(0.00)|(0.00)|(0.00)|\\n|ih i ll i i|||||-0.066***|-0.064***|-0.064***|-0.065***|\\n|Hgest educaton eve s prmary|||||(0.00)|(0.00)|(0.00)|(0.00)|\\n|Highest education level is secondary|||||-0.123***<br>|-0.121***<br>|-0.121***<br>|-0.121***<br>|\\n||||||(0.00)|(0.00)|(0.00)|(0.00)|\\n|Hiht dti ll i tti|||||<br>-0.178***|<br>-0.176***|<br>-0.176***|<br>-0.176***|\\n|ges eucaon eve s erary|||||(0.00)|(0.00)|(0.00)|(0.00)|\\n|Head is married|||||-0.020***<br>|0.007***<br>|0.005**<br>|0.006***<br>|\\n||||||(0.00)|(0.00)|(0.00)|(0.00)|\\n|Hd i ld|||||-0.037***|-0.032***|-0.031***|-0.027***|\\n|ea s empoye|||||(000)|(000)|(000)|(000)|\\n|_Household characteristics_|||||.<br>|.<br>|.<br>|.<br>|\\n|Hhld i|||||0.022***|0.024***|0.023***|0.023***|\\n|ouseo sze|||||(0.00)|(0.00)|(0.00)|(0.00)|\\n|Number of children ae 0-14|0.077***|0.081***|0.076***|0.077***|0.050***|0.054***|0.049***|0.049***|\\n|g|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Sh f hhld b 1524|||||-0.008***|-0.008***|-0.008***|-0.009***|\\n|are o ouseo memers age -|||||(0.00)|(0.00)|(0.00)|(0.00)|\\n||||||-0.022***|-0.019***|-0.017***|-0.020***|\\n|Share of household members age 60 and older|||||(0.00)|(0.00)|(0.00)|(0.00)|\\n|Ub|||||<br>-0.084***|<br>-0.084***|<br>-0.084***|<br>-0.084***|\\n|ran|||||(0.00)|(0.00)|(0.00)|(0.00)|\\n|I|-0.149***|-0.152***|-0.149***|-0.148***|-0.181***|-0.182***|-0.179***|-0.180***|\\n|raq|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Jordan|-0.144***|-0.147***|-0.144***|-0.146***|-0.092***|-0.094***|-0.090***|-0.092***|\\n||(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|\\n||<br>-0.155***|<br>-0.160***|<br>-0.156***|<br>-0.158***|<br>-0.207***|<br>-0.208***|<br>-0.213***|<br>-0.212***|\\n|Mauritania|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|\\n||0181***|0178***|0181***|0180***|0185***|0183***|0186***|0185***|\\n|West Bank and Gaza|.<br>(0.01)|.<br>(0.01)|.<br>(0.01)|.<br>(0.01)|.<br>(0.01)|.<br>(0.01)|.<br>(0.01)|.<br>(0.01)|\\n||<br>-0.060***|<br>-0.063***|<br>-0.060***|<br>-0.061***|<br>-0.054***|<br>-0.055***|<br>-0.054***|<br>-0.054***|\\n|Tunisia|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|_cons|0.127***|0.112***|0.133***|0.126***|0.312***|0.261***|0.270***|0.267***|\\n||(0.00)|(0.00)|(0.00)|(0.00)|(0.01)|(0.01)|(0.01)|(0.01)|\\n|r2_a|0.16|0.17|0.16|0.16|0.21|0.21|0.21|0.21|\\n|N|214931|214931|214931|214931|211069|211069|211069|211069|\\n\\nNote: ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Robust standard errors are in parentheses. All regressions control for survey rounds fixed effects. The reference groups for head`s education without formal education. The reference group for share of household members is share of members aged between 25 and 59 years. The reference country is the Arab Republic of Egypt.\\n\\n54\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [55]}}, {\"page\": 56, \"chunk\": 56, \"input_text\": \"# **<u>Table A.11. Probabilities of Being Poor for Other FHH Types, Linear Probability Model</u>**\\n\\n||||**Specification 1**|||||**Specification 2**|||\\n|---|---|---|---|---|---|---|---|---|---|---|\\n||**(1)**|**(2)**|**(3)**|**(4)**|**(5)**|**(6)**|**(7)**|**(8)**|**(9)**|**(10)**|\\n||-0.003|||||-0.027***|||||\\n|De-jure FHH|<br>|||||<br>|||||\\n||(0.00)|||||(0.00)|||||\\n|De-jure FHH # Number of children age 0-14|0.013***<br>|||||0.009***<br>|||||\\n||(0.00)|||||(0.00)|||||\\n|Married FHH||-0.031***|||||-0.058***||||\\n|||(0.01)|||||(0.01)||||\\n|id b f hild||0.005**|||||0.006**||||\\n|Marre FHH # Numer o cren age 0-14||(0.00)|||||(0.00)||||\\n||||0.009**|||||0.008**|||\\n|Employed FHH|||(0.00)|||||(0.00)|||\\n|Employed FHH # Number of children age 0-14|||-0.001<br>|||||-0.004*<br>|||\\n||||(0.00)|||||(0.00)|||\\n|A FHH||||-0.019***|||||-0.018***||\\n|sset||||(0.00)|||||(0.00)||\\n|A FHH # Nb f hild 014||||0.014***|||||0.011***||\\n|sset umer o cren age -||||(0.00)|||||(0.00)||\\n||||||-0023***|||||-0010**|\\n|Core FHH|||||.<br>(0.00)|||||.<br>(0.00)|\\n|Core FHH # Number of children age 0-14|||||0.006**<br>(0.00)|||||0.007***<br>(0.00)|\\n|_Household head`s characteristics_|||||||||||\\n|`||||||-0.001***|-0.001***|-0.001***|-0.001***|-0.001***|\\n|Heads age||||||(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Hihest education level is rimar||||||-0.065***|-0.065***|-0.064***|-0.064***|-0.064***|\\n|g py||||||(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Hihest edcation leel is secondar||||||-0.122***|-0.122***|-0.121***|-0.121***|-0.121***|\\n|g u v y||||||(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Highest education level is tertiary||||||-0.177***<br>|-0.177***<br>|-0.176***<br>|-0.177***<br>|-0.176***<br>|\\n|||||||(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Head is married||||||-0.008**|0.011***|0.006***|0.006**|0.006***|\\n|||||||(0.00)<br>|(0.00)<br>|(0.00)<br>|(0.00)<br>|(0.00)<br>|\\n|||||||-0.033***|-0.038***|-0.032***|-0.031***|-0.032***|\\n|Head is employed||||||(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|_Household characteristics_|||||||||||\\n|hld i||||||0.023***|0.023***|0.023***|0.023***|0.023***|\\n|Houseo sze||||||(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Nb f hild 014|0.077***|0.078***|0.078***|0.077***|0.078***|0.050***|0.050***|0.050***|0.049***|0.050***|\\n|umer o cren age -|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Share of household members age 15-24||||||-0.009***<br>|-0.008***<br>|-0.009***<br>|-0.009***<br>|-0.009***<br>|\\n|||||||(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Sh f hhld b 60 d ld||||||-0.021***|-0.021***|-0.021***|-0.018***|-0.021***|\\n|are o ouseo memers age an oer||||||(0.00)<br>|(0.00)<br>|(0.00)<br>|(0.00)<br>|(0.00)<br>|\\n|||||||-0.084***|-0.084***|-0.084***|-0.084***|-0.084***|\\n|Urban||||||(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|I|-0.149***|-0.150***|-0.149***|-0.149***|-0.149***|-0.181***|-0.182***|-0.181***|-0.180***|-0.181***|\\n|raq|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n||-0.145***|-0.145***|-0.145***|-0.144***|-0.145***|-0.091***|-0.093***|-0.093***|-0.090***|-0.092***|\\n|Jordan|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|\\n||<br>0.122***|<br>0.123***|<br>0.123***|<br>0.125***|<br>0.124***|<br>0.058***|<br>0.059***|<br>0.059***|<br>0.061***|<br>0.059***|\\n|Mauritania|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|\\n||0.181***|0.180***|0.180***|0.180***|0.180***|0.185***|0.184***|0.184***|0.184***|0.184***|\\n|West Bank and Gaza|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|\\n||-0060***|-0061***|-0061***|-0060***|-0060***|-0054***|-0054***|-0054***|-0054***|-0054***|\\n|Tunisia|.<br>(0.00)|.<br>(0.00)|.<br>(0.00)|.<br>(0.00)|.<br>(0.00)|.<br>(0.00)|.<br>(0.00)|.<br>(0.00)|.<br>(0.00)|.<br>(0.00)|\\n|_cons|0.126***|0.126***|0.125***|0.126***|0.127***|0.289***|0.284***|0.272***|0.268***|0.273***|\\n||(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|\\n|Adjuster R2|0.16|0.16|0.16|0.16|0.16|0.21|0.21|0.21|0.21|0.21|\\n|<br>Number of observations|214931|214931|214931|214931|214931|211069|211069|211069|211069|211069|\\n\\nNote: ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Robust standard errors are in parentheses. All regressions control for survey rounds fixed effects. The reference groups for head`s education without formal education. The reference group for share of household members is share of members aged between 25 and 59 years. The reference country is the Arab Republic of Egypt.\\n\\n55\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [56]}}, {\"page\": 57, \"chunk\": 57, \"input_text\": \"# **<u>Table A.12. Probabilities of Being Poor, Logit Model</u>**\\n\\n|||**Specif**|**ication 1**|||**Speci**|**fication 2**||\\n|---|---|---|---|---|---|---|---|---|\\n||**FHH Type 1**<br>**Self-reported**|**FHH Type 2**<br>**Majority-female-adult**|**FHH Type 3**<br>**Potential**|**FHH Type 4**<br>**Most-educated-female-adult**|**FHH Type 1**<br>**Self-reported**|**FHH Type 2**<br>**Majority-female-adult**|**FHH Type 3**<br>**Potential**|**FHH Type 4**<br>**Most-educated-female-adult**|\\n|Self-reported FHH|-0.110***<br>||||-0.290***<br>||||\\n||(0.02)||||(0.03)||||\\n|Self-reported FHH # Number of children age|<br>0.057***||||<br>0.036***||||\\n|0-14|(0.01)||||(0.01)||||\\n|||0.415***||||0.323***|||\\n|Share of female adults>0.5||<br>(0.02)||||<br>(0.02)|||\\n|Share of female adults>0.5# Number of||-0.072***||||-0.116***|||\\n|children age 0-14||(0.01)||||(0.01)|||\\n|Potential FHH|||-0.339***<br>||||-0.165***<br>||\\n||||(0.02)||||(0.03)||\\n|Pil FHH# Nb f hild 014|||0.124***||||0.075***||\\n|otenta umer o cren age -|||(0.01)||||(0.01)||\\n|Educated females||||-0.037||||-0.040|\\n|||||(0.02)||||(0.03)|\\n|Educated females# Number of children age 0-||||0.086***||||0.074***|\\n|14||||(0.01)||||(0.01)|\\n|_Household head`s characteristics_|||||||||\\n|`|||||-0.009***|-0.009***|-0.007***|-0.008***|\\n|Heads age|||||(0.00)|(0.00)|(0.00)|(0.00)|\\n|Hihest education level is rimar|||||-0.451***|-0.443***|-0.443***|-0.448***|\\n|g py|||||(0.02)|(0.02)|(0.02)|(0.02)|\\n||||||-0.911***|-0.903***|-0.900***|-0.903***|\\n|Highest education level is secondary|||||(0.02)|(0.02)|(0.02)|(0.02)|\\n|Hiht dti ll i tti|||||-1.572***|-1.568***|-1.563***|-1.562***|\\n|ges eucaon eve s erary|||||(0.03)|(0.03)|(0.03)|(0.03)|\\n|d i id|||||-0.097***|0.083***|0.059***|0.068***|\\n|Hea s marre|||||(0.03)|(0.02)|(0.02)|(0.02)|\\n|Hd i ld|||||-0.256***|-0.213***|-0.212***|-0.178***|\\n|ea s empoye|||||(0.02)|(0.02)|(0.02)|(0.02)|\\n|_Household characteristics_|||||||||\\n|hld i|||||0.160***|0.167***|0.160***|0.165***|\\n|Houseo sze|||||(0.00)|(0.00)|(0.00)|(0.00)|\\n|Nb f hild 014|0.468***|0.508***|0.456***|0.470***|0.301***|0.338***|0.298***|0.293***|\\n|umer o cren age -|(0.00)|(0.00)|(0.00)|(0.00)|(0.01)<br>|(0.01)<br>|(0.01)<br>|(0.01)<br>|\\n||||||-0.092***|-0.095***|-0.095***|-0.100***|\\n|Share of household members age 15-24|||||(0.01)|(0.01)|(0.01)|(0.01)|\\n|Share of household members age 60 and older|||||-0.325***<br>|-0.316***<br>|-0.286***<br>|-0.315***<br>|\\n||||||(0.02)|(0.02)|(0.02)|(0.02)|\\n||||||<br>-0.614***|<br>-0.610***|<br>-0.611***|<br>-0.611***|\\n|Urban|||||(0.01)|(0.01)|(0.01)|(0.01)|\\n||-1.072***|-1.106***|-1.065***|-1.070***|-1.441***|-1.457***|-1.422***|-1.429***|\\n|Iraq|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|\\n||-1.184***|-1.203***|-1.180***|-1.204***|-0.886***|-0.903***|-0.863***|-0.880***|\\n|Jordan|(0.07)|(0.07)|(0.07)|(0.07)|(0.08)|(0.08)|(0.08)|(0.08)|\\n|Mauritania|-1.117***|-1.152***|-1.125***|-1.137***|-1.622***|-1.628***|-1.665***|-1.664***|\\n||(0.05)|(0.05)|(0.05)|(0.05)|(0.05)|(0.05)|(0.05)|(0.05)|\\n||<br>0.796***|<br>0.784***|<br>0.797***|<br>0.795***|<br>0.887***|<br>0.880***|<br>0.896***|<br>0.892***|\\n|West Bank and Gaza|(0.04)|(0.04)|(0.04)|(0.04)|(0.04)|(0.04)|(0.04)|(0.04)|\\n|Tii|-0.468***|-0.481***|-0.466***|-0.473***|-0.438***|-0.445***|-0.433***|-0.434***|\\n|unsa|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|\\n|_cons|<br>-1.917***|<br>-2.062***|<br>-1.861***|<br>-1.932***|<br>-0.606***|<br>-0.975***|<br>-0.877***|<br>-0.915***|\\n||(0.04)|(0.04)|(0.04)|(0.04)|(0.06)|(0.06)|(0.06)|(0.06)|\\n|N|214931|214931|214931|214931|211069|211069|211069|211069|\\n\\nNote: ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Robust standard errors are in parentheses. All regressions control for survey rounds fixed effects. The reference groups for head`s education without formal education. The reference group for share of household members is share of members aged between 25 and 59 years. The reference country is the Arab Republic of Egypt.\\n\\n56\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [57]}}, {\"page\": 58, \"chunk\": 58, \"input_text\": \"# **<u>Table A.13. Probabilities of Being Poor for Other FHH Types, Logit Model</u>**\\n\\n||||**Specification 1**|||||**Specification 2**|||\\n|---|---|---|---|---|---|---|---|---|---|---|\\n||**(1)**|**(2)**|**(3)**|**(4)**|**(5)**|**(6)**|**(7)**|**(8)**|**(9)**|**(10)**|\\n||-0098***|||||-0098**|||||\\n|De-jure FHH|.<br>(0.02)|||||.<br>(0.05)|||||\\n|De-jure FHH # Number of children ae 0-14|0.094***|||||0.059***|||||\\n|g|(0.01)|||||(0.01)|||||\\n|Mid FHH||-0.182***|||||-0.392***||||\\n|arre||(0.05)|||||(0.06)||||\\n|Married FHH # Number of children age 0-14||0.024<br>|||||0.029*<br>||||\\n|||(0.02)|||||(0.02)||||\\n|Emloed FHH|||0.036|||||-0.135***|||\\n|py|||(0.03)<br>|||||(0.05)<br>|||\\n|Eld FHH # Nb f hild 014|||-0.002|||||0.058***|||\\n|mpoye umer o cren age -|||(0.01)|||||(0.02)|||\\n|||||-0.235***|||||-0.103***||\\n|Asset FHH||||<br>(0.03)|||||<br>(0.03)||\\n|Asset FHH # Number of children ae 0-14||||0.106***|||||0.056***||\\n|g||||(0.01)|||||(0.01)||\\n|C FHH|||||-0.259***|||||0.028|\\n|ore|||||(0.04)|||||(0.03)|\\n|Core FHH # Number of children age 0-14|||||0.060***<br>|||||-0.016<br>|\\n|<br>_Household head`s characteristics_|||||(0.02)|||||(0.01)|\\n|`||||||-0.008***|-0.009***|-0.008***|-0.008***|-0.008***|\\n|Heads age||||||(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Hih di ll i i||||||-0.443***|-0.449***|-0.442***|-0.442***|-0.442***|\\n|gest eucaton eve s prmary||||||(0.02)|(0.02)|(0.02)|(0.02)|(0.02)|\\n|Hiht dti ll i d||||||-0.902***|-0.909***|-0.900***|-0.900***|-0.901***|\\n|ges eucaon eve s seconary||||||(0.02)|(0.02)|(0.02)|(0.02)|(0.02)|\\n|||||||-1.563***|-1.570***|-1.562***|-1.562***|-1.564***|\\n|Highest education level is tertiary||||||(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|\\n|Head is married||||||0.061|0.104***|0.065***|0.065***|0.066***|\\n|||||||(0.04)<br>|(0.02)<br>|(0.02)<br>|(0.02)<br>|(0.02)<br>|\\n|||||||-0.216***|-0.267***|-0.216***|-0.210***|-0.216***|\\n|Head is employed||||||(0.02)|(0.02)|(0.02)|(0.02)|(0.02)|\\n|_Household characteristics_|||||||||||\\n|Hhld i||||||0.162***|0.161***|0.163***|0.162***|0.163***|\\n|ouseo sze||||||(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Nb f hild 014|0.470***|0.477***|0.478***|0.465***|0.474***|0.298***|0.303***|0.300***|0.298***|0.303***|\\n|umer o cren age -|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|\\n|Share of household members ae 15-24||||||-0.096***|-0.092***|-0.100***|-0.099***|-0.100***|\\n|g||||||(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|\\n|Sh f hhld b 60 d ld||||||<br>-0.316***|<br>-0.326***|<br>-0.323***|<br>-0.304***|<br>-0.324***|\\n|are o ouseo memers age an oer||||||(0.02)|(0.02)|(0.02)|(0.02)|(0.02)|\\n|||||||-0.613***|-0.616***|-0.611***|-0.610***|-0.612***|\\n|Urban||||||(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|\\n||-1.077***|-1.081***|-1.079***|-1.071***|-1.076***|<br>-1.434***|<br>-1.446***|<br>-1.436***|<br>-1.432***|<br>-1.441***|\\n|Iraq|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|\\n||-1.188***|-1.194***|-1.191***|-1.183***|-1.189***|-0.876***|-0.895***|-0.881***|-0.872***|-0.885***|\\n|Jordan|(0.07)|(0.07)|(0.07)|(0.07)|(0.07)|(0.08)|(0.08)|(0.08)|(0.08)|(0.08)|\\n||<br>-1.115***|<br>-1.080***|<br>-1.097***|<br>-1.098***|<br>-1.091***|<br>-1.649***|<br>-1.597***|<br>-1.647***|<br>-1.649***|<br>-1.636***|\\n|Mauritania|(0.05)|(0.05)|(0.05)|(0.05)|(0.05)|(0.05)|(0.05)|(0.05)|(0.05)|(0.05)|\\n|West Bank and Gaza|<br>0.795***|<br>0.794***|<br>0.794***|<br>0.786***|<br>0.795***|<br>0.889***|<br>0.883***|<br>0.889***|<br>0.887***|<br>0.887***|\\n||(0.04)|(0.04)|(0.04)|(0.04)|(0.04)|(0.04)|(0.04)|(0.04)|(0.04)|(0.04)|\\n||<br>-0.468***|<br>-0.469***|<br>-0.469***|<br>-0.465***|<br>-0.467***|<br>-0.434***|<br>-0.436***|<br>-0.434***|<br>-0.433***|<br>-0.435***|\\n|Tunisia|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|\\n|_cons|-1.923***<br>|-1.935***<br>|-1.943***<br>|-1.926***<br>|-1.927***<br>|-0.849***<br>|-0.781***<br>|-0.866***<br>|-0.888***<br>|-0.871***<br>|\\n||(0.04)|(0.04)|(0.04)|(0.04)|(0.04)|(0.07)|(0.06)|(0.06)|(0.06)|(0.06)|\\n|N|214931|214931|214931|214931|214931|211069|211069|211069|211069|211069|\\n\\nNote: ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Robust standard errors are in parentheses. All regressions control for survey rounds fixed effects. The reference groups for head`s education without formal education. The reference group for share of household members is share of members aged between 25 and 59 years. The reference country is the Arab Republic of Egypt.\\n\\n57\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [58]}}, {\"page\": 59, \"chunk\": 59, \"input_text\": \"# **-** **<u>Table A.14. Probabilities of Being Poor for Combination of Main Female Headed Household Types, Linear Probability Model</u>**\\n\\n||**Specific**|**ation 1**|**Specif**|**ication 2**|\\n|---|---|---|---|---|\\n||**(1)**|**(2)**|**(3)**|**(4)**|\\n||-0037***||-0033***||\\n|Overlap of self-reported FHHs, potential FHHs, and most-educated-female-adult FHHs|.<br>(0.00)<br>||.<br>(0.00)<br>||\\n|Overlap of self-reported FHHs, potential FHHs, and most-educated-female-adult FHHs # Number of children age 0-14|0.005**<br>(0.00)||0.007***<br>(0.00)||\\n|Overlap of self-reported FHHs, potential FHHs, majority of females and most-educated-female-adult FHHs||-0.037***<br>(0.00)||-0.033***<br>(0.00)|\\n|Overlap of self-reported FHHs, potential FHHs, majority of females and most-educated-female-adult FHHs # Number of children age 0-14||0.005**<br>(0.00)||0.007***<br>(0.00)|\\n|_Household head`s characteristics_|||||\\n|`|||-0.001***|-0.001***|\\n|Heads age|||(0.00)|(0.00)|\\n|Highest education level is primary|||-0.064***<br>(0.00)|-0.064***<br>(0.00)|\\n||||-0.121***|-0.121***|\\n|Highest education level is secondary|||(0.00)|(0.00)|\\n||||-0.176***|-0.176***|\\n|Highest education level is tertiary|||(0.00)|(0.00)|\\n|Head is married|||0.001<br>|0.001<br>|\\n||||(0.00)<br>|(0.00)<br>|\\n||||-0.033***|-0.033***|\\n|Head is employed|||<br>(0.00)|<br>(0.00)|\\n|_Household characteristics_|||||\\n|Household size|||0.023***|0.023***|\\n||||(0.00)|(0.00)|\\n|Nb f hild 014|0.078***|0.078***|0.050***|0.050***|\\n|umer o cren age -|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Sh f hhld b 1524|||-0.009***|-0.009***|\\n|are o ouseo memers age -|||(0.00)|(0.00)|\\n|Share of household members ae 60 and older|||-0.021***|-0.021***|\\n|g|||(0.00)|(0.00)|\\n||||-0.084***|-0.084***|\\n|Urban|||(0.00)|(0.00)|\\n||-0.150***|-0.150***|-0.181***|-0.181***|\\n|Iraq|(0.00)|(0.00)|(0.00)|(0.00)|\\n||-0.145***|-0.145***|-0.093***|-0.093***|\\n|Jordan|(0.01)|(0.01)|(0.01)|(0.01)|\\n||-0.150***|-0.150***|-0.209***|-0.209***|\\n|Mauritania|(0.01)|(0.01)|(0.01)|(0.01)|\\n||0.180***|0.180***|0.184***|0.184***|\\n|West Bank and Gaza|<br>(0.01)|<br>(0.01)|<br>(0.01)|<br>(0.01)|\\n||-0.061***|-0.061***|-0.054***|-0.054***|\\n|Tunisia|(0.00)|(0.00)|(0.00)|(0.00)|\\n|_cons|0.129***|0.129***|0.283***|0.283***|\\n||(0.00)|(0.00)|(0.01)|(0.01)|\\n|Adjuster R2|0.16|0.16|0.21|0.21|\\n|Number of observations|214931|214931|211069|211069|\\n\\nNote: ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Robust standard errors are in parentheses.\\n\\n58\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [59]}}, {\"page\": 60, \"chunk\": 60, \"input_text\": \"# **<u>Table A.15. Average change for mobility per year between the survey rounds</u>**\\n\\n||**Upward mobility (%)**|**Downward mobility (%)**|\\n|---|---|---|\\n|Jordan (2010-2013)|17.8|3.5|\\n|Egypt, Arab Rep. (2017-2020)|9.7|5.3|\\n|Iraq (2007-2012)|9.0|2.7|\\n|Mauritania (2014-2019)|8.1|2.8|\\n|West Bank and Gaza (2011-2017)|5.1|8.1|\\n|Tunisia (2015-2021)|3.5|1.3|\\n\\n**Note:** Countries are ranked in a decreasing order of upward mobility.\\n\\n59\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [60]}}, {\"page\": 61, \"chunk\": 61, \"input_text\": \"**Figure A.1. Headcount Poverty Rates in Self-Reported Male-Headed Households (%), by Number of Female Adults**\\n\\n<!-- Start of picture text --> Egypt Iraq<br>2012<br>2007<br>2015<br>2017<br>2012<br>2020<br>0 10 20 30 40<br>0 50 100<br>Jordan Mauritania<br>2004<br>2007<br>2008<br>2014<br>2012<br>2019<br>0 10 20 30 40 0 20 40 60 80<br>Palestine Tunisia<br>2007 2005<br>2009 2010<br>2011 2015<br>2017 2021<br>0 20 40 60 80 100 0 20 40 60<br>0 1 2 3 4 >4<br><!-- End of picture text -->\\n\\n**Note:** Headcount poverty rates are estimated using per capita household expenditures. The numbers of female adults are shown for 0, 1, 2, 3, 4, and more than 4 adults. The years are shown on the y-axis and the poverty rates are shown on the x-axis.\\n\\n60\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [61]}}, {\"page\": 62, \"chunk\": 62, \"input_text\": \"# **Figure A.2. FHH–non-FHH Differences in Headcount Poverty Rate in MENA, Pooled Cross Sections (percentage points)**\\n\\n<!-- Start of picture text --> Egypt Iraq<br>***<br>***<br>***<br>**<br>***<br>*** *** ***<br>***<br>*** *** ***<br>Self-reported Married *** Core ***Females>0.5 Educated Self-reported Married Core Females>0.5 Educated<br>De jure Potential Asset Employed De jure Potential Asset Employed<br>Jordan Mauritania<br>***<br>***<br>***<br>*** ***<br>*** ***<br>** *** *** *** *** *** ***<br>*** ***<br>Self-reported Married Core Females>0.5 Educated Self-reported Married Core Females>0.5 Educated<br>De jure Potential Asset Employed De jure Potential Asset Employed<br>Palestine Tunisia<br>***<br>***<br>***<br>** ***<br>*** *** ***<br>*** *** **<br>*** *** *** *** ***<br>Self-reported Married Core Females>0.5 Educated Self-reported Married Core Females>0.5 Educated<br>De jure Potential Asset Employed De jure Potential Asset Employed<br>0 5<br>-5 0<br>-10 -5<br>difference (pps) difference (pps)<br>-15 -10<br>10 5<br>5<br>0<br>0<br>-5 -5<br>difference (pps) difference (pps)<br>-10 -10<br>10 10<br>0<br>5<br>-10<br>difference (pps) -20 0 difference (pps)<br>-30 -5<br><!-- End of picture text -->\\n\\n**Note:** Authors’ calculation based on pooled cross sections. The four main types of FHHs are shown in darker color, the five sub-types of FHHs are shown in lighter color. The headcount poverty rate is applied to per capita household expenditures, of FHHs versus the rest of the households. Stars indicate statistically significantly higher headcount poverty ratio between FHHs and non-FHHs in each category. ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels respectively.\\n\\n61\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [62]}}, {\"page\": 63, \"chunk\": 63, \"input_text\": \"# **Figure A.3. Correlation Between Probabilities of Female-Headed Households Escaping Poverty in Second Year and Number of Children (percentage)**\\n\\n<!-- Start of picture text --> Egypt 2017-2020 Iraq 2007-2012<br>0 1 2 3 4 0 1 2 3 4<br>Number of children Number of children<br>Jordan 2010-2013 Mauritania 2014-2019<br>0 1 2 3 4 0 1 2 3 4<br>Number of children Number of children<br>Palestine 2011-2017 Tunisia 2015-2021<br>0 1 2 3 4 0 1 2 3 4<br>Number of children Number of children<br>Self-reported FHH Females>0.5<br>Potential FHH Educated females<br>.38 .5<br>.36 .48<br>.34 .46<br>.32 .44<br>.3 .42<br>Percentage (%) Percentage (%)<br>.28 .4<br>.55 .48<br>.5 .46<br>.44<br>.45<br>.42<br>Percentage (%) Percentage (%)<br>.4 .4<br>.45<br>.25<br>.4<br>.2<br>.35 .15<br>Percentage (%) Percentage (%)<br>.3 .1<br><!-- End of picture text --> Note: Estimates are obtained with synthetic panel data and weighted with population weights where the second survey round is used as the base year. Household heads' ages are restricted to 25-55 for the first survey round and adjusted accordingly for the second survey round.\\n\\n62\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [63]}}, {\"page\": 64, \"chunk\": 64, \"input_text\": \"# **Figure A.4. Probability of Other FHH Types Escaping Poverty in Second Year Conditional on Being Poor in First Year (percentage)**\\n\\n<!-- Start of picture text --> Egypt 2017-2020 Iraq 2007-2012<br>De-jure Employed fem. Core De-jure Employed fem. Core<br>Married Assets Married Assets<br>Jordan 2010-2013 Mauritania 2014-2019<br>De-jure Employed fem. Core De-jure Employed fem. Core<br>Married Assets Married Assets<br>Palestine 2011-2017 Tunisia 2015-2021<br>De-jure Employed fem. De-jure Employed fem.<br>Married Core Married Core<br>FHH wo/children FHH w/children Non-FHH<br>45 70<br>40<br>60<br>35<br>50<br>30<br>Percentage (%) Percentage (%)<br>25 40<br>70 50<br>48<br>60<br>46<br>44<br>50<br>42<br>Percentage (%) 40 Percentage (%)<br>40<br>50 25<br>40 20<br>30 15<br>Percentage (%) Percentage (%)<br>20 10<br><!-- End of picture text --> Note: Estimates are obtained with synthetic panel data and weighted with population weights where the second survey round is used as the base year. The figure shows the percentage of the population that moves out of poverty in the second year. FHH characteristics are measured in second period. Dashed red lines represent the national average for each period. Household heads' ages are restricted to 25-55 for the first survey round and adjusted accordingly for the second survey round. Standard errors are estimated with 1,000 bootstraps.\\n\\n63\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [64]}}, {\"page\": 65, \"chunk\": 65, \"input_text\": \"**Figure A.5. Probabilities of Female-Headed Households Escaping Poverty in Second Year Conditional on Being Poor in First Year (percentage), Arab Republic of Egypt**\\n\\n<!-- Start of picture text --> Egypt 2012-2015 Egypt 2015-2017 Egypt 2017-2020<br>FHH wo/children FHH w/children Non-FHH<br>Self-reported Females>0.5 Potential Educated fem. Self-reported Females>0.5 Potential Educated fem. Self-reported Females>0.5 Potential Educated fem.<br>40 40 40<br>30 30 30<br>Percentage (%) Percentage (%) Percentage (%)<br>20 20 20<br>10 10 10<br><!-- End of picture text --> Note: Estimates are obtained with synthetic panel data and weighted with population weights where the second survey round is used as the base year. The figure shows the percentage of the population that moves out of poverty in the second year. FHH characteristics are measured in second period. Dashed red lines represent the national average for each period. Household heads' ages are restricted to 25-55 for the first survey round and adjusted accordingly for the second survey round. Standard errors are estimated with 1,000 bootstraps.\\n\\n64\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [65]}}, {\"page\": 66, \"chunk\": 66, \"input_text\": \"**Figure A.6. Probabilities of Female-Headed Households Escaping Poverty in Second Year Conditional on Being Poor in First Year (percentage), Mauritania**\\n\\n<!-- Start of picture text --> Mauritania 2004-2008 Mauritania 2008-2014 Mauritania 2014-2019<br>FHH wo/children FHH w/children Non-FHH<br>Self-reported Females>0.5 Potential Educated fem. Self-reported Females>0.5 Potential Educated fem. Self-reported Females>0.5 Potential Educated fem.<br>50 50 50<br>45 45 45<br>40 40 40<br>Percentage (%) Percentage (%) Percentage (%)<br>35 35 35<br>30 30 30<br><!-- End of picture text --> Note: Estimates are obtained with synthetic panel data and weighted with population weights where the second survey round is used as the base year. The figure shows the percentage of the population that moves out of poverty in the second year. FHH characteristics are measured in second period. Dashed red lines represent the national average for each period. Household heads' ages are restricted to 25-55 for the first survey round and adjusted accordingly for the second survey round. Standard errors are estimated with 1,000 bootstraps.\\n\\n65\", \"datasets\": [{\"mention_name\": {\"text\": \"synthetic panel data\", \"confidence\": 0.8388901352882385, \"start\": 619, \"end\": 639, \"clean_text\": \"synthetic panel data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 619, \"end\": 639, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 619, \"end\": 639, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 619, \"end\": 639}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 619, \"end\": 639, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9220925569534302, \"start\": 619, \"end\": 639}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8388901352882385, \"start\": 619, \"end\": 639}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5956490635871887, \"start\": 619, \"end\": 639}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5956490635871887, \"start\": 619, \"end\": 639}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [66]}}, {\"page\": 67, \"chunk\": 67, \"input_text\": \"**Figure A.7. Probabilities of Female-Headed Households Escaping Poverty in Second Year Conditional on Being Poor in First Year (percentage), West Bank and Gaza**\\n\\n<!-- Start of picture text --> Palestine 2007-2009 Palestine 2009-2011 Palestine 2011-2017<br>FHH wo/children FHH w/children Non-FHH<br>Self-reported Females>0.5 Potential Educated fem. Self-reported Females>0.5 Potential Educated fem. Self-reported Females>0.5 Potential Educated fem.<br>45 45 45<br>40 40 40<br>35 35 35<br>Percentage (%) Percentage (%) Percentage (%)<br>30 30 30<br>25 25 25<br><!-- End of picture text --> Note: Estimates are obtained with synthetic panel data and weighted with population weights where the second survey round is used as the base year. The figure shows the percentage of the population that moves out of poverty in the second year. FHH characteristics are measured in second period. Dashed red lines represent the national average for each period. Household heads' ages are restricted to 25-55 for the first survey round and adjusted accordingly for the second survey round. Standard errors are estimated with 1,000 bootstraps.\\n\\n66\", \"datasets\": [{\"mention_name\": {\"text\": \"synthetic panel data\", \"confidence\": 0.8645969033241272, \"start\": 624, \"end\": 644, \"clean_text\": \"synthetic panel data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 624, \"end\": 644, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 624, \"end\": 644, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 624, \"end\": 644}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 624, \"end\": 644, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9220925569534302, \"start\": 624, \"end\": 644}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8645969033241272, \"start\": 624, \"end\": 644}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5956490635871887, \"start\": 624, \"end\": 644}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5956490635871887, \"start\": 624, \"end\": 644}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [67]}}, {\"page\": 68, \"chunk\": 68, \"input_text\": \"**Figure A.8. Probabilities of Female-Headed Households Escaping Poverty in Second Year Conditional on Being Poor in First Year (percentage), Tunisia**\\n\\n<!-- Start of picture text --> Tunisia 2005-2010 Tunisia 2010-2015 Tunisia 2015-2021<br>FHH wo/children FHH w/children Non-FHH<br>Self-reported Females>0.5 Potential Educated fem. Self-reported Females>0.5 Potential Educated fem. Self-reported Females>0.5 Potential Educated fem.<br>60 60 60<br>50 50 50<br>40 40 40<br>Percentage (%) Percentage (%) Percentage (%)<br>30 30 30<br>20 20 20<br>10 10 10<br><!-- End of picture text --> Note: Estimates are obtained with synthetic panel data and weighted with population weights where the second survey round is used as the base year. The figure shows the percentage of the population that moves out of poverty in the second year. FHH characteristics are measured in second period. Dashed red lines represent the national average for each period. Household heads' ages are restricted to 25-55 for the first survey round and adjusted accordingly for the second survey round. Standard errors are estimated with 1,000 bootstraps.\\n\\n67\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [68]}}, {\"page\": 69, \"chunk\": 69, \"input_text\": \"# **Figure A.9. Probabilities of Female-Headed Households Escaping Poverty in Second Year Conditional on Being Poor in First Year (percentage)**\\n\\n<!-- Start of picture text --> Egypt 2017-2020 Iraq 2007-2012<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>Jordan 2010-2013 Mauritania 2014-2019<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>Palestine 2011-2017 Tunisia 2015-2021<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>FHH Non-FHH<br>34<br>50<br>32 48<br>30 46<br>Percentage (%) Percentage (%)<br>28 44<br>46<br>56<br>54 44<br>52<br>42<br>50<br>40<br>48<br>Percentage (%) Percentage (%)<br>46 38<br>45<br>23<br>40 22<br>21<br>35<br>20<br>Percentage (%) Percentage (%)<br>30 19<br><!-- End of picture text --> Note: Estimates are obtained with synthetic panel data and weighted with population weights where the second survey round is used as the base year. The figure shows the percentage of the population that moves out of poverty in the second year. FHH characteristics are measured in second period. Dashed red lines represent the national average for each period. Household heads' ages are restricted to 25-55 for the first survey round and adjusted accordingly for the second survey round. Standard errors are estimated with 1,000 bootstraps.\\n\\n68\", \"datasets\": [{\"mention_name\": {\"text\": \"synthetic panel data\", \"confidence\": 0.852776825428009, \"start\": 935, \"end\": 955, \"clean_text\": \"synthetic panel data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 935, \"end\": 955, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 935, \"end\": 955, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 935, \"end\": 955}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 935, \"end\": 955, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9220925569534302, \"start\": 300, \"end\": 320}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.852776825428009, \"start\": 300, \"end\": 320}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5956490635871887, \"start\": 300, \"end\": 320}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5956490635871887, \"start\": 300, \"end\": 320}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [69]}}, {\"page\": 70, \"chunk\": 70, \"input_text\": \"# **Figure A.10. Probabilities of Female-Headed Households Falling in Poverty in Second Year Conditional on Being Non-poor in First Year (percentage)**\\n\\n<!-- Start of picture text --> Egypt 2017-2020 Iraq 2007-2012<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>Jordan 2010-2013 Mauritania 2014-2019<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>Palestine 2011-2017 Tunisia 2015-2021<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>FHH Non-FHH<br>17 14<br>16<br>13<br>15<br>12<br>14<br>11<br>13<br>Percentage (%) Percentage (%)<br>12 10<br>13 16<br>12 15<br>14<br>11<br>13<br>10<br>12<br>Percentage (%) 9 Percentage (%)<br>11<br>50 8.5<br>45 8<br>40 7.5<br>7<br>Percentage (%) Percentage (%)<br>35<br><!-- End of picture text --> Note: Estimates are obtained with synthetic panel data and weighted with population weights where the second survey round is used as the base year. The figure shows the percentage of the population that enters poverty in the second year. FHH characteristics are measured in second period. Dashed red lines represent the national average for each period. Household heads' ages are restricted to 25-55 for the first survey round and adjusted accordingly for the second survey round. Standard errors are estimated with 1,000 bootstraps.\\n\\n69\", \"datasets\": [{\"mention_name\": {\"text\": \"synthetic panel data\", \"confidence\": 0.8606802225112915, \"start\": 950, \"end\": 970, \"clean_text\": \"synthetic panel data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 950, \"end\": 970, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 950, \"end\": 970, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 950, \"end\": 970}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 950, \"end\": 970, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9220925569534302, \"start\": 308, \"end\": 328}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8606802225112915, \"start\": 308, \"end\": 328}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5956490635871887, \"start\": 308, \"end\": 328}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5956490635871887, \"start\": 308, \"end\": 328}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [70]}}, {\"page\": 71, \"chunk\": 71, \"input_text\": \"# **Appendix B: Additional Descriptive Statistics for Self-Reported FHHs and MHHs**\\n\\n**<u>Table B.1. Descriptive Statistics, Arab Republic of Egypt 2012-2020</u>**\\n\\n||||||**Egypt, A**<br>|**rab Rep.**<br>|||||\\n|---|---|---|---|---|---|---|---|---|---|---|\\n||**2**|**012**|**2**|**017**|**2**|**017**|**2**|**020**|**Pooled**|**2012-2020**|\\n||**MHH**<br>|**FHH**<br>|**MHH**<br>|**FHH**<br>|**MHH**<br>|**FHH**<br>|**MHH**<br>|**FHH**<br>|**MHH**<br>|**FHH**<br>|\\n|Heads age|47.17<br>(13.44)|54.11<br>(15.13)|48.71<br>(13.13)|56.05<br>(14.43)|49.99<br>(12.87)|57.19<br>(14.01)|47.01<br>(13.79)|55.49<br>(16.04)|48.37<br>(13.34)|55.91<br>(14.89)|\\n||0.40|0.69|0.36|0.63|0.34|0.63|0.31|0.58|0.35|0.63|\\n|Head does not complete primary school|(0.49)|(0.46)|(0.48)|(0.48)|(0.47)|(0.48)|(0.46)|(0.49)|(0.48)|(0.48)|\\n|'|0.13|0.09|0.15|0.12|0.15|0.11|0.14|0.11|0.15|0.11|\\n|Heads highest education level is primary|(0.33)|(0.29)|(0.36)|(0.32)|(0.36)|(0.31)|(0.35)|(0.31)|(0.35)|(0.31)|\\n|'|0.28|0.14|0.29|0.16|0.30|0.15|0.33|0.19|0.30|0.16|\\n|Heads highest education level is secondary|(0.45)|(0.35)|(0.45)|(0.36)|(0.46)|(0.36)|(0.47)|(0.39)|(0.46)|(0.37)|\\n|Hd' hih di ll i i|0.19|0.08|0.20|0.09|0.20|0.10|0.22|0.12|0.20|0.10|\\n|eas gest eucaton eve s tertary|(0.40)|(0.27)|(0.40)|(0.29)|(0.40)|(0.30)|(0.41)|(0.33)|(0.40)|(0.30)|\\n||0.02|0.03|0.01|0.02|0.01|0.03|0.01|0.03|0.01|0.03|\\n|Head is never married|(0.13)|(0.16)|(0.12)|(0.14)|(0.11)|(0.16)|(0.12)|(0.18)|(0.12)|(0.16)|\\n|Hd i id|<br>0.95|<br>0.21|<br>0.93|<br>0.16|<br>0.95|<br>0.14|<br>0.95|<br>0.20|<br>0.95|<br>0.17|\\n|ea s mono marre|(0.22)|(0.41)|(0.25)|(0.37)|(0.22)|(0.34)|(0.21)|(0.40)|(0.23)|(0.38)|\\n||0.01|0.00|0.02|0.00|0.00|0.00|0.00|0.00|0.01|0.00|\\n|Head is poly married|<br>(0.07)|<br>(0.00)|<br>(0.12)|<br>(0.00)|<br>(0.06)|<br>(0.00)|<br>(0.00)|<br>(0.00)|<br>(0.08)|<br>(0.00)|\\n||0.01|0.05|0.01|0.07|0.01|0.08|0.01|0.10|0.01|0.08|\\n|Head is divorced/separated|(0.08)|(0.22)|(0.09)|(0.26)|(0.09)|(0.28)|(0.09)|(0.30)|(0.09)|(0.27)|\\n|Hd i idd|0.02|0.71|0.03|0.74|0.03|0.75|0.02|0.66|0.03|0.72|\\n|ea s wowe|(0.15)|(0.46)|(0.17)|(0.44)|(0.16)|(0.43)|(0.15)|(0.47)|(0.16)|(0.45)|\\n|i l|0.86|0.19|0.83|0.20|0.81|0.19|0.85|0.21|0.84|0.20|\\n|Head s empoyed|(0.34)|(0.40)|(0.37)|(0.40)|(0.39)|(0.39)|(0.36)|(0.41)|(0.37)|(0.40)|\\n||0.01|0.00|0.01|0.00|0.01|0.01|0.01|0.02|0.01|0.01|\\n|Head is unemployed|(0.09)|(0.05)|(0.07)|(0.06)|(0.08)|(0.08)|(0.08)|(0.13)|(0.08)|(0.08)|\\n|Head is homemaker/housewife|0.00<br>|0.20<br>|0.00<br>|0.48<br>|0.00<br>|0.51<br>|0.00<br>|0.49<br>|0.00<br>|0.44<br>|\\n||(0.00)|(0.40)|(0.00)|(0.50)|(0.00)|(0.50)|(0.00)|(0.50)|(0.00)|(0.50)|\\n|i|<br>0.00|<br>0.00|<br>0.00|<br>0.00|<br>0.00|<br>0.00|<br>0.00|<br>0.01|<br>0.00|<br>0.00|\\n|Head s student|(0.05)|(0.00)|(0.04)|(0.04)|(0.01)|(0.04)|(0.03)|(0.07)|(0.03)|(0.05)|\\n||0.11|0.51|0.15|0.31|0.18|0.29|0.14|0.27|0.15|0.33|\\n|Head is pensioner/retired/disabled|(0.31)|(0.50)|(0.36)|(0.46)|(0.38)|(0.46)|(0.35)|(0.45)|(0.36)|(0.47)|\\n||0.02|0.09|0.01|0.00|0.00|0.00|0.00|0.00|0.01|0.02|\\n|Head is other activities|(0.13)|(0.29)|(0.08)|(0.06)|(0.07)|(0.02)|(0.06)|(0.03)|(0.08)|(0.13)|\\n||671896|840976|1022128|1311456|1435090|1935674|1674403|2153011|1248320|1627927|\\n|Per capita consumption|,.<br>(5,294)|,.<br>(6,273)|,.<br>(10,335)|,.<br>(12,830)|,.<br>(11,865)|,.<br>(13,604)|,.<br>(18,062)|,.<br>(16,090)|,.<br>(13,131)|,.<br>(14,000)|\\n|P i f|<br>1,201.68|<br>5,234.46|<br>2,281.21|<br>8,402.00|<br>3,864.17|<br>12,784.78|<br>4,060.60|<br>15,846.50|<br>3,007.98|<br>11,034.81|\\n|er capta transers|(3,334)|(5,406)|(5,429)|(9,957)|(8,474)|(11,940)|(8,253)|(15,900)|(7,046)|(12,376)|\\n|Hhld i|4.63|2.97|4.54|3.00|4.49|2.72|4.35|2.63|4.50|2.82|\\n|ouseo sze|(1.81)|(1.85)|(1.71)|(1.92)|(1.71)|(1.73)|(1.60)|(1.67)|(1.70)|(1.80)|\\n|Nb f hild 014|1.47|0.72|1.43|0.72|1.39|0.56|1.53|0.69|1.45|0.67|\\n|umer o cren age -|(1.35)|(1.14)|(1.39)|(1.21)|(1.40)|(1.04)|(1.38)|(1.18)|(1.38)|(1.15)|\\n||<br>0.20|<br>0.31|<br>0.21|<br>0.32|<br>0.25|<br>0.36|<br>0.21|<br>0.36|<br>0.22|<br>0.34|\\n|Number of seniors|(0.49)|(0.47)|(0.50)|(0.47)|(0.55)|(0.49)|(0.51)|(0.49)|(0.51)|(0.48)|\\n|1-2 adults no child|0.12|0.42|0.13|0.42|0.15|0.49|0.14|0.50|0.14|0.46|\\n|,|(0.32)|(0.49)|(0.34)|(0.49)|(0.36)|(0.50)|(0.35)|(0.50)|(0.34)|(0.50)|\\n|12 dl 12 hild|<br>0.23|<br>0.18|<br>0.19|<br>0.16|<br>0.16|<br>0.15|<br>0.25|<br>0.18|<br>0.20|<br>0.16|\\n|- auts, - cren|(0.42)|(0.38)|(0.39)|(0.37)|(0.36)|(0.35)|(0.43)|(0.38)|(0.40)|(0.37)|\\n||<br>0.23|<br>0.11|<br>0.25|<br>0.10|<br>0.26|<br>0.09|<br>0.25|<br>0.11|<br>0.25|<br>0.10|\\n|1-2 adult, 3 or more children|(0.42)|(0.31)|(0.43)|(0.30)|(0.44)|(0.29)|(0.43)|(0.31)|(0.43)|(0.30)|\\n||0.23|0.20|0.25|0.21|0.26|0.20|0.21|0.16|0.24|0.19|\\n|3 adults or more, 0-1 child|(0.42)|(0.40)|(0.43)|(0.41)|(0.44)|(0.40)|(0.41)|(0.36)|(0.43)|(0.40)|\\n||014|007|014|008|014|005|012|004|014|006|\\n|3 adults or more, 2-3 children|.<br>(0.35)|.<br>(0.25)|.<br>(0.34)|.<br>(0.28)|.<br>(0.35)|.<br>(0.23)|.<br>(0.33)|.<br>(0.21)|.<br>(0.34)|.<br>(0.24)|\\n|3 adults or more 4 children or more|<br>0.05|<br>0.02|<br>0.04|<br>0.02|<br>0.03|<br>0.01|<br>0.03|<br>0.01|<br>0.04|<br>0.02|\\n|,|(0.21)|(0.14)|(0.19)|(0.15)|(0.18)|(0.12)|(0.17)|(0.11)|(0.19)|(0.13)|\\n|Rl|<br>0.56|<br>0.55|<br>0.55|<br>0.52|<br>0.55|<br>0.49|<br>0.57|<br>0.54|<br>0.56|<br>0.52|\\n|ura area|(0.50)<br>|(0.50)<br>|(0.50)<br>|(0.50)<br>|(0.50)<br>|(0.50)<br>|(0.50)<br>|(0.50)<br>|(0.50)<br>|(0.50)<br>|\\n||0.44|0.45|0.45|0.48|0.45|0.51|0.43|0.46|0.44|0.48|\\n|Urban area|(0.50)|(0.50)|(0.50)|(0.50)|(0.50)|(0.50)|(0.50)|(0.50)|(0.50)|(0.50)|\\n\\nNote: Household sampling weights are applied. Standard deviations are in parentheses. FHHs and MHHs are self-reported.\\n\\n70\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [71]}}, {\"page\": 72, \"chunk\": 72, \"input_text\": \"# **<u>Table B.2. Descriptive Statistics, West Bank and Gaza, 2007-2017</u>**\\n\\n||**20**|**07**|**2**|**009**|**West Ban**<br>**2**|**k and Gaza**<br>**011**|**2**|**017**|**Pooled **|**2007-2017**|\\n|---|---|---|---|---|---|---|---|---|---|---|\\n||**MHH**|**FHH**|**MHH**|**FHH**|**MHH**|**FHH**|**MHH**|**FHH**|**MHH**|**FHH**|\\n||44.45|57.75|43.99|57.96|45.06|58.27|45.37|59.97|44.77|58.61|\\n|Heads age|<br>(13.29)|<br>(16.31)|<br>(13.21)|<br>(15.03)|<br>(12.95)|<br>(13.61)|<br>(12.96)|<br>(13.16)|<br>(13.08)|<br>(14.16)|\\n|Hd d l i hl|0.15|0.52|0.14|0.53|0.12|0.50|0.12|0.43|0.13|0.49|\\n|ea oes not compete prmary scoo|(0.35)|(0.50)|(0.35)|(0.50)|(0.33)|(0.50)|(0.32)|(0.50)|(0.34)|(0.50)|\\n|Hd' hiht dti ll i i|0.52|0.24|0.48|0.28|0.49|0.31|0.49|0.37|0.49|0.31|\\n|eas ges eucaon eve s prmary|(0.50)|(0.43)|(0.50)|(0.45)|(0.50)|(0.46)|(0.50)|(0.48)|(0.50)|(0.46)|\\n|Head's hihest education level is secondar|0.16|0.12|0.17|0.11|0.18|0.10|0.16|0.08|0.17|0.10|\\n|g y|(0.36)|(0.32)|(0.38)|(0.31)|(0.38)|(0.30)|(0.37)|(0.28)|(0.38)|(0.30)|\\n|'|0.17|0.11|0.21|0.08|0.21|0.09|0.23|0.11|0.21|0.10|\\n|Heads highest education level is tertiary|(0.38)|(0.32)|(0.41)|(0.28)|(0.41)|(0.29)|(0.42)|(0.31)|(0.41)|(0.30)|\\n|Hd i id|0.01|0.06|0.01|0.11|0.01|0.10|0.00|0.00|0.01|0.07|\\n|ea s never marre|(0.10)|(0.25)|(0.09)|(0.31)|(0.08)|(0.30)|(0.00)|(0.00)|(0.08)|(0.26)|\\n|Head is mono married|<br>0.96|<br>0.09|<br>0.97|<br>0.12|<br>0.97|<br>0.10|<br>0.99|<br>0.20|<br>0.97|<br>0.13|\\n||(0.19)|(0.29)|(0.18)|(0.33)|(0.17)|(0.30)|(0.10)|(0.40)|(0.16)|(0.34)|\\n|Hd i l id|0.02|0.00|0.01|0.00|0.01|0.00|0.00|0.00|0.01|0.00|\\n|ea s poy marre|(0.12)|(0.00)|(0.11)|(0.00)|(0.10)|(0.00)|(0.00)|(0.00)|(0.09)|(0.00)|\\n|Hd i did/d|0.00|0.11|0.00|0.06|0.00|0.10|0.01|0.72|0.00|0.26|\\n|ea s vorceseparate|(0.00)|(0.31)|(0.05)|(0.25)|(0.06)|(0.30)|(0.09)|(0.45)|(0.06)|(0.44)|\\n|Hd i idd|0.01|0.73|0.01|0.71|0.01|0.70|0.00|0.08|0.01|0.53|\\n|ea s wowe|(0.11)|(0.44)|(0.10)|(0.46)|(0.10)|(0.46)|(0.05)|(0.27)|(0.09)|(0.50)|\\n|Hd i ld|<br>0.76|<br>0.21|<br>0.76|<br>0.23|<br>0.78|<br>0.24|||<br>0.77|<br>0.23|\\n|ea s empoye|(0.43)|(0.41)|(0.43)|(0.42)|(0.42)|(0.43)|||(0.42)|(0.42)|\\n|Hd i ld|0.12|0.02|0.08|0.01|0.08|0.02|||0.08|0.02|\\n|ea s unempoye|(0.32)|(0.14)|(0.27)|(0.11)|(0.27)|(0.15)|||(0.28)|(0.13)|\\n|Hd i hk/hif|0.00|0.56|0.00|0.47|0.14|0.08|||0.06|0.29|\\n|ea s omemaerousewe|(0.05)|(0.50)|(0.05)|(0.50)|(0.34)|(0.28)|||(0.24)|(0.45)|\\n|Hd i tdt|0.00|0.00|0.00|0.00|0.00|0.00|||0.00|0.00|\\n|ea s suen|(0.06)|(0.00)|(0.03)|(0.06)|(0.04)|(0.04)|||(0.04)|(0.04)|\\n||0.08|0.18|0.10|0.27|0.00|0.00|||0.05|0.13|\\n|Head is pensioner/retired/disabled|(0.27)|(0.39)|(0.30)|(0.44)|(0.00)|(0.00)|||(0.22)|(0.33)|\\n|Hd i h iii|0.04|0.03|0.06|0.02|0.01|0.66|||0.03|0.33|\\n|ea s oter actvtes|(0.20)|(0.18)|(0.24)|(0.13)|(0.10)|(0.48)|||(0.18)|(0.47)|\\n||9,147.45|14,179.85|11,758.84|15,823.72|13,164.33|16,983.40|12,202.69|16,567.52|12,095.51|16,304.18|\\n|Per capita consumption|<br>(9,668)|<br>(13,757)|<br>(10,971)|<br>(12,594)|<br>(12,794)|<br>(13,025)|<br>(9,668)|<br>(13,757)|<br>(11,135)|<br>(12,622)|\\n|Hhld i|<br>6.62|<br>3.84|<br>6.31|<br>3.44|<br>6.27|<br>3.63|<br>5.76|<br>3.34|<br>6.17|<br>3.51|\\n|ouseo sze|(2.75)|(2.86)|(2.67)|(2.34)|(2.54)|(2.86)|(2.31)|(2.51)|(2.55)|(2.62)|\\n||2.69|0.99|2.42|0.71|2.38|0.81|2.31|0.70|2.40|0.77|\\n|Number of children age 0-14|(1.95)|(1.60)|(1.93)|(1.41)|(1.85)|(1.52)|(1.86)|(1.32)|(1.89)|(1.44)|\\n|Nb f i|<br>0.22|<br>0.40|<br>0.19|<br>0.42|<br>0.19|<br>0.42|<br>0.17|<br>0.41|<br>0.19|<br>0.42|\\n|umer o senors|(0.51)|(0.50)|(0.50)|(0.51)|(0.49)|(0.53)|(0.46)|(0.50)|(0.49)|(0.51)|\\n||0.07|0.37|0.07|0.42|0.07|0.44|||0.07|0.43|\\n|1-2 adults, no child|(0.25)|(0.49)|(0.25)|(0.49)|(0.25)|(0.50)|||(0.25)|(0.49)|\\n||0.09|0.15|0.13|0.08|0.10|0.08|||0.11|0.09|\\n|1-2 adults, 1-2 children|(0.29)|(0.36)|(0.34)|(0.28)|(0.31)|(0.27)|||(0.32)|(0.28)|\\n||0.37|0.12|0.33|0.11|0.34|0.09|||0.34|0.10|\\n|1-2 adult, 3 or more children|(0.48)|(0.32)|(0.47)|(0.32)|(0.47)|(0.29)|||(0.47)|(0.31)|\\n|dl 1 hild|<br>0.13|<br>0.22|<br>0.14|<br>0.26|<br>0.17|<br>0.22|||<br>0.15|<br>0.23|\\n|3 auts or more, 0- c|(0.33)|(0.42)|(0.35)|(0.44)|(0.37)|(0.42)|||(0.36)|(0.42)|\\n||<br>0.15|<br>0.07|<br>0.15|<br>0.08|<br>0.14|<br>0.09|||<br>0.15|<br>0.09|\\n|3 adults or more, 2-3 children|(0.36)|(0.25)|(0.36)|(0.28)|(0.35)|(0.29)|||(0.35)|(0.28)|\\n||0.20|0.08|0.18|0.04|0.18|0.07|||0.18|0.06|\\n|3 adults or more, 4 children or more|<br>(0.40)|<br>(0.27)|<br>(0.38)|<br>(0.20)|<br>(0.38)|<br>(0.26)|||<br>(0.38)|<br>(0.24)|\\n|Rral area|<br>0.29|<br>0.26|<br>0.17|<br>0.22|<br>0.17|<br>0.18|0.18|0.17|<br>0.18|<br>0.20|\\n|u|(0.45)<br>|(0.44)<br>|(0.37)<br>|(0.42)<br>|(0.38)<br>|(0.39)<br>|(0.38)<br>|(0.37)<br>|(0.39)<br>|(0.40)<br>|\\n||0.57|0.60|0.74|0.68|0.74|0.72|0.73|0.72|0.72|0.70|\\n|Urban area|(0.50)|(0.49)|(0.44)|(0.47)|(0.44)|(0.45)|(0.44)|(0.45)|(0.45)|(0.46)|\\n\\nNote: Household sampling weights are applied. Standard deviations are in parentheses. FHHs and MHHs are self-reported.\\n\\n71\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [72]}}, {\"page\": 73, \"chunk\": 73, \"input_text\": \"**<u>Table B.3. Descriptive Statistics, Tunisia 2005-2021</u>**\\n\\n||**2**|**005**|**2**|**010**|**Tu**<br>**20**|**nisia**<br>**15**|**2**|**021**|**Pooled **|**2005-2021**|\\n|---|---|---|---|---|---|---|---|---|---|---|\\n||**MHH**|**FHH**|**MHH**|**FHH**|**MHH**|**FHH**|**MHH**|**FHH**|**MHH**|**FHH**|\\n|Heads ae|52.12|57.74|53.44|59.52|53.38|60.99|55.40|61.95|53.72|60.31|\\n|g|(14.07)|(15.18)|(13.74)|(15.60)|(13.65)|(14.86)|(13.53)|(14.47)|(13.78)|(15.04)|\\n|Head does not complete|0.75|0.92|0.77|0.93|0.19|0.57|0.13|0.48|0.42|0.68|\\n|primary school|(0.43)|(0.27)|(0.42)|(0.26)|(0.39)|(0.50)|(0.34)|(0.50)|(0.49)|(0.47)|\\n|Head's highest education level|0.04|0.01|0.03|0.02|0.39|0.25|0.42|0.30|0.25|0.17|\\n|is primary|(0.20)|(0.12)|(0.18)|(0.12)|(0.49)|(0.43)|(0.49)|(0.46)|(0.43)|(0.38)|\\n|Head's highest education level|0.12|0.05|0.11|0.04|0.29|0.13|0.32|0.16|0.22|0.10|\\n|is secondary|(0.32)|(0.21)|(0.31)|(0.20)|(0.46)|(0.34)|(0.47)|(0.36)|(0.42)|(0.31)|\\n|Head's highest education level|0.09|0.02|0.09|0.02|0.13|0.05|0.13|0.06|0.11|0.04|\\n|is tertiary|(0.28)|(0.14)|(0.29)|(0.14)|(0.33)|(0.22)|(0.33)|(0.23)|(0.31)|(0.20)|\\n|<br>|<br>0.02|<br>0.06|<br>0.01|<br>0.04|<br>0.03|<br>0.08|<br>0.03|<br>0.08|<br>0.02|<br>0.07|\\n|Head is never married|(0.13)|(0.24)|(0.11)|(0.19)|(0.17)|(0.27)|(0.17)|(0.27)|(0.15)|(0.25)|\\n|Head is mono married|0.96|0.19|0.97|0.20|0.95|0.13|0.95|0.10|0.95|0.15|\\n||(0.20)|(0.39)|(0.18)|(0.40)|(0.22)|(0.34)|(0.22)|(0.30)|(0.21)|(0.35)|\\n||0.01|0.08|0.00|0.08|0.01|0.09|0.01|0.10|0.00|0.09|\\n|Head is divorced/separated|(0.07)|(0.26)|(0.06)|(0.27)|(0.07)|(0.29)|(0.07)|(0.30)|(0.07)|(0.29)|\\n|Head is widowed|0.02|0.67|0.02|0.69|0.02|0.70|0.02|0.71|0.02|0.69|\\n||(0.14)|(0.47)|(0.14)|(0.46)|(0.13)|(0.46)|(0.13)|(0.45)|(0.13)|(0.46)|\\n|Hd i ld|<br>0.74|<br>0.24|<br>0.71|<br>0.17|||||<br>0.72|<br>0.21|\\n|ea s empoye|(0.44)|(0.43)|(0.46)|(0.38)|||||(0.45)|(0.40)|\\n|Hd i ld|0.02|0.01|0.02|0.01|||||0.02|0.01|\\n|ea s unempoye|(0.15)|(0.10)|(0.15)|(0.07)|||||(0.15)|(0.09)|\\n|Head is|<br>0.00|<br>0.41|<br>0.01|<br>0.49|||||<br>0.00|<br>0.45|\\n|homemaker/housewife|(0.05)|(0.49)|(0.07)|(0.50)|||||(0.06)|(0.50)|\\n|Head is student|0.00|0.01|0.00|0.00|||||0.00|0.00|\\n||(0.04)|(0.09)|(0.02)|(0.02)|||||(0.03)|(0.06)|\\n|Head is|<br>0.22|<br>0.32|<br>0.26|<br>0.33|||||<br>0.24|<br>0.32|\\n|pensioner/retired/disabled|(0.42)|(0.46)|(0.44)|(0.47)|||||(0.43)|(0.47)|\\n||000|002|000|001|||||000|001|\\n|Head is other activities|.<br>|.<br>|.<br>|.<br>|||||.<br>|.<br>|\\n||(0.07)|(0.15)|(0.05)|(0.08)|||||(0.06)|(0.12)|\\n|P it ti|<br>2,027.76|<br>2,252.64|<br>2,823.94|<br>3,156.40|4,310.95|4,892.92|5,999.25|6,886.89|<br>3,976.12|<br>4,634.09|\\n|er capa consumpon|(2,206.66)|(2,473.81)|(2,604.83)|(2,640.10)|(4,590.12)|(4,002.85)|(7,378.56)|(6,515.51)|(5,136.62)|(4,910.49)|\\n|Household size|4.77<br>|3.25<br>|4.55<br>|3.11<br>|4.26<br>|2.79<br>|3.97<br>|2.61<br>|4.35<br>|2.89<br>|\\n||(1.88)|(1.96)|(1.73)|(1.75)|(1.60)|(1.60)|(1.45)|(1.39)|(1.68)|(1.67)|\\n|Nb f hild 014|1.21|0.57|1.01|0.50|1.08|0.40|0.95|0.36|1.05|0.44|\\n|umer o cren age -|(1.26)|(1.07)|(1.19)|(0.99)|(1.23)|(0.86)|(1.18)|(0.83)|(1.21)|(0.93)|\\n||0.38|0.43|0.38|0.46|0.35|0.47|||0.37|0.45|\\n|Number of seniors|(0.66)|(0.52)|(0.66)|(0.54)|(0.66)|(0.53)|||(0.66)|(0.53)|\\n|12 dlt hild|0.10|0.38|0.11|0.40|||0.19|0.52|0.14|0.45|\\n|- aus, no c|(0.31)<br>|(0.49)<br>|(0.32)<br>|(0.49)<br>|||(0.39)<br>|(0.50)<br>|(0.35)<br>|(0.50)<br>|\\n||0.20|0.12|0.20|0.13|||0.18|0.09|0.19|0.11|\\n|1-2 adults, 1-2 children|(0.40)|(0.32)|(0.40)|(0.33)|||(0.38)|(0.29)|(0.39)|(0.31)|\\n|1-2 adult 3 or more children|0.17|0.08|0.14|0.08|||0.10|0.03|0.13|0.06|\\n|,|(0.37)|(0.28)|(0.35)|(0.27)|||(0.30)|(0.17)|(0.34)|(0.23)|\\n||<br>0.32|<br>0.32|<br>0.39|<br>0.33|||<br>0.44|<br>0.32|<br>0.39|<br>0.32|\\n|3 adults or more, 0-1 child|(0.47)|(0.47)|(0.49)|(0.47)|||(0.50)|(0.47)|(0.49)|(0.47)|\\n|3 adults or more 2-3 children|0.16<br>|0.08<br>|0.14<br>|0.06<br>|||0.09<br>|0.04<br>|0.13<br>|0.06<br>|\\n|,|(0.37)|(0.27)|(0.35)|(0.24)|||(0.28)|(0.19)|(0.33)|(0.23)|\\n|3 adults or more, 4 children or|<br>0.04|<br>0.02|<br>0.02|<br>0.01|||<br>0.01|<br>0.00|<br>0.02|<br>0.01|\\n|more|(0.20)|(0.14)|(0.15)|(0.08)|||(0.07)|(0.04)|(0.14)|(0.09)|\\n||0.32|0.33|0.32|0.34|0.30|0.26|0.31|0.27|0.31|0.29|\\n|Rural area|(0.46)|(0.47)|(0.47)|(0.47)|(0.46)|(0.44)|(0.46)|(0.44)|(0.46)|(0.45)|\\n|Ub|0.68|0.67|0.68|0.66|0.70|0.74|0.69|0.73|0.69|0.71|\\n|ran area|(0.46)|(0.47)|(0.47)|(0.47)|(0.46)|(0.44)|(0.46)|(0.44)|(0.46)|(0.45)|\\n\\nNote: Household sampling weights are applied. Standard deviations are in parentheses. FHHs and MHHs are self-reported.\\n\\n72\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [73]}}, {\"page\": 74, \"chunk\": 74, \"input_text\": \"# **<u>Table B.4. Descriptive Statistics, Jordan 2010-2013</u>**\\n\\n||**20**|**10**|**Jo**<br>**2**|**rdan**<br>**013**|**Pooled **|**2010-2013**|\\n|---|---|---|---|---|---|---|\\n||**MHH**|**FHH**|**MHH**|**FHH**|**MHH**|**FHH**|\\n||47.62|58.63|47.49|58.96|47.55|58.80|\\n|Heads age|(14.13)|(13.68)|(14.04)|(13.24)|(14.08)|(13.45)|\\n|Hd d t lt i hl|0.15|0.53|0.12|0.49|0.13|0.51|\\n|ea oes no compee prmary scoo|(0.36)|(0.50)|(0.32)|(0.50)|(0.34)|(0.50)|\\n|'|0.47|0.27|0.47|0.30|0.47|0.28|\\n|Heads highest education level is primary|(0.50)|(0.44)|(0.50)|(0.46)|(0.50)|(0.45)|\\n|Head's hihest education level is secondar|0.15|0.08|0.14|0.09|0.15|0.09|\\n|g y|(0.36)|(0.28)|(0.34)|(0.28)|(0.35)|(0.28)|\\n|'|0.22|0.12|0.28|0.12|0.25|0.12|\\n|Heads highest education level is tertiary|(0.42)|(0.33)|(0.45)|(0.33)|(0.44)|(0.33)|\\n|Hd i id|0.02|0.06|0.02|0.07|0.02|0.07|\\n|ea s never marre|(0.13)|(0.25)|(0.14)|(0.25)|(0.14)|(0.25)|\\n|Head is mono married|<br>0.97|<br>0.19|<br>0.96|<br>0.17|<br>0.97|<br>0.18|\\n||(0.17)|(0.39)|(0.20)|(0.37)|(0.18)|(0.38)|\\n|Hd i l id|0.00|0.00|0.00|0.00|0.00|0.00|\\n|ea s poy marre|(0.06)|(0.00)|(0.06)|(0.00)|(0.06)|(0.00)|\\n|Hd i did/td|0.00|0.04|0.00|0.05|0.00|0.04|\\n|ea s vorceseparae|(0.03)|(0.19)|(0.06)|(0.22)|(0.05)|(0.20)|\\n|Hd i idd|0.01|0.71|0.01|0.72|0.01|0.72|\\n|ea s wowe|(0.08)|(0.45)|(0.11)|(0.45)|(0.10)|(0.45)|\\n|Hd i ld|0.66|0.04|0.67|0.05|0.66|0.05|\\n|ea s empoye|(0.47)|(0.20)|(0.47)|(0.22)|(0.47)|(0.21)|\\n|Hd i ld|0.04|0.01|0.07|0.04|0.06|0.02|\\n|ea s unempoye|(0.20)|(0.09)|(0.26)|(0.19)|(0.24)|(0.15)|\\n|Hd i hk/hif|0.00|0.79|0.00|0.75|0.00|0.77|\\n|ea s omemaerousewe|(0.02)|(0.41)|(0.00)|(0.43)|(0.01)|(0.42)|\\n|Head is student|0.00|0.00|0.00|0.00|0.00|0.00|\\n||(0.01)<br>|(0.00)<br>|(0.05)<br>|(0.04)<br>|(0.04)<br>|(0.03)<br>|\\n||0.12|0.15|0.14|0.14|0.13|0.14|\\n|Head is pensioner/retired/disabled|(0.33)|(0.35)|(0.35)|(0.34)|(0.34)|(0.35)|\\n||0.17|0.02|0.12|0.02|0.14|0.02|\\n|Head is other activities|(0.38)|(0.13)|(0.32)|(0.15)|(0.35)|(0.14)|\\n||206345|291621|223836|310082|215559|301097|\\n|Per capita consumption|,.<br>(3,263.14)|,.<br>(2,417.35)|,.<br>(1,578.64)|,.<br>(2,442.96)|,.<br>(2,521.48)|,.<br>(2,431.02)|\\n|P i f|<br>377.77|<br>1,107.40|<br>469.79|<br>1,114.52|<br>426.24|<br>1,111.05|\\n|er capta transers|(703.55)|(1,489.48)|(839.28)|(1,364.42)|(779.30)|(1,425.88)|\\n|Hhld i|5.66|3.64|5.36|3.49|5.50|3.56|\\n|ouseo sze|(2.18)|(2.28)|(2.12)|(2.51)|(2.15)|(2.40)|\\n||<br>1.93|<br>0.59|<br>1.80|<br>0.57|<br>1.86|<br>0.58|\\n|Number of children age 0-14|(1.70)|(1.14)|(1.66)|(1.42)|(1.68)|(1.29)|\\n||<br>0.26|<br>0.38|<br>0.23|<br>0.41|<br>0.24|<br>0.40|\\n|Number of seniors|(0.57)|(0.49)|(0.54)|(0.51)|(0.55)|(0.50)|\\n|1-2 adults no child|0.07<br>|0.36<br>|0.09<br>|0.39<br>|0.08<br>|0.37<br>|\\n|,|(0.26)|(0.48)|(0.28)|(0.49)|(0.27)|(0.48)|\\n|1-2 adlts 1-2 children|<br>0.13|<br>0.14|<br>0.15|<br>0.10|<br>0.14|<br>0.12|\\n|u,|(0.34)<br>|(0.35)<br>|(0.35)<br>|(0.30)<br>|(0.35)<br>|(0.32)<br>|\\n||0.30|0.06|0.29|0.06|0.29|0.06|\\n|1-2 adult, 3 or more children|(046)|(025)|(045)|(023)|(046)|(024)|\\n||.<br>0.23|.<br>0.29|.<br>0.25|.<br>0.34|.<br>0.24|.<br>0.32|\\n|3 adults or more, 0-1 child|(0.42)|(0.46)|(0.43)|(0.47)|(0.43)|(0.47)|\\n||015|010|013|008|014|009|\\n|3 adults or more, 2-3 children|.<br>(0.36)|.<br>(0.30)|.<br>(0.34)|.<br>(0.27)|.<br>(0.35)|.<br>(0.29)|\\n|3 adults or more 4 children or more|<br>0.11|<br>0.05|<br>0.09|<br>0.04|<br>0.10|<br>0.04|\\n|,|(0.31)|(0.22)|(0.29)|(0.18)|(0.30)|(0.20)|\\n|Rl|<br>0.16|<br>0.13|<br>0.17|<br>0.19|<br>0.17|<br>0.16|\\n|ura area|(0.37)<br>|(0.34)<br>|(0.37)<br>|(0.39)<br>|(0.37)<br>|(0.37)<br>|\\n||0.84|0.87|0.83|0.81|0.83|0.84|\\n|Urban area|(0.37)|(0.34)|(0.37)|(0.39)|(0.37)|(0.37)|\\n\\nNote: Household sampling weights are applied. Standard deviations are in parentheses. FHHs and MHHs are self-reported.\\n\\n73\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [74]}}, {\"page\": 75, \"chunk\": 75, \"input_text\": \"# **<u>Table B.5. Descriptive Statistics, Iraq 2007-2013</u>**\\n\\n||**20**|**07**|**Ir**<br>**2**|**aq**<br>**012**|**Pooled **|**2007-2013**|\\n|---|---|---|---|---|---|---|\\n||**MHH**|**FHH**|**MHH**|**FHH**|**MHH**|**FHH**|\\n||45.15|54.29|46.93|54.28|46.73|54.28|\\n|Heads age|(13.71)|(13.04)|(13.06)|(12.51)|(13.14)|(12.58)|\\n|Hd d t lt i hl|0.31|0.75|0.34|0.72|0.34|0.72|\\n|ea oes no compee prmary scoo|(0.46)|(0.43)|(0.47)|(0.45)|(0.47)|(0.45)|\\n|'|0.41|0.17|0.41|0.20|0.41|0.20|\\n|Heads highest education level is primary|(0.49)|(0.38)|(0.49)|(0.40)|(0.49)|(0.40)|\\n|Hd' hih di ll i d|0.11|0.02|0.09|0.03|0.09|0.03|\\n|eas gest eucaton eve s seconary|(0.31)|(0.14)|(0.28)|(0.17)|(0.28)|(0.17)|\\n|Hd' hiht dti ll i tti|0.17|0.06|0.16|0.05|0.16|0.05|\\n|eas ges eucaon eve s erary|(0.38)|(0.23)|(0.36)|(0.22)|(0.37)|(0.22)|\\n|Head i neer married|0.02|0.04|0.01|0.02|0.01|0.02|\\n|s v|(0.13)|(0.20)|(0.09)|(0.14)|(0.10)|(0.15)|\\n|Head is mono married|<br>0.95|<br>0.08|<br>0.95|<br>0.15|<br>0.95|<br>0.14|\\n||(0.21)|(0.27)|(0.22)|(0.35)|(0.22)|(0.34)|\\n|Hd i l id|0.01|0.00|0.03|0.00|0.03|0.00|\\n|ea s poy marre|(0.11)|(0.00)|(0.17)|(0.00)|(0.16)|(0.00)|\\n|Hd i did/td|0.00|0.07|0.00|0.06|0.00|0.06|\\n|ea s vorceseparae|(0.04)|(0.25)|(0.04)|(0.25)|(0.04)|(0.25)|\\n|Hd i idd|0.01|0.81|0.01|0.77|0.01|0.77|\\n|ea s wowe|(0.12)|(0.39)|(0.10)|(0.42)|(0.11)|(0.42)|\\n|Head is emloed|<br>0.79|<br>0.19|<br>0.78|<br>0.15|<br>0.78|<br>0.16|\\n|py|(0.41)|(0.40)|(0.42)|(0.36)|(0.42)|(0.36)|\\n|Head is unemloed|0.03|0.02|0.02|0.01|0.02|0.01|\\n|py|(0.18)|(0.13)|(0.15)|(0.08)|(0.15)|(0.09)|\\n|Hd i hk/hif|0.00|0.69|0.00|0.53|0.00|0.55|\\n|ea s omemaerousewe|(0.02)|(0.46)|(0.03)|(0.50)|(0.03)|(0.50)|\\n|Hd i tdt|0.00|0.00|0.00|0.00|0.00|0.00|\\n|ea s suen|(0.04)|(0.00)|(0.04)|(0.02)|(0.04)|(0.02)|\\n|i iiil|<br>0.14|<br>0.08|<br>0.18|<br>0.30|<br>0.18|<br>0.27|\\n|Head s pensoner/retred/dsabed|(0.35)|(0.27)|(0.39)|(0.46)|(0.38)|(0.45)|\\n||0.03|0.03|0.02|0.01|0.02|0.01|\\n|Head is other activities|(0.17)|(0.16)|(0.12)|(0.08)|(0.13)|(0.10)|\\n||187883975|201149891|285561371|307240417|274644313|293369042|\\n|Per capita consumption|,,.<br>(1,662,430)|,,.<br>(1,562,966)|,,.<br>(2,612,256)|,,.<br>(2,728,793)|,,.<br>(2,542,574)|,,.<br>(2,630,373)|\\n|P it tf|166.71|409.56|249.20|647.95|239.98|616.79|\\n|er capa ransers|(415.53)|(941.59)|(902.11)|(1,940.46)|(861.86)|(1,842.51)|\\n|Hhld i|7.00|5.76|8.49|7.64|8.33|7.39|\\n|ouseo sze|(3.43)|(3.46)|(4.21)|(4.22)|(4.16)|(4.18)|\\n|Nb f hild 014|2.69|1.64|3.46|2.75|3.37|2.61|\\n|umer o cren age -|(2.12)|(2.00)|(2.55)|(2.68)|(2.51)|(2.63)|\\n||0.22|0.27|0.24|0.24|0.24|0.25|\\n|Number of seniors|(0.51)|(0.49)|(0.52)|(0.46)|(0.52)|(0.46)|\\n|1-2 adults no child|0.04|0.14|0.01|0.04|0.01|0.05|\\n|,|(0.20)|(0.35)|(0.11)|(0.19)|(0.12)|(0.22)|\\n|12 dlt 12 hild|<br>0.14|<br>0.08|<br>0.06|<br>0.05|<br>0.07|<br>0.06|\\n|- aus, - cren|(0.34)|(0.28)|(0.25)|(0.23)|(0.26)|(0.23)|\\n||<br>0.29|<br>0.13|<br>0.27|<br>0.15|<br>0.27|<br>0.15|\\n|1-2 adult, 3 or more children|(0.45)|(0.34)|(0.44)|(0.35)|(0.44)|(0.35)|\\n||0.15|0.30|0.11|0.21|0.11|0.22|\\n|3 adults or more, 0-1 child|(0.35)|(0.46)|(0.31)|(0.41)|(0.32)|(0.42)|\\n||016|018|018|026|018|025|\\n|3 adults or more, 2-3 children|.<br>(0.37)|.<br>(0.38)|.<br>(0.39)|.<br>(0.44)|.<br>(0.39)|.<br>(0.43)|\\n|3 adults or more 4 children or more|<br>0.22|<br>0.16|<br>0.36|<br>0.29|<br>0.35|<br>0.27|\\n|,|(0.42)|(0.37)|(0.48)|(0.45)|(0.48)|(0.45)|\\n|Rl|<br>0.27|<br>0.20|<br>0.33|<br>0.22|<br>0.32|<br>0.22|\\n|ura area|(0.44)<br>|(0.40)<br>|(0.47)<br>|(0.42)<br>|(0.47)<br>|(0.41)<br>|\\n||0.73|0.80|0.67|0.78|0.68|0.78|\\n|Urban area|(0.44)|(0.40)|(0.47)|(0.42)|(0.47)|(0.41)|\\n\\nNote: Household sampling weights are applied. Standard deviations are in parentheses. FHHs and MHHs are self-reported.\\n\\n74\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [75]}}, {\"page\": 76, \"chunk\": 76, \"input_text\": \"# **<u>Table B.6. Descriptive Statistics, Mauritania 2004-2019</u>**\\n\\n||||||**Maurita**|**nia**|||||\\n|---|---|---|---|---|---|---|---|---|---|---|\\n||**20**|**04 **|**20**|**08**|**20**|**14**|**2**|**019**|**Pooled **|**2004-2019**|\\n||**MHH**|**FHH**|**MHH**|**FHH**|**MHH**|**FHH**|**MHH**|**FHH**|**MHH**|**FHH**|\\n|Hd|48.10|53.29|48.22|47.33|49.11|48.05|49.63|46.83|48.83|48.01|\\n|eas age|(13.60)|(14.09)|(14.06)|(15.24)|(14.49)|(15.80)|(14.29)|(15.25)|(14.16)|(15.40)|\\n|Head does not complete primary school|0.77<br>|0.93<br>|0.72<br>|0.85<br>|0.70<br>|0.81<br>|0.48<br>|0.50<br>|0.66<br>|0.72<br>|\\n||(0.42)|(0.25)|(0.45)|(0.35)|(0.46)|(0.39)|(0.50)|(0.50)|(0.47)|(0.45)|\\n|'|0.08|0.03|0.11|0.09|0.11|0.11|0.26|0.33|0.15|0.18|\\n|Heads highest education level is primary|(0.28)|(0.17)|(0.31)|(0.29)|(0.32)|(0.32)|(0.44)|(0.47)|(0.35)|(0.38)|\\n|d' hih di ll i d|0.11|0.03|0.12|0.05|0.13|0.07|0.18|0.16|0.14|0.09|\\n|Heas gest eucaton eve s seconary|(0.31)|(0.18)|(0.33)|(0.21)|(0.34)|(0.25)|(0.38)|(0.36)|(0.34)|(0.29)|\\n|Hd' hih di ll i i|<br>0.04|<br>0.01|<br>0.05|<br>0.01|<br>0.06|<br>0.01|<br>0.08|<br>0.01|<br>0.06|<br>0.01|\\n|eas gest eucaton eve s tertary|(0.19)|(0.08)|(0.22)|(0.08)|(0.24)|(0.08)|(0.27)|(0.10)|(0.23)|(0.09)|\\n|Hd i id|<br>0.03|<br>0.02|<br>0.02|<br>0.01|<br>0.03|<br>0.02|<br>0.03|<br>0.02|<br>0.03|<br>0.02|\\n|ea s never marre|(0.17)|(0.15)|(0.16)|(0.12)|(0.17)|(0.13)|(0.17)|(0.13)|(0.16)|(0.13)|\\n|Head is mono married|0.94|0.10|0.95|0.39|0.94|0.43|0.88|0.50|0.93|0.41|\\n||(0.24)|(0.30)|(0.22)|(0.49)|(0.23)|(0.50)|(0.32)|(0.50)|(0.26)|(0.49)|\\n|Hd i l id|0.00|0.00|0.00|0.00|0.00|0.00|0.06|0.03|0.02|0.01|\\n|ea s poy marre|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.23)|(0.17)|(0.12)|(0.11)|\\n|Hd i did/d|<br>0.02|<br>0.33|<br>0.01|<br>0.22|<br>0.02|<br>0.21|<br>0.02|<br>0.18|<br>0.02|<br>0.21|\\n|ea s vorceseparate|(0.14)|(0.47)|(0.12)|(0.42)|(0.13)|(0.40)|(0.13)|(0.39)|(0.13)|(0.41)|\\n|Hd i idd|0.01|0.55|0.01|0.37|0.01|0.35|0.01|0.27|0.01|0.35|\\n|ea s wowe|(0.11)|(0.50)|(0.11)|(0.48)|(0.11)|(0.48)|(0.12)|(0.44)|(0.11)|(0.48)|\\n|Hd i ld|0.86|0.54|0.78|0.40|0.87|0.40|0.81|0.37|0.83|0.41|\\n|ea s empoye|(0.35)|(0.50)|(0.41)|(0.49)|(0.33)|(0.49)|(0.39)|(0.48)|(0.37)|(0.49)|\\n||0.02|0.02|0.04|0.02|0.01|0.02|0.01|0.01|0.02|0.02|\\n|Head is unemployed|(0.15)|(0.12)|(0.19)|(0.13)|(0.08)|(0.13)|(0.11)|(0.11)|(0.14)|(0.12)|\\n|Head is not searched and not work|0.12<br>|0.45<br>|0.18<br>|0.59<br>|0.12<br>|0.58<br>|0.18<br>|0.61<br>|0.15<br>|0.58<br>|\\n||(0.32)|(0.50)|(0.38)|(0.49)|(0.32)|(0.49)|(0.38)|(0.49)|(0.36)|(0.49)|\\n|i i|<br>184,984.25|<br>147,570.28|<br>240,551.50|<br>219,258.68|<br>341,658.92|<br>350,431.69|<br>373,240.01|<br>403,696.14|<br>294,662.55|<br>315,850.25|\\n|Per capta consumpton|(2587509)|(205663)|(234972)|(168764)|(299510)|(247036)|(264805)|(268529)|(1,189,714)|(252,048)|\\n|Hhld i|5.94|4.40|5.87|4.68|6.05|5.03|6.46|5.59|6.10|5.08|\\n|ouseo sze|(2.77)|(2.44)|(2.88)|(2.47)|(3.44)|(2.61)|(3.59)|(2.98)|(3.24)|(2.73)|\\n|Number of children age 0-14|2.56<br>|1.63<br>|2.53<br>|2.12<br>|2.68<br>|2.33<br>|2.86<br>|2.66<br>|2.67<br>|2.32<br>|\\n||(2.01)|(1.73)|(2.03)|(1.84)|(2.29)|(1.98)|(2.41)|(2.01)|(2.21)|(1.96)|\\n||0.19|0.27|0.21|0.20|0.25|0.23|0.27|0.23|0.23|0.23|\\n|Number of seniors|(0.44)|(0.46)|(0.47)|(0.41)|(0.51)|(0.44)|(0.56)|(0.46)|(0.50)|(0.44)|\\n|12 dlt hild|0.08|0.19|0.10|0.13|0.10|0.11|0.09|0.07|0.09|0.11|\\n|- aus, no c|(0.28)|(0.39)|(0.29)|(0.33)|(0.29)|(0.31)|(0.28)|(0.25)|(0.29)|(0.31)|\\n|1-2 adults 1-2 children|0.16|0.21|0.16|0.22|0.14|0.19|0.12|0.19|0.14|0.20|\\n|,|(0.37)|(0.40)|(0.36)|(0.41)|(0.35)|(0.40)|(0.33)|(0.39)|(0.35)|(0.40)|\\n||<br>0.20|<br>0.14|<br>0.20|<br>0.24|<br>0.23|<br>0.26|<br>0.20|<br>0.26|<br>0.21|<br>0.24|\\n|1-2 adult, 3 or more children|(0.40)|(0.35)|(0.40)|(0.43)|(0.42)|(0.44)|(0.40)|(0.44)|(0.41)|(0.43)|\\n|dl 1 hild|<br>0.19|<br>0.26|<br>0.19|<br>0.20|<br>0.18|<br>0.20|<br>0.18|<br>0.17|<br>0.18|<br>0.19|\\n|3 auts or more, 0- c|(0.39)|(0.44)|(0.39)|(0.40)|(0.38)|(0.40)|(0.38)|(0.37)|(0.39)|(0.40)|\\n||017|013|018|013|017|014|019|017|018|015|\\n|3 adults or more, 2-3 children|.<br>(0.38)|.<br>(0.34)|.<br>(0.38)|.<br>(0.34)|.<br>(0.38)|.<br>(0.34)|.<br>(0.39)|.<br>(0.37)|.<br>(0.38)|.<br>(0.35)|\\n||<br>0.19|<br>0.07|<br>0.18|<br>0.09|<br>0.18|<br>0.11|<br>0.23|<br>0.15|<br>0.20|<br>0.12|\\n|3 adults or more, 4 children or more|(0.39)|(0.25)|(0.38)|(0.28)|(0.39)|(0.31)|(0.42)|(0.36)|(0.40)|(0.32)|\\n||<br>0.62|<br>0.57|<br>0.54|<br>0.64|<br>0.49|<br>0.56|<br>0.52|<br>0.57|<br>0.54|<br>0.58|\\n|Rural area|||||||||||\\n||(0.49)|(0.49)|(0.50)|(0.48)|(0.50)|(0.50)|(0.50)|(0.50)|(0.50)|(0.49)|\\n|Ub|0.38|0.43|0.46|0.36|0.51|0.44|0.48|0.43|0.46|0.42|\\n|ran area|(0.49)|(0.49)|(0.50)|(0.48)|(0.50)|(0.50)|(0.50)|(0.50)|(0.50)|(0.49)|\\n\\nNote: Household sampling weights are applied. Standard deviations are in parentheses. FHHs and MHHs are self-reported.\\n\\n75\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [76]}}, {\"page\": 77, \"chunk\": 77, \"input_text\": \"# **-** **<u>Table B.7. Descriptive Statistics, all countries years</u>**\\n\\n||**MHH**<br>|**FHH**<br>|\\n|---|---|---|\\n|Heads ae|48.14|55.52|\\n|g|(13.61)<br>|(14.10)<br>|\\n||0.36|0.70|\\n|Head does not complete primary school|(0.48)<br>|(0.46)<br>|\\n|Head's highest education level is primary|0.38<br>(0.48)<br>|0.19<br>(0.40)<br>|\\n|Head's hihest education level is secondar|0.12|0.06|\\n|g y|(0.32)<br>|(0.23)<br>|\\n|Hd' hiht dti ll i tti|0.15|0.05|\\n|eas ges eucaon eve s erary|(0.36)<br>|(0.21)<br>|\\n||0.01|0.04|\\n|Head is never married|(0.11)|(0.19)|\\n|Hd i id|0.95|0.17|\\n|ea s mono marre|(0.22)|(0.37)|\\n||0.02|0.03|\\n|Head is poly married|(0.15)|(0.16)|\\n||0.01|0.25|\\n|Head is divorced/separated|<br>(0.07)|<br>(0.43)|\\n|Head is widowed|0.01<br>|0.71<br>|\\n||(0.11)<br>|(0.45)<br>|\\n|Hd i ld|0.77|0.19|\\n|ea s empoye|(0.42)|(0.39)|\\n||0.03|0.01|\\n|Head is unemployed|<br>(0.16)|<br>(0.10)|\\n|Head is homemaker/housewife|0.01<br>|0.55<br>|\\n||(0.08)<br>|(0.50)<br>|\\n|d i d|0.00|0.00|\\n|Hea s stuent|(0.04)|(0.03)|\\n||0.18|0.27|\\n|Head is pensioner/retired/disabled|<br>(0.39)|<br>(0.45)|\\n|Head is other activities|0.02|0.01|\\n||(0.15)<br>|(0.11)<br>|\\n|P it ti|2,034,242.98|1,731,419.27|\\n|er capa consumpon|(2,496,523)|(2,451,185)|\\n|Per capita transfers|204.12<br>|471.50<br>|\\n||(805.10)|(1,655.70)|\\n|Household size|7.39|5.75|\\n||(4.05)<br>|(3.99)<br>|\\n||2.85|1.89|\\n|Number of children|(2.46)|(2.38)|\\n||0.26|0.29|\\n|Number of seniors|<br>(0.54)|<br>(0.48)|\\n|12 dlt hild|0.04|0.16|\\n|- aus, no c|(0.19)<br>|(0.37)<br>|\\n||0.10|0.09|\\n|1-2 adults, 1-2 children|(0.29)|(0.28)|\\n|1-2 adult, 3 or more children|0.25<br>|0.13<br>|\\n||(0.43)|(0.34)|\\n|3 adults or more 0-1 child|0.16|0.25|\\n|,|(0.37)<br>|(0.43)<br>|\\n|3 adults or more, 2-3 children|0.17<br>(0.38)|0.19<br>(0.39)|\\n||028|019|\\n|3 adults or more, 4 children or more|.<br>(0.45)|.<br>(0.39)|\\n|Rural area|0.32|0.27|\\n||(0.47)<br>|(0.45)<br>|\\n||0.68|0.73|\\n|Urban area|(0.47)|(0.45)|\\n\\nNote: Household sampling weights are applied. Standard deviations are in parentheses. FHHs and MHHs are self-reported.\\n\\n76\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [77]}}, {\"page\": 78, \"chunk\": 78, \"input_text\": \"# **Appendix C: Synthetic Panel Method**\\n\\nThis appendix offers a brief overview of the synthetic panel method based on Dang and Lanjouw (2023). Recent validations and applications of the synthetic panel methods by various researchers for different country contexts ranging from Africa to Latin America, the Middle East, and Europe have been encouraging in terms of accurate projections of economic status (Ferreira _et al._ , 2012; Beegle _et al._ , 2016; UNDP, 2016; OECD, 2018; Salvuci and Tarp, 2021; Ghomi, 2022). Let i be a vector of household characteristics observed in survey round _j_ (j= 1 or 2) that are also observed in the other survey round for household _i_ , i= 1,..., N.<sup>17</sup> These household characteristics can include such time-invariant variables as ethnicity, religion, language, place of xx birth, parental education, and other time-varying household characteristics if retrospective questions about the round-1 values of such characteristics are asked in the second round survey. To reduce spurious changes due to changes in household composition over time, we usually restrict the estimation samples to household heads in a certain age range, say 25 to 55, in the first cross section and adjust this age range accordingly in the second cross section. This restriction also helps ensure certain variables such as heads’ education attainment remains relatively stable over time (assuming most heads are finished with their schooling).<sup>18</sup> This age range is usually used in traditional pseudo-panel analysis but can vary depending on the cultural and economic factors in each specific setting. Population weights are then employed to provide estimates that represent the whole population. Then let i represent household consumption or income in survey round _j_ , j= 1 or 2. The linear projection of household consumption (or income) on household characteristics for each survey round is given by yy ′ i i i (C1) Let be the poverty line in period _j_ . We are interested in knowing the unconditional measures ii yy = ββ xx + εε of poverty mobility such as ii zz 1 2) (C2) which represents the percentage of households that are poor in the first survey round (year) but ii1 ii2 nonpoor in the second survey round, or the conditional measures such as PP(yy < zz aai yy > zz 2 1) (C3) which represents the percentage of poor households in the first round that escape poverty in the ii2 ii1 second round. PP(yy > zz | yy < zz\\n\\nIf true panel data are available, we can straightforwardly estimate the quantities in (C2) and (C3); but in the absence of such data, we can use synthetic panels to study mobility. To operationalize the framework, we make two standard assumptions. First, we assume that the underlying population being sampled in survey rounds 1 and 2 are identical such that their timeinvariant characteristics remain the same over time. More specifically, coupled with equation (C1), this implies the conditional distribution of expenditure in a given period is identical whether it is conditional on the given household characteristics in period 1 or period 2 (i.e., implies and have identical distributions) (Assumption 1). Second, we assume that i1 and ii1 ii2 i2 have a bivariate normal distribution with positive correlation coefficient ρ and standard xx = xx ii1 ii1 ii1 ii2 yydeviations |xx yy1 and |xx σ 2 respectively (Assumption 2). Quantity (2) can be estimated by εε εε εε εε σσ\\n\\n> 17 We suppress the index for countries and FHH types to make notation less cluttered in this appendix.\\n\\n> 18 While household heads may still increase their educational achievement in theory, this rarely happens in practice.\\n\\n77\", \"datasets\": [{\"mention_name\": {\"text\": \"panel data\", \"confidence\": 0.9062867164611816, \"start\": 2469, \"end\": 2479, \"clean_text\": \"panel data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2469, \"end\": 2479, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2469, \"end\": 2479, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2469, \"end\": 2479}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2469, \"end\": 2479, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.7969213128089905, \"start\": 176, \"end\": 186}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9062867164611816, \"start\": 176, \"end\": 186}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5028738975524902, \"start\": 176, \"end\": 186}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5028738975524902, \"start\": 176, \"end\": 186}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [78]}}, {\"page\": 79, \"chunk\": 79, \"input_text\": \"1−β1<sup>′</sup> <u>i2</u> 2−β2<sup>′</sup> <u>i2</u> 1 2) = Φ2 , − (C4) ~~�~~ zz xx zz xx where Φ 2(. ) stands for the bivariate normal cumulative distribution function (cdf), and PP(yyii1 < zz aai yyii2 > zz σσεε1 σσεε2 , −ρρ� 2(. ) stands for the bivariate normal probability density function (pdf). Note that in Equation (1), the estimated parameters obtained from data in both survey rounds are applied to data from the second φφ survey round (x2) (or the base year) for prediction, but we can use data from the first survey round as the base year as well. It is then straightforward to estimate quantity (C3) by dividing quantity 1−β1<sup>′</sup> <u>i2</u> (C2) by Φ , where Φ (. ) stands for the univariate normal cumulative distribution � ~~�~~ zz xx function (cdf). σσεε1 In Equation (4), the parameters and are estimated from Equation (C1), and ρ can be estimated using an approximation of the correlation of the cohort-aggregated household ii εεii ββ σσ consumption between the two surveys ( ). In particular, given an approximation of , where c indexes the cohorts constructed from the household survey data, the partial correlation yycc1yycc2 yycc1yycc2 coefficient ρ can be estimated by ρρ ρρ <u>i1 i2</u> ) 1<sup>′</sup> 2 (C5) ρρyy yy �vvaavv(yyii1)vvaavv(yyii2 −ββ vvaavv(xxii)ββ An alternative way to estimate ρρ= is to further assume that there is a cohort fixed effect in the σσεε1σσεε2 error terms and aggregate all the time-invariant variables to the cohort level and use the following equation ρρ i i i (C6) where the error term i includes a cohort fixed effect and the error i . ii yy = ββ ′xx + εε Note that the standard errors of estimates based on the synthetic panels can in fact be even ii εε ττ νν smaller than that of the true (or design-based) rate if there is a good model fit (or the sample size in the target survey is significantly larger than that in the base survey; see Dang and Lanjouw, 2023, for discussion).\\n\\nTables C.1-C.6 present the estimation results using Equation (C1) for all the countries and survey rounds. Tables C.7-C.12 present the descriptive statistics of the estimation sample. These tables show that while most of the time-invariant characteristics show similar distributions across survey rounds (and satisfy Assumption 1), some do not. For example, these include the shares of household heads achieving primary education or secondary education in the Arab Republic of Egypt during 2012-2015 (Table C.7). But the differences are practically very close to 0. Table C.13 presents the estimates for using Equations (C5) and (C6), where are estimated using Equation (C5) for all countries with cohorts being defined by age interacted with household heads’ education. ρρ ρρ We also provide alternative estimates for using Equation (C6). Using these estimates, Figures C.1 and C.2 offer qualitatively similar results to Figures 5 and 6. ρρ 78\", \"datasets\": [{\"mention_name\": {\"text\": \"household survey data\", \"confidence\": 0.8135799169540405, \"start\": 1098, \"end\": 1119, \"clean_text\": \"household survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1098, \"end\": 1119, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1098, \"end\": 1119, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1098, \"end\": 1119}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1098, \"end\": 1119, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8653478622436523, \"start\": 599, \"end\": 620}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8135799169540405, \"start\": 599, \"end\": 620}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.502353310585022, \"start\": 599, \"end\": 620}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.502353310585022, \"start\": 599, \"end\": 620}}, {\"mention_name\": {\"text\": \"synthetic panels\", \"confidence\": 0.34477362036705017, \"start\": 1679, \"end\": 1695, \"clean_text\": \"synthetic panels\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1679, \"end\": 1695, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1679, \"end\": 1695, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1679, \"end\": 1695}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1679, \"end\": 1695, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9455071091651917, \"start\": 494, \"end\": 510}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.34477362036705017, \"start\": 494, \"end\": 510}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.3596612513065338, \"start\": 494, \"end\": 510}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.3596612513065338, \"start\": 494, \"end\": 510}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [79]}}, {\"page\": 80, \"chunk\": 80, \"input_text\": \"# **<u>Table C.1 First-stage regressions, Arab Republic of Egypt</u>**\\n\\n||**2012-2**|**015**|**2015-2**|**017**|**2017-2**|**020**|\\n|---|---|---|---|---|---|---|\\n||**2012**|**2015**|**2015**|**2017**|**2017**|**2020**|\\n|Head`s age|0.006***|0.012***|0.010***|0.011***|0.009***|0.012***|\\n||(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Head is female|0.141***|0.141***|0.136***|0.234***|0.226***|0.198***|\\n||(0.02)|(0.02)|(0.02)|(0.02)|(0.02)|(0.02)|\\n|Highest education level is primary|0.142***|0.126***|0.123***|0.097***|0.099***|0.091***|\\n||(0.02)|(0.01)|(0.01)|(0.01)|(0.01)|(0.02)|\\n|Highest education level is secondary|0.199***|0.215***|0.209***|0.179***|0.173***|0.194***|\\n||(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|\\n|Highest education level is tertiary|0.461***|0.482***|0.481***|0.395***|0.394***|0.486***|\\n||(0.02)|(0.01)|(0.01)|(0.01)|(0.01)|(0.02)|\\n|Urban|0.241***|0.210***|0.215***|0.126***|0.126***|0.162***|\\n||(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|(0.01)|\\n|_cons|7.995***|8.091***|8.196***|8.555***|8.604***|8.632***|\\n||(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|\\n|adjusted R2|0.23|0.25|0.24|0.17|0.16|0.21|\\n|N|5102|8338|7836|8301|7799|7286|\\n\\n**Note:** ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Standard errors are in parentheses. Regression is estimated using Weighted Ordinary Least Squares. Household heads’ ages are restricted to between 25 and 55 for the first survey round and adjusted accordingly for the second round. The reference groups are household with no primary education and living in rural areas.\\n\\n79\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [80]}}, {\"page\": 81, \"chunk\": 81, \"input_text\": \"# **<u>Table C.2 First-stage regressions, Iraq</u>**\\n\\n||**2007-**<br>**2007**|**2012**<br>**2012**|\\n|---|---|---|\\n|Head`s age|-0.001**<br>(0.00)|0.005***<br>(0.00)|\\n|Head is female|0.012|0.114***|\\n||(0.02)|(0.02)|\\n|Highest education level is primary|-0.015|0.069***|\\n||(0.01)|(0.01)|\\n|Highest education level is secondary|0.059***|0.305***|\\n||(0.02)|(0.02)|\\n|Highest education level is tertiary|0.183***|0.441***|\\n||(0.02)|(0.01)|\\n|Urban|0.326***|0.317***|\\n||(0.01)|(0.01)|\\n|_cons|13.917***|13.966***|\\n||(0.03)|(0.03)|\\n|adjusted R2|0.08|0.13|\\n|N|12895|18552|\\n\\n**Note:** ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Standard errors are in parentheses. Regression is estimated using Weighted Ordinary Least Squares. Household heads’ ages are restricted to between 25 and 55 for the first survey round and adjusted accordingly for the second round. The reference groups are household with no primary education and living in rural areas.\\n\\n80\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [81]}}, {\"page\": 82, \"chunk\": 82, \"input_text\": \"# **<u>Table C.3 First-stage regressions, Jordan</u>**\\n\\n||**2010-**<br>**2010**|**2013**<br>**2013**|\\n|---|---|---|\\n|Head`s age|-0.002|0.004***|\\n||(0.00)|(0.00)|\\n|Head is female|0.213***|0.107***|\\n||(0.05)|(0.03)|\\n|Highest education level is primary|0.171***|0.280***|\\n||(0.04)|(0.03)|\\n|Highest education level is secondary|0.320***|0.449***|\\n||(0.05)|(0.04)|\\n|Highest education level is tertiary|0.666***|0.729***|\\n||(0.05)|(0.03)|\\n|Urban|0.022|0.037*|\\n||(0.03)|(0.02)|\\n|_cons|6.976***|6.718***|\\n||(0.08)|(0.06)|\\n|adjusted R2|0.16|0.18|\\n|N|1873|3437|\\n\\n**Note:** ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Standard errors are in parentheses. Regression is estimated using Weighted Ordinary Least Squares. Household heads’ ages are restricted to between 25 and 55 for the first survey round and adjusted accordingly for the second round. The reference groups are household with no primary education and living in rural areas.\\n\\n81\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [82]}}, {\"page\": 83, \"chunk\": 83, \"input_text\": \"# **<u>Table C.4 First-stage regressions, Mauritania</u>**\\n\\n||**2004-**|**2008**|**2008-**|**2014**|**2014-**|**2019**|\\n|---|---|---|---|---|---|---|\\n||**2004**|**2008**|**2008**|**2014**|**2014**|**2019**|\\n|Head`s age|-0.007***|-0.003***|-0.006***|-0.004***|-0.007***|-0.004***|\\n||(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Head is female|0.022|0.075***|0.065***|0.128***|0.103***|0.130***|\\n||(0.02)|(0.01)|(0.01)|(0.02)|(0.02)|(0.01)|\\n|Highest education level is primary|0.187***|0.143***|0.146***|0.073***|0.079***|-0.014|\\n||(0.03)|(0.02)|(0.02)|(0.02)|(0.02)|(0.02)|\\n|Highest education level is secondary|0.384***|0.383***|0.382***|0.223***|0.245***|0.145***|\\n||(0.03)|(0.02)|(0.02)|(0.02)|(0.02)|(0.02)|\\n|Highest education level is tertiary|0.708***|0.609***|0.621***|0.382***|0.408***|0.382***|\\n||(0.04)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|\\n|Urban|0.342***|0.605***|0.596***|0.353***|0.380***|0.368***|\\n||(0.02)|(0.01)|(0.01)|(0.01)|(0.02)|(0.01)|\\n|_cons|11.532***|11.710***|11.819***|12.279***|12.356***|12.421***|\\n||(0.05)|(0.04)|(0.04)|(0.04)|(0.04)|(0.04)|\\n|adjusted R2|0.18|0.32|0.32|0.16|0.18|0.18|\\n|N|6065|9269|9088|6672|6219|6425|\\n\\n**Note:** ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Standard errors are in parentheses. Regression is estimated using Weighted Ordinary Least Squares. Household heads’ ages are restricted to between 25 and 55 for the first survey round and adjusted accordingly for the second round. The reference groups are household with no primary education and living in rural areas.\\n\\n82\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [83]}}, {\"page\": 84, \"chunk\": 84, \"input_text\": \"# **<u>Table C.5 First-stage regressions, West Bank and Gaza</u>**\\n\\n||**2007-**|**2009**|**2009-**|**2011**|**2011**|**-2017**|\\n|---|---|---|---|---|---|---|\\n||**2007**|**2009**|**2009**|**2011**|**2011**|**2017**|\\n|Head`s age|0.005|0.006***|0.005***|0.002|-0.000|0.012***|\\n||(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Head is female|0.226**|0.206***|0.226***|0.133***|0.141***|0.160***|\\n||(0.11)|(0.05)|(0.06)|(0.05)|(0.05)|(0.06)|\\n|Highest education level is primary|0.177**|0.220***|0.214***|0.218***|0.210***|0.260***|\\n||(0.07)|(0.04)|(0.04)|(0.04)|(0.04)|(0.04)|\\n|Highest education level is secondary|0.276***|0.344***|0.346***|0.339***|0.341***|0.373***|\\n||(0.09)|(0.04)|(0.04)|(0.04)|(0.04)|(0.05)|\\n|Highest education level is tertiary|0.585***|0.607***|0.606***|0.602***|0.606***|0.518***|\\n||(0.08)|(0.04)|(0.04)|(0.04)|(0.04)|(0.04)|\\n|Urban|-0.034|0.014|0.012|-0.064**|-0.070**|-0.240***|\\n||(0.05)|(0.03)|(0.03)|(0.03)|(0.03)|(0.03)|\\n|Refugee|-0.327***|-0.039|-0.034|-0.314***|-0.314***|-0.538***|\\n||(0.07)|(0.05)|(0.05)|(0.04)|(0.04)|(0.05)|\\n|_cons|8.288***|8.388***|8.457***|8.718***|8.828***|8.381***|\\n||(0.15)|(0.07)|(0.07)|(0.08)|(0.08)|(0.08)|\\n|adjusted R2|0.08|0.09|0.09|0.09|0.09|0.11|\\n|N|962|2944|2938|3288|3229|2815|\\n\\n**Note:** ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Standard errors are in parentheses. Regression is estimated using Weighted Ordinary Least Squares. Household heads’ ages are restricted to between 25 and 55 for the first survey round and adjusted accordingly for the second round. The reference groups are household with no primary education and living in rural areas.\\n\\n83\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [84]}}, {\"page\": 85, \"chunk\": 85, \"input_text\": \"# **<u>Table C.6 First-stage regressions, Tunisia</u>**\\n\\n||**2005-**|**2010**|**2010-**|**2015**|**2015-**|**2021**|\\n|---|---|---|---|---|---|---|\\n||**2005**|**2010**|**2010**|**2015**|**2015**|**2021**|\\n|Head`s age|0.004***|0.008***|0.005***|0.008***|0.006***|0.014***|\\n||(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|(0.00)|\\n|Head is female|0.143***|0.053**|0.006|0.197***|0.184***|0.113***|\\n||(0.03)|(0.02)|(0.03)|(0.02)|(0.02)|(0.02)|\\n|Highest education level is primary|0.312***|0.305***|0.300***|0.210***|0.227***|0.142***|\\n||(0.04)|(0.04)|(0.04)|(0.01)|(0.02)|(0.02)|\\n|Highest education level is secondary|0.386***|0.293***|0.266***|0.443***|0.460***|0.354***|\\n||(0.02)|(0.02)|(0.02)|(0.02)|(0.02)|(0.02)|\\n|Highest education level is tertiary|0.945***|0.764***|0.754***|0.904***|0.911***|0.704***|\\n||(0.03)|(0.02)|(0.03)|(0.02)|(0.02)|(0.02)|\\n|Urban|0.495***|0.497***|0.499***|0.355***|0.370***|0.287***|\\n||(0.02)|(0.01)|(0.02)|(0.01)|(0.01)|(0.01)|\\n|_cons|6.547***|6.747***|6.881***|7.006***|7.084***|7.183***|\\n||(0.05)|(0.04)|(0.05)|(0.03)|(0.03)|(0.04)|\\n|adjusted R2|0.30|0.29|0.29|0.29|0.29|0.21|\\n|N|6769|7507|6425|16456|13635|10520|\\n\\n**Note:** ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Standard errors are in parentheses. Regression is estimated using Weighted Ordinary Least Squares. Household heads’ ages are restricted to between 25 and 55 for the first survey round and adjusted accordingly for the second round. The reference groups are household with no primary education and living in rural areas.\\n\\n84\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [85]}}, {\"page\": 86, \"chunk\": 86, \"input_text\": \"# **<u>Table C.7. Descriptive statistics of estimation sample, Arab Republic of Egypt</u>**\\n\\n|**Variables**||**2012-2015**|||**2015-2017**|||**2017-2020**||\\n|---|---|---|---|---|---|---|---|---|---|\\n||**2012**|**2015**|**diff**|**2015**|**2017**|**diff**|**2017**|**2020**|**diff**|\\n|L f r it nmtin|8.58|8.97|0.4***|8.95|9.32|0.4***|9.30|9.48|0.2***|\\n|og o pe capa cosupo|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|\\n|Hd`|41.32|43.97|2.7***|42.23|44.20|2.0***|43.14|42.89|-0.3*|\\n|eas age|(0.12)|(0.09)|(0.1)|(0.09)|(0.09)|(0.1)|(0.09)|(0.10)|(0.1)|\\n|Hd i fml|0.13|0.13|0.0|0.12|0.13|0.0|0.12|0.13|0.0|\\n|ea s eae|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|\\n|Hd' hiht dti ll i i|0.13|0.15|0.0***|0.15|0.15|0.0|0.15|0.14|-0.0***|\\n|eas ges eucaon eve s prmary|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|\\n|Hd' hiht dti ll i d|0.31|0.34|0.0**|0.35|0.36|0.0|0.37|0.37|0.0|\\n|eas ges eucaon eve s seconary|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|\\n|Hd' hiht dti ll i tti|0.20|0.19|-0.0|0.19|0.18|-0.0|0.18|0.21|0.0***|\\n|eas ges eucaon eve s erary|(0.01)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|\\n|Ub|0.42|0.42|-0.0|0.41|0.42|0.0|0.41|0.43|0.0**|\\n|ran area|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|\\n\\nNote: ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Standard errors are in parentheses. The data are unweighted. Household heads' ages are restricted to between 25 and 55 for the first survey round and adjusted accordingly for the second survey round. The reference groups are household with no primary education and living in rural areas.\\n\\n85\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [86]}}, {\"page\": 87, \"chunk\": 87, \"input_text\": \"# **<u>Table C.8 Descriptive statistics of estimation sample, Iraq</u>**\\n\\n||**2007**|**2007-2012**<br>**2012**|**diff**|\\n|---|---|---|---|\\n|Lo of er caita consumtion|14.22|14.67|0.1***|\\n|g p p p|(0.01)|(0.01)|(0.0)|\\n|Head` ae|40.46|44.28|1.7***|\\n|s g|(0.07)|(0.06)|(0.2)|\\n|Hd i fl|0.09|0.09|0.0|\\n|ea s emae|(0.00)|(0.00)|(0.0)|\\n|Hd' hiht dti ll i i|0.41|0.40|0.0***|\\n|eas ges eucaon eve s prmary|(0.00)|(0.00)|(0.0)|\\n|Hd' hiht dti ll i d|0.12|0.08|-0.0|\\n|eas ges eucaon eve s seconary|(0.00)|(0.00)|(0.0)|\\n|Hd' hih di ll i i|0.19|0.15|-0.0***|\\n|eas gest eucaton eve s tertary|(0.00)|(0.00)|(0.0)|\\n||0.68|0.60|-0.1***|\\n|Urban area|(0.00)|(0.00)|(0.0)|\\n\\nNote: ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Standard errors are in parentheses. The data are unweighted. Household heads' ages are restricted to between 25 and 55 for the first survey round and adjusted accordingly for the second survey round. The reference groups are household with no primary education and living in rural areas.\\n\\n86\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [87]}}, {\"page\": 88, \"chunk\": 88, \"input_text\": \"# **<u>Table C.9 Descriptive statistics of estimation sample, Jordan</u>**\\n\\n||**2010**|**2010-2013**<br>**2013**|**diff**|\\n|---|---|---|---|\\n|Lo of er caita consumtion|7.29|7.35|0.1***|\\n|g p p p|(0.01)|(0.01)|(0.0)|\\n|Hd`|41.00|42.72|1.7***|\\n|eas age|(0.18)|(0.14)|(0.2)|\\n|Hd i fl|0.09|0.09|0.0|\\n|ea s emae|(0.01)|(0.00)|(0.0)|\\n|Hd' hiht dti ll i i|0.51|0.55|0.0***|\\n|eas ges eucaon eve s prmary|(0.01)|(0.01)|(0.0)|\\n|Hd' hiht dti ll i d|0.16|0.15|-0.0|\\n|eas ges eucaon eve s seconary|(0.01)|(0.01)|(0.0)|\\n|Hd' hih di ll i i|0.23|0.19|-0.0***|\\n|eas gest eucaton eve s tertary|(0.01)|(0.01)|(0.0)|\\n||0.74|0.63|-0.1***|\\n|Urban area|(0.01)|(0.01)|(0.0)|\\n\\nNote: ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Standard errors are in parentheses. The data are unweighted. Household heads' ages are restricted to between 25 and 55 for the first survey round and adjusted accordingly for the second survey round. The reference groups are household with no primary education and living in rural areas.\\n\\n87\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [88]}}, {\"page\": 89, \"chunk\": 89, \"input_text\": \"**<u>Table C.10 Descriptive statistics of estimation sample, Mauritania</u>**\\n\\n|||**2004-2008**|||**2008-2014**|||**2014-201**|**9**|\\n|---|---|---|---|---|---|---|---|---|---|\\n||**2004**|**2008**|**diff**|**2008**|**2014**|**diff**|**2014**|**2019**|**diff**|\\n|L f i i|11.66|12.09|0.4***|12.10|12.54|0.4***|12.56|12.68|0.1***|\\n|og o per capta consumpton|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|\\n|Hd`|42.41|43.90|1.5***|41.90|46.40|4.5***|42.42|45.49|3.1***|\\n|eas age|(0.10)|(0.08)|(0.1)|(0.08)|(0.10)|(0.1)|(0.10)|(0.11)|(0.1)|\\n|Hd i fl|0.16|0.29|0.1***|0.30|0.30|0.0|0.31|0.37|0.1***|\\n|ea s emae|(0.00)|(0.00)|(0.0)|(0.00)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|\\n|Hd' hih di ll i i|0.10|0.11|0.0|0.12|0.12|-0.0|0.14|0.29|0.1***|\\n|eas gest eucaton eve s prmary|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|(0.00)|(0.01)|(0.0)|\\n|Hd' hiht dti ll i d|0.13|0.13|-0.0|0.13|0.13|0.0|0.15|0.19|0.0***|\\n|eas ges eucaon eve s seconary|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|\\n|Hd' hih di ll i i|0.05|0.05|0.0|0.05|0.05|0.0**|0.05|0.06|0.0|\\n|eas gest eucaton eve s tertary|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|\\n|Ub|0.51|0.48|-0.0***|0.47|0.59|0.1***|0.59|0.50|-0.1***|\\n|ran area|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|\\n\\nNote: ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Standard errors are in parentheses. The data are unweighted. Household heads' ages are restricted to between 25 and 55 for the first survey round and adjusted accordingly for the second survey round. The reference groups are household with no primary education and living in rural areas.\\n\\n88\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [89]}}, {\"page\": 90, \"chunk\": 90, \"input_text\": \"**<u>Table C.11. Descriptive statistics of estimation sample, West Bank and Gaza</u>**\\n\\n|||**2007-20**|**09**||**2009-20**|**11**||**2011-2**|**017**|\\n|---|---|---|---|---|---|---|---|---|---|\\n||**2007**|**2009**|**difference**|**2009**|**2011**|**difference**|**2011**|**2017**|**difference**|\\n|L f i i|8.78|9.06|0.3***|9.06|9.13|0.1***|9.13|9.23|0.1***|\\n|og o per capta consumpton|(0.02)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|\\n|Hd`|40.03|40.72|0.7**|39.80|41.65|1.9***|40.71|44.94|4.2***|\\n|eas age|(0.25)|(0.15)|(0.3)|(0.15)|(0.14)|(0.2)|(0.14)|(0.16)|(0.2)|\\n|Hd i fl|0.06|0.06|0.0|0.06|0.07|0.0*|0.07|0.08|0.0|\\n|ea s emae|(0.01)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|\\n|Hd' hiht dti ll i i|0.55|0.50|-0.0***|0.50|0.51|0.0|0.51|0.51|-0.0|\\n|eas ges eucaon eve s prmary|(0.02)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|\\n|Hd' hiht dti ll i d|0.18|0.17|-0.0|0.17|0.18|0.0|0.18|0.16|-0.0**|\\n|eas ges eucaon eve s seconary|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|\\n|Hd' hiht dti ll i tti|0.17|0.22|0.0***|0.21|0.22|0.0|0.22|0.23|0.0|\\n|eas ges eucaon eve s erary|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|\\n|Ub|0.54|0.70|0.2***|0.70|0.53|-0.2***|0.53|0.56|0.0***|\\n|ran area|(0.02)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|\\n|Rf|0.18|0.12|-0.1***|0.12|0.21|0.1***|0.21|0.12|-0.1***|\\n|eugee area|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|(0.01)|(0.01)|(0.0)|\\n\\nNote: ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Standard errors are in parentheses. The data are unweighted. Household heads' ages are restricted to between 25 and 55 for the first survey round and adjusted accordingly for the second survey round. The reference groups are household with no primary education and living in rural areas.\\n\\n89\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [90]}}, {\"page\": 91, \"chunk\": 91, \"input_text\": \"**<u>Table C.12. Descriptive statistics of estimation sample, Tunisia</u>**\\n\\n|||**2005-2010**|||**2010-2015**|||**2015-2021**||\\n|---|---|---|---|---|---|---|---|---|---|\\n||**2005**|**2010**|**diff**|**2010**|**2015**|**diff**|**2015**|**2021**|**diff**|\\n|L f it ti|7.23|7.58|0.4***|7.55|7.98|0.4***|7.96|8.36|0.4***|\\n|og o per capa consumpon|(0.01)|(0.01)|(0.0)|(0.01)|(0.00)|(0.0)|(0.01)|(0.01)|(0.0)|\\n|Hd`|43.55|46.72|3.2***|44.24|47.23|3.0***|44.32|48.40|4.1***|\\n|eas age|(0.09)|(0.09)|(0.1)|(0.09)|(0.06)|(0.1)|(0.06)|(0.08)|(0.1)|\\n|Hd i fl|0.12|0.11|-0.0*|0.11|0.11|0.0|0.10|0.12|0.0***|\\n|ea s emae|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|\\n|Hd' hiht dti ll i i|0.04|0.03|-0.0***|0.03|0.46|0.4***|0.44|0.44|0.0|\\n|eas ges eucaon eve s prmary|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|\\n|Hd' hiht dti ll i d|0.14|0.11|-0.0***|0.12|0.30|0.2***|0.32|0.32|0.0|\\n|eas ges eucaon eve s seconary|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|\\n|Hd' hiht dti ll i tti|0.08|0.09|0.0|0.09|0.11|0.0***|0.12|0.12|0.0|\\n|eas ges eucaon eve s erary|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|\\n|Ub|0.65|0.66|0.0|0.65|0.62|-0.0***|0.61|0.62|0.0**|\\n|ran area|(0.01)|(0.01)|(0.0)|(0.01)|(0.00)|(0.0)|(0.00)|(0.00)|(0.0)|\\n\\nNote: ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.10 levels. Standard errors are in parentheses. The data are unweighted. Household heads' ages are restricted to between 25 and 55 for the first survey round and adjusted accordingly for the second survey round. The reference groups are household with no primary education and living in rural areas.\\n\\n90\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [91]}}, {\"page\": 92, \"chunk\": 92, \"input_text\": \"|**Table C.13. Estimated rho (**ρ**) from**<br>|**cross-sectional data**<br><br>||\\n|---|---|---|\\n|**Country**|**Period**<br>ρ|**Alternative**ρ|\\n||2012-2015<br>0.84|0.52|\\n|**Egypt, Arab Rep.**|2015-2017<br>0.89|0.46|\\n||2017-2020<br>0.79|0.61|\\n||2007-2009<br>0.54|0.56|\\n|**West Bank and Gaza**|2009-2011<br>0.62|0.66|\\n||2011-2017<br>0.34|0.59|\\n||2005-2010<br>0.57|0.67|\\n|**Tunisia**|2010-2015<br>0.73|0.65|\\n||2015-2021<br>0.89|0.61|\\n||2004-2008<br>0.77|0.57|\\n|**Mauritania**|2008-2014<br>0.63|0.56|\\n||2014-2019<br>0.70<br>|0.61<br>|\\n|**Iraq**<br><br>|2007-2012<br>0.68<br><br>|0.37<br>|\\n|**Jordan**<br><br>|2010-2013<br>0.63<br><br>|0.63<br>|\\n|**Note:** ρare estimated using Equati<br>education. Alternativeρ’s are estima|on (C5) for all countries with cohorts being defined by age i<br>ted using Equation (C6).|nteracted with household heads’|\\n\\n91\", \"datasets\": [{\"mention_name\": {\"text\": \"cross-sectional data\", \"confidence\": 0.9488121867179871, \"start\": 50, \"end\": 70, \"clean_text\": \"cross-sectional data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 50, \"end\": 70, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 50, \"end\": 70, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 50, \"end\": 70}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 50, \"end\": 70, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9965872764587402, \"start\": 50, \"end\": 70}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9488121867179871, \"start\": 50, \"end\": 70}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5685274600982666, \"start\": 50, \"end\": 70}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5685274600982666, \"start\": 50, \"end\": 70}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [92]}}, {\"page\": 93, \"chunk\": 93, \"input_text\": \"# **Figure C.1. Probabilities of Female-Headed Households Escaping Poverty in Second Year Conditional on Being Poor in First Year (percentage)**\\n\\n<!-- Start of picture text --> Egypt 2017-2020 Iraq 2007-2012<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>Jordan 2010-2013 Mauritania 2014-2019<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>Palestine 2011-2017 Tunisia 2015-2021<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>FHH wo/children FHH w/children Non-FHH<br>75<br>45<br>70<br>40<br>65<br>35<br>Percentage (%) Percentage (%)<br>60<br>30<br>60 55<br>55 50<br>50<br>45<br>Percentage (%) Percentage (%)<br>45<br>40<br>40 55<br>35<br>50<br>30<br>45<br>25<br>Percentage (%) Percentage (%)<br>20 40<br><!-- End of picture text --> Note: Estimates are obtained with synthetic panel data and weighted with population weights where the second survey round is used as the base year. The figure shows the percentage of the population that moves out of poverty in the second year. FHH characteristics are measured in second period. Dashed red lines represent the national average for each period. Household heads' ages are restricted to 25-55 for the first survey round and adjusted accordingly for the second survey round. Standard errors are estimated with 1,000 bootstraps. ’s are estimated using Equation (C6). ρρ 92\", \"datasets\": [{\"mention_name\": {\"text\": \"synthetic panel data\", \"confidence\": 0.8587033152580261, \"start\": 953, \"end\": 973, \"clean_text\": \"synthetic panel data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 953, \"end\": 973, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 953, \"end\": 973, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 953, \"end\": 973}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 953, \"end\": 973, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9220925569534302, \"start\": 301, \"end\": 321}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8587033152580261, \"start\": 301, \"end\": 321}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5956490635871887, \"start\": 301, \"end\": 321}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5956490635871887, \"start\": 301, \"end\": 321}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [93]}}, {\"page\": 94, \"chunk\": 94, \"input_text\": \"**Figure C.2. Probabilities of Female-Headed Households Falling in Poverty in Second Year Conditional on Being Non-poor in First Year (percentage)**\\n\\n<!-- Start of picture text --> Egypt 2017-2020 Iraq 2007-2012<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>Jordan 2010-2013 Mauritania 2014-2019<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>Palestine 2011-2017 Tunisia 2015-2021<br>Self-reported Potential Self-reported Potential<br>Females>0.5 Educated fem. Females>0.5 Educated fem.<br>FHH wo/children FHH w/children Non-FHH<br>20 20<br>18 18<br>16 16<br>14 14<br>Percentage (%) 12 12 Percentage (%)<br>10 10<br>14 18<br>12 16<br>10 14<br>8 12<br>Percentage (%) Percentage (%)<br>6 10<br>45 16<br>40 14<br>35 12<br>30<br>10<br>Percentage (%) Percentage (%)<br>25 8<br><!-- End of picture text --> Note: Estimates are obtained with synthetic panel data and weighted with population weights where the second survey round is used as the base year. The figure shows the percentage of the population that enters poverty in the second year. FHH characteristics are measured in second period. Dashed red lines represent the national average for each period. Household heads' ages are restricted to 25-55 for the first survey round and adjusted accordingly for the second survey round. Standard errors are estimated with 1,000 bootstraps. ’s are estimated using Equation (C6). ρρ 93\", \"datasets\": [{\"mention_name\": {\"text\": \"synthetic panel data\", \"confidence\": 0.8386054635047913, \"start\": 960, \"end\": 980, \"clean_text\": \"synthetic panel data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 960, \"end\": 980, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 960, \"end\": 980, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 960, \"end\": 980}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 960, \"end\": 980, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9220925569534302, \"start\": 304, \"end\": 324}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8386054635047913, \"start\": 304, \"end\": 324}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5956490635871887, \"start\": 304, \"end\": 324}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5956490635871887, \"start\": 304, \"end\": 324}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [94]}}, {\"page\": 95, \"chunk\": 95, \"input_text\": \"# **Appendix D: Further Analysis with Equivalence Scale**\\n\\nGiven our consistent observation of self-reported FHHs having an advantage in terms of greater mobility, the question arises as to whether this conclusion remains valid if we extend to selecting a measure of household members' welfare that goes beyond household expenditure per capita.\\n\\nWithin the context of poverty dynamics, we show two scenarios for self-reported FHHs as an example: one in which FHHs have a greater probability to escape poverty compared to non-FHHs (denoted by the orange-shaded region in Figure D.1), and another where FHHs are less likely to escape poverty than non-FHHs (represented by the blue-shaded area in Figure D.1). Importantly, the selection of specific scale parameters can significantly alter the conclusions drawn regarding poverty dynamics among FHHs.\\n\\nIn particular, when assessing consumption on a per capita basis (i.e., when β=1 and θ=1), selfreported FHHs consistently exhibit a higher probability of escaping poverty than non-FHHs and it holds true across all countries. Intriguingly, these findings align with those in Abanokova _et al._ (2022), which demonstrated a persistent upward mobility when income is evaluated on a per capita basis.\\n\\nThe conclusions regarding poverty dynamics shift when adopting OECD-recommended (modified) equivalence scales, which assign a value of 0.3 to each child aged 0-13 (indicated by the green dashed line) and/or the \\\"square root scale\\\" set at 0.5 (represented by the red dashed line). Under the \\\"square root scale,\\\" self-reported FHHs become less likely to escape poverty than nonFHHs in Jordan, the West Bank and Gaza, and Tunisia, regardless of the child parameter value. The use of a lower scale parameter than the “square root scale” alters the conclusion in the Arab Republic of Egypt, but the sensitivity to the child parameter is also observed. Significant sensitivity to the child parameter is found in the West Bank and Gaza. When the child parameter is set to 0.4 or lower, there is a shift in the scenario from FHHs experiencing upward mobility to FHHs facing downward mobility. However, varying the parameters of economies of scale and child parameters from 0 to 1 does not alter the conclusions regarding poverty dynamics for Mauritania, Iraq and Jordan.\\n\\nThe absolute difference in the percentage of the population transitioning out of poverty between FHHs and non-FHHs is also influenced by the scale parameters. In the case of Mauritania, where self-reported FHHs are more likely to escape poverty than non-FHHs, fluctuations in scale parameters can result in significant changes in the percentage of self-reported FHHs escaping poverty. These variations can yield a discrepancy of up to 6.8 percentage points, depending on the scale parameters applied.\\n\\nThe overarching finding is that the parameter dictating the economies of scale and the private– public nature of household consumption contributes non-trivially to the poverty ranking between FHHs and non-FHHs across most countries and FHH definitions, while the child parameter having a comparatively smaller impact compared to household size. These results mirror our earlier observation in Abanokova _et al._ (2022) regarding the sensitivity of income dynamics to scale parameters.\\n\\n94\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [95]}}, {\"page\": 96, \"chunk\": 96, \"input_text\": \"**Figure D.1. Self-reported FHHs– non-FHHs Differences in Probabilities of Escaping Poverty in Second Year Conditional on Being Poor in First Year (percentage points), by Scale Parameters**\\n\\nδ **Note:** Each figure shows 2-parameter equivalence scale that adjusts household consumption: (aij + αkij) where a– number of adults in the household, k – number of children in the household, α is “child parameter” that accounts for the needs of children aged 0-13 and δ is “size parameter” that measures the degree of economies of scale in household consumption. Both parameters are varying between 0 and 1. The blue zone indicates lower probabilities of escaping poverty among FHHs compared to non-FHHs. The orange zone indicates higher probabilities of escaping poverty among FHHs compared to non-FHHs. Each bar shows the difference in the percentage of the population that moves out of poverty among FHHs compared to non-FHHs in the second year (expressed in percentage points). We use OECD recommended (modified) equivalence scale that assigns a value of 0.3 to each child aged 0-13 (green dashed line) and “square root scale” that equals to 0.5 (red dashed line). The top right corner of the box (marked ×) illustrates the case when δ = 1 and α= 1, which represents per capita expenditure (“Per Capita”).\\n\\n95\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099916001162436719/pdf/IDU11b4274c21f7ee1444b192b010c62ef384a79.pdf\", \"pages\": [96]}}]}"}}},{"rowIdx":79,"cells":{"doc_id":{"kind":"string","value":"34240433"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34240433\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"admreg\": \"Other,Other\", \"admreg_key\": \"912972,912972\", \"authors\": {\"0\": {\"author\": \"Stacy,Brian William\"}, \"1\": {\"author\": \"Kitzmüller,Lucas\"}, \"2\": {\"author\": \"Wang,Xiaoyu\"}, \"3\": {\"author\": \"Mahler,Daniel Gerszon\"}, \"4\": {\"author\": \"Serajuddin,Umar\"}}, \"count\": \"World\", \"count_key\": \"517191\", \"docna\": {\"0\": {\"docna\": \"Missing Evidence : Tracking Academic Data\\n Use around the World\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"Development Indicators and Data (DECID)\", \"projn\": \"1W-Measuring The Use Of Data In Academia -- P178728\", \"prdln\": \"Research Activity\", \"prdln_key\": \"1582145\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34240433\"}, \"repnb\": \"WPS10673\", \"docdt\": \"2024-01-17T00:00:00Z\", \"datestored\": \"2024-01-17T03:03:03Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"Data-driven research on a country is key\\n to producing evidence-based public policies. Yet little is\\n known about where data-driven research is lacking and how it\\n could be expanded. This paper proposes a method for tracking\\n academic data use by country of subject, applying natural\\n language processing to open-access research papers. The\\n model’s predictions produce country estimates of\\n the number of articles using data that are highly correlated\\n with a human-coded approach, with a correlation of 0.99.\\n Analyzing more than 1 million academic articles, the paper\\n finds that the number of articles on a country is strongly\\n correlated with its gross domestic product per capita,\\n population, and the quality of its national statistical\\n system. The paper identifies data sources that are strongly\\n associated with data-driven research and finds that\\n availability of subnational data appears to be particularly\\n important. Finally, the paper classifies countries into\\n groups based on whether they could most benefit from\\n increasing their supply of or demand for data. The findings\\n show that the former applies to many low- and\\n lower-middle-income countries, while the latter applies to\\n many upper-middle- and high-income countries.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10673; RRR; RSB; DIGITAL\", \"display_title\": \"Missing Evidence : Tracking Academic\\n Data Use around the World\", \"disclosure_date\": \"2024-01-17T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099503001172493286/text/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.txt\", \"listing_relative_url\": \"/research/2024/01/34240433/missing-evidence-tracking-academic-data-use-around-world\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099503001172493286/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286\", \"new_url\": \"2024/01/34240433/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286\", \"ext_pub_date\": \"2024-01-17T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Public Funded Research and Development\\n m1504101 2787,Statistical Surveys m1504545 3123\", \"docm_id\": \"IDU-ad0286b6-aec0-4d8c-9d3e-e651d3fe6286\", \"chronical_docm_id\": \"IDU-ad0286b6-aec0-4d8c-9d3e-e651d3fe6286\", \"publishtoextweb_dt\": \"2024-01-17T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10673\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"projectid\": \"P178728\", \"guid\": \"099503001172493286\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"39\", \"author_keys\": {\"0\": {\"author_key\": \"1312545\"}, \"1\": {\"author_key\": \"NA\"}, \"2\": {\"author_key\": \"2172451\"}, \"3\": {\"author_key\": \"1539953\"}, \"4\": {\"author_key\": \"539942\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099503001172493286/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10673 Missing Evidence Tracking Academic Data Use around the World _Brian Stacy Lucas Kitzmüller Xiaoyu Wang Daniel Gerszon Mahler Umar Serajuddin_ Development Economics Development Data Group January 2024 A verified reproducibility package for this paper is available at http://reproducibility.worldbank.org, click **here** for direct access.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10673\\n\\n# **Abstract**\\n\\nData-driven research on a country is key to producing evidence-based public policies. Yet little is known about where data-driven research is lacking and how it could be expanded. This paper proposes a method for tracking academic data use by country of subject, applying natural language processing to open-access research papers. The model’s predictions produce country estimates of the number of articles using data that are highly correlated with a human-coded approach, with a correlation of 0.99. Analyzing more than 1 million academic articles, the paper finds that the number of articles on a country is strongly correlated with its gross domestic product per capita, population, and the quality of its national statistical system. The paper identifies data sources that are strongly associated with data-driven research and finds that availability of subnational data appears to be particularly important. Finally, the paper classifies countries into groups based on whether they could most benefit from increasing their supply of or demand for data. The findings show that the former applies to many low- and lower-middle-income countries, while the latter applies to many upper-middle- and high-income countries.\\n\\nThis paper is a product of the Development Data Group, Development Economics. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at bstacy@worldbank.org. A verified reproducibility package for this paper is available at http://reproducibility. worldbank.org, click **here** for direct access.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# **Missing Evidence: Tracking Academic Data Use around the World**\\n\\nBrian Stacy, Lucas Kitzmüller, Xiaoyu Wang, Daniel Gerszon Mahler, and Umar Serajuddin<sup>1</sup> Keywords: Data, academia, research, natural language processing JEL codes: C45, C52, O30\\n\\n> 1 Stacy, Wang, Mahler, and Serajuddin are with the World Bank’s Development Data Group. Lucas Kitzmüller completed the work while at the European Bank for Reconstruction and Development (EBRD). Corresponding author: Brian Stacy (bstacy@worldbank.org). We are grateful for comments from Dean Jolliffe, Jishnu Das, Olivier Dupriez, and Patrick Brock. We acknowledge financial support from a World Bank Research Support Grant (P178728). The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Views presented are those of the authors and not necessarily of the EBRD.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **1 Introduction**\\n\\nIn recent decades, the amount of data produced has exploded, generating boundless opportunities for policies to improve people’s lives (World Bank 2021). Though data can be valuable in their raw form, the full value of data is only realized when they are analyzed to create insights, and these insights are converted to public policies or increased accountability.\\n\\nResearchers have a vital role to play in this regard. Many researchers spend countless hours digesting data, using data to create new knowledge, and communicating this knowledge with the intent of impacting public discourse and public policies. There are numerous examples of data-driven analyses having real and important impacts on people’s lives (Jolliffe et al. 2023). One example from Brazil explicitly looks at researchers’ ability to influence policy outcomes. There, evidence from 2,150 municipalities found that informing municipal mayors of research findings on the effectiveness of a simple policy change increased the probability that their municipality implemented the policy by 10 percentage points (Hjort et al. 2021).\\n\\nWithout research, there is a risk that the return of data to society will be reduced, and policies to improve lives unrealized. Yet very little is known about where there is missing data-driven evidence and how governments can best stimulate an evidence base for local decision makers. This paper attempts to fill these gaps by addressing two questions: (1) Which countries are the subject of research papers using data? (2) How can countries increase their national evidence base? We focus on data-driven research due to the increasing importance of data for policy making and the specific policies that are needed to increase the supply and demand of data, such as boosting statistical capacity and improving data literacy.\\n\\nTo answer the first question, we introduce a new method for measuring data use in research articles based on 1 million English-language articles spanning 216 countries and various academic fields. These articles are made available by Semantic Scholar Open Research Corpus (s2orc), which has digitized millions of research papers worldwide and made its raw text accessible via APIs (Lo et al. 2020). With the aid of Amazon Mturk workers, we manually code 900 of these articles as using data or not, upon which we train a natural language model to predict the coding of the Mturk workers (Devlin et al. 2018). The model achieved an 87% out-of-sample accuracy rate, and when the articles are aggregated to the country level, the model had a correlation of 0.99 with the number of articles classified by Mturk workers. The model was then applied to 1 million academic articles from the s2orc database from 2000-2020. The model estimates the amount of data-driven research on a country regardless of where the researchers may be located, _not_ the amount of data-driven research by the citizens of a country. We argue that the former is the relevant quantity to understand the evidence base available to national decision makers.\\n\\nWe find that data-driven research is strongly correlated with GDP per capita and population, which together account for around 75% of the variation across countries. High-income countries are the subject of nearly 50% of all papers using data despite representing only around 15% of the world's population, while low-income countries, comprising approximately 10% of the world population, account for only about 5% of articles using data.\\n\\n2\", \"datasets\": [{\"mention_name\": {\"text\": \"s2orc database\", \"confidence\": 0.6825814247131348, \"start\": 2732, \"end\": 2746, \"clean_text\": \"s2orc database\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2732, \"end\": 2746, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2732, \"end\": 2746, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2732, \"end\": 2746}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2732, \"end\": 2746, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9945186972618103, \"start\": 1058, \"end\": 1072}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6825814247131348, \"start\": 1058, \"end\": 1072}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6581195592880249, \"start\": 1058, \"end\": 1072}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6581195592880249, \"start\": 1058, \"end\": 1072}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"To answer the second question – how countries can increase their national evidence base – we first establish that a country’s statistical capacity is predictive of data research even after controlling for population and GDP, and articles not using data (which we use as a proxy for the general research interest in the country). To understand which part of a country’s statistical capacity is most important for increasing data-driven research, we explore the data sources most related to academic data use. We find that the availability of geospatial data at the first administrative level is associated with a 1.1% increase in data use, that a population census in the past decade is associated with a 0.3% increase in data use, and that two or more labor force (agriculture) surveys over the past 10 years is associated with a 0.4% (0.2%) increase in data use. Though we are unable to establish these links causally, there are concrete data products that governments could supply to likely increase the evidence-base at its disposal.\\n\\nBoosting the supply of data is one way for a country to increase the data-driven research it is subject to, another is to increase the demand for its data. This is particularly relevant for countries that have already invested in relevant data products but are nonetheless subject to relatively little data-driven research. These are cases where existing data are underutilized and where it may be relevant to make existing data more accessible to researchers and possibly boost data literacy in the country. To explore this distinction between boosting data supply and data demand, and building upon Porteous (2020), we classify countries into four groups: _Deserts_ have little data demand and little data supply, _swamps_ have high data supply but little data demand, _oases_ have high data demand but little data supply, and _lakes_ have high data demand and high data supply. Nearly two-thirds of low-income countries and countries in Sub-Saharan Africa are oases, suggesting that these countries are getting relatively large evidence from their data supply and that they have relatively little issue with lack of demand, but that they could benefit from increasing the data available to researchers. By contrast, nearly half of countries in Europe are data swamps, suggesting a priority on increasing the use of existing data.\\n\\nPrevious research has highlighted gaps between countries in economic research output and noted that richer countries are the subject of more economic research. For instance, Robinson, Hartley, and Schneider (2006), Das et al. (2013), and Porteous (2020) examine which countries are studied most by economists using the EconLit database. Cameron, Mishra, and Brown (2016) and Sabet and Brown (2018) extend this to note that impact evaluations are highly uneven across countries as well. Phillips and Greene (2022) show conflict research is biased towards Western countries, while Courtioux et al. (2022) show that academic research is highly related to public investments in scientific research.\\n\\nWe contribute to the literature by applying NLP to improve our understanding of what countries are under-researched. Using NLP allows us to go beyond the existing literature in three ways: (1) scale up the sample size and look at all fields of interest rather than only economics, (2) identify papers that use data, which is key to understanding whether data demand or data supply could be explaining a country’s lack of research, and (3) indicate measures countries can take to increase data research.\\n\\n3\", \"datasets\": [{\"mention_name\": {\"text\": \"geospatial data\", \"confidence\": 0.5393737554550171, \"start\": 541, \"end\": 556, \"clean_text\": \"geospatial data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 541, \"end\": 556, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 541, \"end\": 556, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 541, \"end\": 556}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 541, \"end\": 556, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.8829935193061829, \"start\": 541, \"end\": 556}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5393737554550171, \"start\": 541, \"end\": 556}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4219647943973541, \"start\": 541, \"end\": 556}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4219647943973541, \"start\": 541, \"end\": 556}}, {\"mention_name\": {\"text\": \"population census\", \"confidence\": 0.7281874418258667, \"start\": 646, \"end\": 663, \"clean_text\": \"population census\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 646, \"end\": 663, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 646, \"end\": 663, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 646, \"end\": 663}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 646, \"end\": 663, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.9930364489555359, \"start\": 646, \"end\": 663}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7281874418258667, \"start\": 646, \"end\": 663}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.561593770980835, \"start\": 646, \"end\": 663}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.561593770980835, \"start\": 646, \"end\": 663}}, {\"mention_name\": {\"text\": \"EconLit database\", \"confidence\": 0.9328306913375854, \"start\": 2691, \"end\": 2707, \"clean_text\": \"EconLit database\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2691, \"end\": 2707, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2691, \"end\": 2707, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2691, \"end\": 2707}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2691, \"end\": 2707, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9985999464988708, \"start\": 1066, \"end\": 1082}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.9328306913375854, \"start\": 1066, \"end\": 1082}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9665244817733765, \"start\": 1066, \"end\": 1082}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9665244817733765, \"start\": 1066, \"end\": 1082}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"The remainder of the paper is structured as follows. Section 2 discusses our data sources, section 3 details our methodology, section 4 introduces a theoretical framework, section 5 presents our empirical results, section 6 conducts robustness checks, and section 7 concludes.\\n\\n# **2 Data**\\n\\nOur data source of academic articles is the Semantic Scholar Open Research Corpus (S2ORC) (Lo et al. 2020). The corpus contains more than 130 million English language academic papers across multiple disciplines. The papers included in the Semantic Scholar corpus are gathered directly from publishers, from open archives such as arXiv or PubMed, and crawled from the internet.\\n\\nWe placed some restrictions on the articles to make them usable and relevant for our purposes. First, only articles with an abstract and parsed PDF or latex file are included in the analysis. The full text of the abstract is necessary to classify the country of study and whether the article uses data. The parsed PDF and latex file are important for extracting important information like the date of publication and field of study. This restriction eliminated a large number of articles in the original corpus. Around 30 million articles remain after keeping only articles with a parsable (i.e., suitable for digital processing) PDF, and around 26% of those 30 million are eliminated when removing articles without an abstract. Second, only articles from the years 2000 to 2020 were considered. This restriction eliminated an additional 9% of the remaining articles. Finally, articles from the following fields of study were excluded, as we aim to focus on fields that are likely to use data produced by countries’ national statistical system: Biology, Chemistry, Engineering, Physics, Materials Science, Environmental Science, Geology, History, Philosophy, Math, Computer Science, and Art. Fields that _are_ included are: Economics, Political Science, Business, Sociology, Medicine, and Psychology. This third restriction eliminated around 34% of the remaining articles. From an initial corpus of 136 million articles, this resulted in a final corpus of around 10 million articles.\\n\\nDue to the intensive computer resources required, a set of 1,037,748 articles were randomly selected from the 10 million articles in our restricted corpus as a convenience sample. Summary statistics of the final sample of 1 million articles are available in Table 1.\\n\\n_Table 1. Summary Statistics of Article Corpus. 2000-2020_\\n\\n|Field|Published in Journal<br>(1=yes)|Data Use<br>(1=yes)|Country Identified<br>(1=yes)|Articles|Share of<br>Articles|\\n|---|---|---|---|---|---|\\n|Business|0.56|0.64|0.30|28,571|2.8|\\n|Economics|0.79|0.68|0.28|62,241|6.0|\\n|Medicine|0.96|0.85|0.10|840,920|81.0|\\n|Political Science|0.42|0.33|0.34|26,185|2.5|\\n|Psychology|0.75|0.70|0.14|44,191|4.3|\\n|Sociology|0.90|0.33|0.25|35,640|3.4|\\n\\n4\", \"datasets\": [{\"mention_name\": {\"text\": \"Semantic Scholar Open Research Corpus\", \"confidence\": 0.687567949295044, \"start\": 336, \"end\": 373, \"clean_text\": \"Semantic Scholar Open Research Corpus\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.7122364640235901, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 336, \"end\": 373, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 336, \"end\": 373}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 336, \"end\": 373, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9544560313224792, \"start\": 336, \"end\": 373}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.687567949295044, \"start\": 336, \"end\": 373}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.844089925289154, \"start\": 336, \"end\": 373}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.844089925289154, \"start\": 336, \"end\": 373}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"# **3 Empirical Strategy**\\n\\nThe empirical approach employed in this project utilizes text mining with natural language processing (NLP). The goal of NLP is to extract structured information from raw, unstructured text. In this project, NLP is used to extract the country of study and whether the paper makes use of data. We will discuss each of these in turn.\\n\\nTo determine the country or countries of study in each academic article, two approaches are employed based on information found in the title, abstract, or topic fields. The first approach uses regular expression searches based on the presence of ISO3166 country names. A defined set of country names is compiled, and the presence of these names is checked in the relevant fields. This approach is transparent, widely used in social science research, and easily extended to other languages. However, there is a potential for exclusion errors if a country’s name is spelled nonstandardly.\\n\\nThe second approach is based on Named Entity Recognition (NER), which uses machine learning to identify objects from text, utilizing the spaCy Python library. The Named Entity Recognition algorithm splits text into named entities, and NER is used in this project to identify countries of study in the academic articles. SpaCy supports multiple languages and has been trained on multiple spellings of countries, overcoming some of the limitations of the regular expression approach. If a country is identified by either the regular expression search or NER, it is linked to the article. Note that one article can be linked to more than one country.\\n\\nThe second task is to classify whether the paper uses data. A supervised machine learning approach is employed, where 3,500 publications were first randomly selected and manually labeled by human raters using the Mechanical Turk service (Paszke et al. 2019).<sup>2</sup> To make sure the human raters had a similar and appropriate definition of data in mind, they were given the following instructions before seeing their first paper: _<mark>Each of these documents is an academic article. The goal of this study is to measure whether a specific academic article is using data and from which country the data came.</mark>_ _<mark>There are two classification tasks in this exercise:</mark>_ _1._ _<mark>Identifying whether an academic article is using data from any country</mark>_ _2._ _<mark>Identifying from which country that data came.</mark>_ _<mark>For task 1, we are looking specifically at the use of data. Data is any information that has been collected, observed, generated or created to produce research findings. As an example, a study that reports findings or analysis using a survey data, uses data. Some clues to</mark>_\\n\\n> 2 A stratified sample of articles for the training set was taken. Of the articles 15% were medical, 15% psychology, 25% economics, 25% political science, 10% were tagged as business, and 10% sociology articles.\\n\\n5\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"_<mark>indicate that a study does use data includes whether a survey or census is described, a statistical model estimated, or a table or means or summary statistics is reported.</mark>_ _<mark>After an article is classified as using data, please note the type of data used. The options are population or business census, survey data, administrative data, geospatial data, private sector data, and other data. If no data is used, then mark \\\"Not applicable\\\". In cases where multiple data types are used, please click multiple options.</mark>_<sup>_3_</sup> _<mark>For task 2, we are looking at the country or countries that are studied in the article. In some cases, no country may be applicable. For instance, if the research is theoretical and has no specific country application. In some cases, the research article may involve multiple countries. In these cases, select all countries that are discussed in the paper</mark>_ <mark>.</mark>\\n\\n# _<mark>We expect between 10 and 35 percent of all articles to use data.</mark>_\\n\\nAn image of the screen facing the MTurk workers when classifying an article is presented in Figure A.1 in the Appendix. The median amount of time that a worker spent on an article, measured as the time between when the article was accepted to be classified by the worker and when the classification was submitted was 25.4 minutes. If human raters were exclusively used rather than machine learning tools, then the corpus of 1,037,748 articles examined in this study would take around 50 years of human work time to review at a cost of $3,113,244, which assumes a cost of $3 per article as was paid to MTurk workers.\\n\\nA model is next trained on the 3,500 labeled articles. We use a distilled version of the BERT (bidirectional Encoder Representations for transformers) model to encode raw text into a numeric format suitable for predictions (Devlin et al. (2018)). BERT is pre-trained on a large corpus comprising the Toronto Book Corpus and Wikipedia. The distilled version (DistilBERT) is a compressed model that is 60% the size of BERT and retains 97% of the language understanding capabilities and is 60% faster (Sanh, Debut, Chaumond, Wolf 2019). We use PyTorch to produce a model to classify articles based on the labeled data. Of the 3,500 articles that were hand coded by the MTurk workers, 900 are fed to the machine learning model. Nine hundred articles were selected because of computational limitations in training the NLP model. A classification of “uses data” was assigned if the model predicted an article used data with at least 90% confidence.\\n\\nThe performance of the models classifying articles to countries and as using data or not can be compared to the classification by the human raters. We consider the human raters as giving us the ground truth. This may underestimate the model performance if the workers at times got the allocation wrong in a way that would not apply to the model. For instance, a human rater could mistake the Republic of Korea for the Democratic People’s Republic of Korea. If both humans and the model perform the same kind of errors, then the performance reported here will be overestimated.\\n\\n> 3 We ended up not using this in the paper.\\n\\n6\", \"datasets\": [{\"mention_name\": {\"text\": \"population or business census\", \"confidence\": 0.6650468111038208, \"start\": 291, \"end\": 320, \"clean_text\": \"population or business census\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 291, \"end\": 320, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 291, \"end\": 320, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 291, \"end\": 320}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 291, \"end\": 320, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.0, \"start\": 291, \"end\": 320}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6650468111038208, \"start\": 291, \"end\": 320}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4557607173919678, \"start\": 291, \"end\": 320}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4557607173919678, \"start\": 291, \"end\": 320}}, {\"mention_name\": {\"text\": \"survey data\", \"confidence\": 0.7223150134086609, \"start\": 322, \"end\": 333, \"clean_text\": \"survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 322, \"end\": 333, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 322, \"end\": 333, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 322, \"end\": 333}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 322, \"end\": 333, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.0, \"start\": 322, \"end\": 333}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7223150134086609, \"start\": 322, \"end\": 333}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4557607173919678, \"start\": 322, \"end\": 333}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4557607173919678, \"start\": 322, \"end\": 333}}, {\"mention_name\": {\"text\": \"administrative data\", \"confidence\": 0.4806603789329529, \"start\": 335, \"end\": 354, \"clean_text\": \"administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 335, \"end\": 354, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 335, \"end\": 354, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 335, \"end\": 354}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 335, \"end\": 354, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.0, \"start\": 335, \"end\": 354}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4806603789329529, \"start\": 335, \"end\": 354}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4557607173919678, \"start\": 335, \"end\": 354}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4557607173919678, \"start\": 335, \"end\": 354}}, {\"mention_name\": {\"text\": \"geospatial data\", \"confidence\": 0.575215220451355, \"start\": 356, \"end\": 371, \"clean_text\": \"geospatial data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 356, \"end\": 371, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 356, \"end\": 371, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 356, \"end\": 371}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 356, \"end\": 371, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.0, \"start\": 356, \"end\": 371}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.575215220451355, \"start\": 356, \"end\": 371}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4557607173919678, \"start\": 356, \"end\": 371}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4557607173919678, \"start\": 356, \"end\": 371}}, {\"mention_name\": {\"text\": \"private sector data\", \"confidence\": 0.7185405492782593, \"start\": 373, \"end\": 392, \"clean_text\": \"private sector data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 373, \"end\": 392, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 373, \"end\": 392, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 373, \"end\": 392}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 373, \"end\": 392, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.4193874001502991, \"start\": 373, \"end\": 392}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7185405492782593, \"start\": 373, \"end\": 392}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4557607173919678, \"start\": 373, \"end\": 392}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4557607173919678, \"start\": 373, \"end\": 392}}, {\"mention_name\": {\"text\": \"Toronto Book Corpus\", \"confidence\": 0.8286852240562439, \"start\": 1943, \"end\": 1962, \"clean_text\": \"Toronto Book Corpus\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1943, \"end\": 1962, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1943, \"end\": 1962, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1943, \"end\": 1962}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1943, \"end\": 1962, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.991349458694458, \"start\": 595, \"end\": 614}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8286852240562439, \"start\": 595, \"end\": 614}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6370512247085571, \"start\": 595, \"end\": 614}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6370512247085571, \"start\": 595, \"end\": 614}}, {\"mention_name\": {\"text\": \"labeled data\", \"confidence\": 0.30249959230422974, \"start\": 2245, \"end\": 2257, \"clean_text\": \"labeled data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2245, \"end\": 2257, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2245, \"end\": 2257, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2245, \"end\": 2257}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2245, \"end\": 2257, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.9384605288505554, \"start\": 897, \"end\": 909}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.30249959230422974, \"start\": 897, \"end\": 909}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8813068270683289, \"start\": 897, \"end\": 909}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8813068270683289, \"start\": 897, \"end\": 909}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"<!-- Start of picture text --> United States<br>00 y=-0.3+ 13x, R2=0.996<br>(02) (00)<br>i<br>5 India United im<br>& Canada alia<br>2.<br>o8 XY 5<br>° Tance<br>5 10 be<br>= Mexico<br>Syrian Arab Republic Port- ort u 2egc a ll Pfe)* °¢ ° dan<br>United Arab Emirates-Libya . Korea, Rep.<br>Slovenia-Somalia Chile<br>Sri Lani ° °<br>Lebanon<br>4 » °Nigerig<br>1 10 100<br>Machine Learning Classification<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"<!-- Start of picture text --> fGapetglue<br>em N\\\\\\\\ Au VAN) Ay WAAAY YY \\\\ ltd 7 Hi wi 93 ___<br>This 6888). which is part of a sfmiposiumledition of the Akron Law Review, explores the practice of metadata<br>mining (i.e., the surfeptitious examination of an electronic documenta€TMs hidden a. The essay (EES<br>the (B@BBethics Spiniohs that have addressed the practice and contends that metadatalmining is simply a<br>variation ofbans the oft-QEB problem of inadvertently disclosed documents. The essay concludes thabifiat<br>inadvertent on meta@ata}miningdisclosures moreare misguided generally. andUnder that this the approach,issue shouldif a statebe treated permitsin thelawyers same to manner FEVIEW as<br>inadvertently disclosed privileged documents, the jurisdiction should also permit lawyers to review the<br>meta@iataicontained in electronic documents. In contrast, if a jurisdiction prohibits the feviewlof misdirected<br>privileged documents, the state should ban metadata mining, but only when recipients have reason to<br>believe that the metadata contains protected information.<br><!-- End of picture text --> <!-- Start of picture text --> ee ees inout \\\" \\\"<br>Objectivasifelestablish 5(ii@and reliable translated version of Smartphone Addiction Scale for adolescent<br>[SAS-SV) in Italian languages is a 10-item) monodimensional, self-administered questionnaire, Which Was<br>validated using 2 large Korean SSmple§Methods:We reported the linguistic validation process and the<br>youngmetric adults\\\\frecruited validity ofthe Italian from version High Public of SAS-SV School in the ItalyTheiid)UniversityEBiapIBin Italy,consistedwith a Meanlageof 633 adolescentsof 18 years{iThe<br>validation process consisted in forward-backward translation {factorstructure testing)Results: Results<br>(BRIER: g000 FBliability (Cronbachs alpha = .75) Sian almost acceptable factor structuresConclusion: A<br>(lid)3nd) reliable Smartphone Addiction Scale in Sho Version Was developed which can be used for<br>fé5earch)and clinical assessment Ofadolescentsiandyoung adults with smartphone addiction|<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> —04 yy sat — baseoot value 05 0.673803f(inputs) 28 it 14<br>Sy»)) inputsscram —_ ((<br>The article discusses theoretical aspects of integration approach to the {of the development of the<br>agrarian sphere of economy. The article presents authors who have made significant scientific contributions<br>to the theory of integration; identifies structural elements of the national agro-industrial complex:<br>determines the value of scientific and technological progress in the development and integration of<br>productive forces; identifies factors affecting the development of the agrarian sphere as a complex<br>integrated system. The article notes that integration as a system possesses characteristics of self-<br>organization, hierarchy, striving for equilibrium, and has the presence of a common goal. The article<br>particularly characterizes the socio-economic factors of integration, influencing the agricultural sphere<br>development and management. Particular emphasis is placed on the problem of integration of Russian<br>agricultural industries in the WTO.<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"|J=BIN<br>)J-C<br>)Vij<br>()|\\n|---|\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"<!-- Start of picture text --> 4 0°\\\". |<br>a aS<br>Papers \\\\<br>= 10 ’ Ud PA<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"<!-- Start of picture text --> High income 71,815<br>_— |“<br>Low income 5,517<br>0 25,000 50,000 75,000<br>Number of Papers using Data<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> ed<br>Suope ScenaAS Ls<br>ee (iii)i<br>a Lda<br>‘ate Ares & careoean |TM<br>Sub-Saharan Africa | 143.1<br>sauA -<br>0 100 200 300 400 500<br>Number of Papers using Data per capita<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"<!-- Start of picture text --> CHN ° USA<br>1,000 NGA, IND . IRN-2—$°IDN ,BRAZAF mys, ITASCAN®JPN ¢>-DEU Lp.AUS<br>g ETH PAK THAN MEX wTUR ee ee NOR<br>8 UGA KEN*® “3BGD VNM cot * RUS FOL. SAU a<br>2 RWA 1A\\\" NPLeGHA JOR\\\\PHL *~ £GY KA CHL RC * BEND a cp<br>53 MMR MAR PER, o.° * PRTQs %e<br>2 100 NER MW! E . ZMB eSDN YKHM NRQ—* . * IKR_GEOARG ROU CHE<br>goO con—+ yo, AFG SEN * ° gO gL AQUTUNSECU oee 5 SRN RV agree...HUN CZEt1St, OAT<br>& MOZ“Teq SNe BEN=GO ae eRiRiy=-0.49 + 0.50 x, R2= 0.12<br>3 LBR——*—BFarh HY) NIC SWZCIM*s poh s—LB (0.86) (0.09)<br>8 pp Som. SE 7 SLB PNG \\\"asd. SLV BINN. *BLR, —TTO \\\"BHR ux®<br>=— 10 Tcp *KIR® GMBLo\\\" . TuKSq}gVUT vorSM UZBf Btz°°? MDAGuy’caB|sURARM, MOV—lMNE* “—~BHSsyc<br>FSM“—4/ WSM veT* mus<br>MRT “KM<br>LCA<br>| cnB° MHL° CPV®<br>1,000 10,000 100,000<br>GDP per capita, PPP (constant 2017 international $)<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> USA CHN °<br>AUS<br>1,000 mys, CAN© DEUSIRN *JPN % BRA_ex ip) 3 IND<br>g& NZL *SWENOR NLD, Sau. eae3. 89 eM te PAK<br>8fo.) IRL tp A ca MeTza “itis 8D<br>23 FIN PR’te N PER NBLS COL RUS<br>5 100 HRV . LKA Q3SARG PHL ~EGY<br>g© . EST\\\\. GEO,° BUNGE MAR «‘SoRENMR-—SDNEMIMR-UKR<br>¢ ISL 7 FJIXCYPPesVN ee Peesea 4”. Bly =-6.60A + 0.67anx, R= 0.59<br>3 vuTLUX*.,Sup2Gme” SLV, TEER cow ize<br>€ —. a s 0 “MD, f BDie .- e?<br>=5 10 SYC,FSM °° KIR % BLZ_-vsSUR 2,{tS3TTO,‘MUS’ GABcoz M Tép “civ<br>< M GUY ° MRT TJK<br>v pi =1SO<br>. TKM<br>MAL LCA cPv . GNB .<br>100,000 1,000,000 10,000,000 100,000,000 1,000,000,000<br>Population, Total<br><!-- End of picture text --> <!-- Start of picture text --> CHN® JPN USA<br>1,000 rawokPAK SAU mpmys_*ZAF tO *BRA GBR,ganitaAUS<br>© VNM eo ° THA TyRLOEY yor<br>® ETH e, BED \\\\\\\\usa ERY brus “Cie,<br>3 KEN ‘GHAeEOb cohSeP* ROW<br>& NPL TZA jon \\\\PER Soy ge? *PRT<br>8 IRQ AFG wvR——e HRV AoePEL ze” AUT<br>2 100 . SDN MRS ‘SRB GEO<br>2 cop . NBN, 28% 2wetY Ut\\\" eisy * HUN, SVK<br>S * .°PNG O08BEN » 2% 6 *RWA © MNG, GRe “Sj SVN<br>&633 ol BY cosHTy=y = 0.52+ 0.06 x,*120 R2=0.36- HReT e pses URpy .PSEsKD %.iatCYP,“ant*.~CR VA EST<br>M<br>€ som (0.37) (0.01) HND DOM aze ALB LuxS<br>3 ° KIR~SAPAD TCD VU TO e Nciv8CO,-BLZinetstv ° MLT af\\\\ NDAARM<br>2 Guy MRT ‘SUR TUK? Ney MUS éz<br>FSM. Dil . vet * sYC_ Lso<br>TKM LCA<br>1 MHL . .GNB QoPV<br>20 40 60 80<br>SPI Overall Score<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"_Note_ : Data on GDP, population and SPI are retrieved from the World Bank’s World Development Indicators For GDP per capita, the figure indicates an elasticity of 0.5, meaning that a 10% increase in GDP is associated with a 5% increase in academic articles using data. The elasticity with respect to population is 0.67, and population alone explains 59% of the variation in academic articles using data. The overall SPI score is also predictive of academic output using data. A 10-point increase in SPI scores, which is approximately the same as moving from the median SPI score to the 65th percentile, translates into around a 0.6% increase in the number of articles using data.\\n\\nTable 2 shows coefficients from a regression of the log number of academic papers using data on log GDP per capita, log population, and the SPI scores. The first column shows a specification including just log GDP per capita and log population. These two indicators alone explain roughly 75% of the variation in the production of articles using data. The second column includes the SPI overall score as an additional control. The performance of a country’s statistical system is associated with greater academic output. Conditional on GDP per capita and population, a tenpoint increase in SPI overall scores (on a scale of 0-100) translates into a 0.2% increase in academic output.\\n\\nIt is possible that omitted variables are behind this result. For example, it could be the case that countries that are of general interest to the research community also happen to have high statistical performance. Das et al. (2013) and Porteous (2022) control for a range of variables which may influence the relationship sought after, such as whether English is a first language, international tourist arrivals, etc. We leverage that we have an estimate of the number of articles not using data – referred to as qualitative articles – to control for all factors that are likely to influence overall research output in a country unrelated to statistical performance. The number of qualitative articles is calculated as the number of total articles minus any articles using data, based on our NLP model.\\n\\nCountries that tend to be over-represented in quantitative articles also tend to be overrepresented in qualitative articles, conditional on GDP per capita and population (Figure A.3). The elasticity is around 0.6, suggesting that a 1% increase in the number of qualitative articles associated with a 0.6% increase in the number of articles using data. We take this to imply that quantitative articles do not crowd out qualitative articles and vice versa. Returning to Equation (2), one prediction from the model is that countries with particularly high costs related to using data, (such as a lower statistical performance score), should see reductions in the amount of quantitative research, but it would not follow that it would see lower levels of qualitative ii XX research. This would predict the share of quantitative articles to total articles increasing in SPI score. Contrary to this, there is almost no relationship between SPI overall scores and the share of qualitative papers to total papers produced overall (Figure A.4). This suggests to us that some omitted variable may in part be driving the impact of statistical performance on academic articles using data, and that the number of qualitative articles can control for such possibly omitted factors.\\n\\n13\", \"datasets\": [{\"mention_name\": {\"text\": \"World Development Indicators\", \"confidence\": 0.5096678733825684, \"start\": 77, \"end\": 105, \"clean_text\": \"World Development Indicators\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 77, \"end\": 105, \"clean_text\": \"\"}, \"producer\": {\"text\": \"World Bank\", \"confidence\": 0.3015117645263672, \"start\": 64, \"end\": 74, \"clean_text\": \"World Bank\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 77, \"end\": 105}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 77, \"end\": 105, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9796669483184814, \"start\": 77, \"end\": 105}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5096678733825684, \"start\": 77, \"end\": 105}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6106045246124268, \"start\": 77, \"end\": 105}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6106045246124268, \"start\": 77, \"end\": 105}}, {\"mention_name\": {\"text\": \"SPI scores\", \"confidence\": 0.37969115376472473, \"start\": 822, \"end\": 832, \"clean_text\": \"SPI scores\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 822, \"end\": 832, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 822, \"end\": 832, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 822, \"end\": 832}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 822, \"end\": 832, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.5484216809272766, \"start\": 58, \"end\": 68}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.37969115376472473, \"start\": 58, \"end\": 68}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7850683927536011, \"start\": 58, \"end\": 68}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7850683927536011, \"start\": 58, \"end\": 68}}, {\"mention_name\": {\"text\": \"qualitative articles\", \"confidence\": 0.5398549437522888, \"start\": 2048, \"end\": 2068, \"clean_text\": \"qualitative articles\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2048, \"end\": 2068, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2048, \"end\": 2068, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2048, \"end\": 2068}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2048, \"end\": 2068, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.7560797333717346, \"start\": 480, \"end\": 500}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5398549437522888, \"start\": 480, \"end\": 500}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8453448414802551, \"start\": 480, \"end\": 500}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8453448414802551, \"start\": 480, \"end\": 500}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"The third column of Table 2 additionally controls for the log number of qualitative articles. The SPI overall scores are still statistically significant at the 5% level, as is country population. Column 4 examines the relationship with one of the pillars of the SPI, data sources, which covers the availability of recent censuses, surveys, academic data, and geospatial data. The data sources index from the SPI is also correlated with data use in academia, conditional on log GDP, population, and the number of qualitative papers. Similar regressions using the other pillars of the SPI, such as data services and data infrastructure are not associated with the number of articles using data in a similar regression, suggesting that it is the availability of particular data sources that is helpful in boosting academic data use.\\n\\n_Table 2. Relationships between Number of Papers Using Data and Statistical Performance_\\n\\n||(1)|(2)|(3)|(4)|\\n|---|---|---|---|---|\\n|(Intercept)|-12.41***|-10.66***|-3.96***|-3.57**|\\n||(0.82)|(0.95)|(1.14)|(1.17)|\\n|Log GDP per capita|0.58***|0.36***|0.12+|0.10|\\n||(0.05)|(0.07)|(0.07)|(0.08)|\\n|Log Population|0.70***|0.63***|0.29***|0.29***|\\n||(0.03)|(0.04)|(0.05)|(0.05)|\\n|SPI Overall Score||0.02***|0.01*||\\n|||(0.01)|(0.01)||\\n|Log Qualitative Papers|||0.64***|0.65***|\\n||||(0.08)|(0.08)|\\n|SPI Data Sources Score||||0.01*<br>(0.00)|\\n|Observations|168|168|168|168|\\n|R2|0.751|0.774|0.852|0.852|\\n|R2 Adj.|0.748|0.770|0.849|0.849|\\n|AIC|414.8|400.1|331.1|330.7|\\n|BIC|424.2|412.6|346.8|346.3|\\n|RMSE<br>_Note:_GDP, population and S|0.82<br>PI data are retr|0.78<br>ieved from th|0.63<br>e World Ba|0.63<br>nk's World|\\n\\n_Note:_ GDP, population and SPI data are retrieved from the World Bank's World Development Indicators. Regressions include all papers using data from 20172019. Standard errors are robust to heteroskedasticity. ***=0.001 level, **=0.01 level, *=0.05 level, +=0.1 level.\\n\\nUsing more detailed data from the SPI makes it possible to assess how the availability of specific data sources relates to academic output. Figure 6 shows the linear regression coefficient for the availability of each of the ten data sources considered by the SPI, conditional on log GDP per capita, log population, and log of qualitative papers. The ten data sources include: population census, agriculture census, business/establishment census, household consumption/income survey, agriculture survey, labor force survey, health survey, business/establishment survey, civil registration and vital statistics system (CRVS), and geospatial data at the Admin 1 (usually province/state) level.\\n\\n14\", \"datasets\": [{\"mention_name\": {\"text\": \"PI data\", \"confidence\": 0.3739616274833679, \"start\": 1564, \"end\": 1571, \"clean_text\": \"PI data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1564, \"end\": 1571, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1564, \"end\": 1571, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1564, \"end\": 1571}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1564, \"end\": 1571, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"microdata\", \"confidence\": 0.30698251724243164, \"start\": 733, \"end\": 740}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3739616274833679, \"start\": 733, \"end\": 740}, \"usage_context\": {\"text\": \"background\", \"confidence\": 0.4325276017189026, \"start\": 733, \"end\": 740}, \"is_used\": {\"text\": \"False\", \"confidence\": 0.4325276017189026, \"start\": 733, \"end\": 740}}, {\"mention_name\": {\"text\": \"World Development Indicators\", \"confidence\": 0.468847393989563, \"start\": 1716, \"end\": 1744, \"clean_text\": \"World Development Indicators\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1716, \"end\": 1744, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1716, \"end\": 1744, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1716, \"end\": 1744}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1716, \"end\": 1744, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9642064571380615, \"start\": 189, \"end\": 217}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.468847393989563, \"start\": 189, \"end\": 217}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.762802243232727, \"start\": 189, \"end\": 217}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.762802243232727, \"start\": 189, \"end\": 217}}, {\"mention_name\": {\"text\": \"population census\", \"confidence\": 0.303668349981308, \"start\": 2290, \"end\": 2307, \"clean_text\": \"population census\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2290, \"end\": 2307, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2290, \"end\": 2307, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2290, \"end\": 2307}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2290, \"end\": 2307, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.9985955357551575, \"start\": 763, \"end\": 780}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.303668349981308, \"start\": 763, \"end\": 780}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6201784610748291, \"start\": 763, \"end\": 780}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6201784610748291, \"start\": 763, \"end\": 780}}, {\"mention_name\": {\"text\": \"agriculture census\", \"confidence\": 0.9149410724639893, \"start\": 2309, \"end\": 2327, \"clean_text\": \"agriculture census\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2309, \"end\": 2327, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2309, \"end\": 2327, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2309, \"end\": 2327}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2309, \"end\": 2327, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.625907838344574, \"start\": 203, \"end\": 221}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.9149410724639893, \"start\": 203, \"end\": 221}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.649949312210083, \"start\": 203, \"end\": 221}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.649949312210083, \"start\": 203, \"end\": 221}}, {\"mention_name\": {\"text\": \"business/establishment census\", \"confidence\": 0.8990249037742615, \"start\": 2329, \"end\": 2358, \"clean_text\": \"business/establishment census\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2329, \"end\": 2358, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2329, \"end\": 2358, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2329, \"end\": 2358}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2329, \"end\": 2358, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.625907838344574, \"start\": 223, \"end\": 252}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8990249037742615, \"start\": 223, \"end\": 252}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.649949312210083, \"start\": 223, \"end\": 252}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.649949312210083, \"start\": 223, \"end\": 252}}, {\"mention_name\": {\"text\": \"household consumption/income survey\", \"confidence\": 0.8885161280632019, \"start\": 2360, \"end\": 2395, \"clean_text\": \"household consumption/income survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2360, \"end\": 2395, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2360, \"end\": 2395, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2360, \"end\": 2395}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2360, \"end\": 2395, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.625907838344574, \"start\": 254, \"end\": 289}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8885161280632019, \"start\": 254, \"end\": 289}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.649949312210083, \"start\": 254, \"end\": 289}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.649949312210083, \"start\": 254, \"end\": 289}}, {\"mention_name\": {\"text\": \"agriculture survey\", \"confidence\": 0.8575627207756042, \"start\": 2397, \"end\": 2415, \"clean_text\": \"agriculture survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2397, \"end\": 2415, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2397, \"end\": 2415, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2397, \"end\": 2415}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2397, \"end\": 2415, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.625907838344574, \"start\": 291, \"end\": 309}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8575627207756042, \"start\": 291, \"end\": 309}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.649949312210083, \"start\": 291, \"end\": 309}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.649949312210083, \"start\": 291, \"end\": 309}}, {\"mention_name\": {\"text\": \"labor force survey\", \"confidence\": 0.8499384522438049, \"start\": 2417, \"end\": 2435, \"clean_text\": \"labor force survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2417, \"end\": 2435, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2417, \"end\": 2435, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2417, \"end\": 2435}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2417, \"end\": 2435, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.625907838344574, \"start\": 311, \"end\": 329}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8499384522438049, \"start\": 311, \"end\": 329}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.649949312210083, \"start\": 311, \"end\": 329}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.649949312210083, \"start\": 311, \"end\": 329}}, {\"mention_name\": {\"text\": \"health survey\", \"confidence\": 0.8690361976623535, \"start\": 2437, \"end\": 2450, \"clean_text\": \"health survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2437, \"end\": 2450, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2437, \"end\": 2450, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2437, \"end\": 2450}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2437, \"end\": 2450, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.625907838344574, \"start\": 331, \"end\": 344}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8690361976623535, \"start\": 331, \"end\": 344}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.649949312210083, \"start\": 331, \"end\": 344}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.649949312210083, \"start\": 331, \"end\": 344}}, {\"mention_name\": {\"text\": \"business/establishment survey\", \"confidence\": 0.934510350227356, \"start\": 2452, \"end\": 2481, \"clean_text\": \"business/establishment survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2452, \"end\": 2481, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2452, \"end\": 2481, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2452, \"end\": 2481}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2452, \"end\": 2481, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.625907838344574, \"start\": 346, \"end\": 375}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.934510350227356, \"start\": 346, \"end\": 375}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.649949312210083, \"start\": 346, \"end\": 375}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.649949312210083, \"start\": 346, \"end\": 375}}, {\"mention_name\": {\"text\": \"civil registration and vital statistics system\", \"confidence\": 0.4981077015399933, \"start\": 2483, \"end\": 2529, \"clean_text\": \"civil registration and vital statistics system\"}, \"acronym\": {\"text\": \"CRVS\", \"confidence\": 0.4119051396846771, \"start\": 2531, \"end\": 2535, \"clean_text\": \"CRVS\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2483, \"end\": 2529, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2483, \"end\": 2529}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2483, \"end\": 2529, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.6259080171585083, \"start\": 377, \"end\": 423}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4981077015399933, \"start\": 377, \"end\": 423}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6499490737915039, \"start\": 377, \"end\": 423}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6499490737915039, \"start\": 377, \"end\": 423}}, {\"mention_name\": {\"text\": \"geospatial data\", \"confidence\": 0.7225856184959412, \"start\": 2542, \"end\": 2557, \"clean_text\": \"geospatial data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2542, \"end\": 2557, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2542, \"end\": 2557, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2542, \"end\": 2557}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2542, \"end\": 2557, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.625907838344574, \"start\": 436, \"end\": 451}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7225856184959412, \"start\": 436, \"end\": 451}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.649949312210083, \"start\": 436, \"end\": 451}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.649949312210083, \"start\": 436, \"end\": 451}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"<!-- Start of picture text --> Population Census in past 10 years<br>Agriculture Census in past 10 years<br>Business/Establishment Census in past 10 years<br>Household surveys: 2 or more in 10 years<br>Agriculture Surveys: 2 or more in 10 years —_—<br>Labor Force Surveys: 2 or more in 10 years ad<br>Health Surveys: 2 or more in 10 years<br>Business/Establishments: 2 or more in 10 years ——_—_-<br>Complete Civil Registration and Vital Statistics System ‘2<br>Availability of Data at ist Admin Level (ODIN) Score — 7 > - ——<br>0 1 2<br>Coefficient estimates and 95% confidence intervals<br><!-- End of picture text -->\", \"datasets\": [{\"mention_name\": {\"text\": \"Population Census in past 10 years\", \"confidence\": 0.658663272857666, \"start\": 31, \"end\": 65, \"clean_text\": \"Population Census in past 10 years\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 31, \"end\": 65, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 31, \"end\": 65, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 31, \"end\": 65}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 31, \"end\": 65, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.821201503276825, \"start\": 31, \"end\": 65}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.658663272857666, \"start\": 31, \"end\": 65}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.43239307403564453, \"start\": 31, \"end\": 65}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.43239307403564453, \"start\": 31, \"end\": 65}}, {\"mention_name\": {\"text\": \"Agriculture Census\", \"confidence\": 0.5468803644180298, \"start\": 69, \"end\": 87, \"clean_text\": \"Agriculture Census\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 69, \"end\": 87, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 69, \"end\": 87, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 69, \"end\": 87}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 69, \"end\": 87, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.821201503276825, \"start\": 69, \"end\": 87}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5468803644180298, \"start\": 69, \"end\": 87}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.43239307403564453, \"start\": 69, \"end\": 87}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.43239307403564453, \"start\": 69, \"end\": 87}}, {\"mention_name\": {\"text\": \"Business/Establishment Census\", \"confidence\": 0.6225113272666931, \"start\": 108, \"end\": 137, \"clean_text\": \"Business/Establishment Census\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 108, \"end\": 137, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 108, \"end\": 137, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 108, \"end\": 137}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 108, \"end\": 137, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.821201503276825, \"start\": 108, \"end\": 137}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6225113272666931, \"start\": 108, \"end\": 137}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.43239307403564453, \"start\": 108, \"end\": 137}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.43239307403564453, \"start\": 108, \"end\": 137}}, {\"mention_name\": {\"text\": \"Household surveys\", \"confidence\": 0.3487548232078552, \"start\": 158, \"end\": 175, \"clean_text\": \"Household surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 158, \"end\": 175, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 158, \"end\": 175, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 158, \"end\": 175}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 158, \"end\": 175, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.821201503276825, \"start\": 158, \"end\": 175}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3487548232078552, \"start\": 158, \"end\": 175}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.43239307403564453, \"start\": 158, \"end\": 175}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.43239307403564453, \"start\": 158, \"end\": 175}}, {\"mention_name\": {\"text\": \"Agriculture Surveys\", \"confidence\": 0.471158504486084, \"start\": 202, \"end\": 221, \"clean_text\": \"Agriculture Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 202, \"end\": 221, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 202, \"end\": 221, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 202, \"end\": 221}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 202, \"end\": 221, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.821201503276825, \"start\": 202, \"end\": 221}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.471158504486084, \"start\": 202, \"end\": 221}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.43239307403564453, \"start\": 202, \"end\": 221}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.43239307403564453, \"start\": 202, \"end\": 221}}, {\"mention_name\": {\"text\": \"Labor Force Surveys\", \"confidence\": 0.5691205263137817, \"start\": 252, \"end\": 271, \"clean_text\": \"Labor Force Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 252, \"end\": 271, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 252, \"end\": 271, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 252, \"end\": 271}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 252, \"end\": 271, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.821201503276825, \"start\": 252, \"end\": 271}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5691205263137817, \"start\": 252, \"end\": 271}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.43239307403564453, \"start\": 252, \"end\": 271}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.43239307403564453, \"start\": 252, \"end\": 271}}, {\"mention_name\": {\"text\": \"Health Surveys\", \"confidence\": 0.5094643235206604, \"start\": 301, \"end\": 315, \"clean_text\": \"Health Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 301, \"end\": 315, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 301, \"end\": 315, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 301, \"end\": 315}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 301, \"end\": 315, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.821201503276825, \"start\": 301, \"end\": 315}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5094643235206604, \"start\": 301, \"end\": 315}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.43239307403564453, \"start\": 301, \"end\": 315}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.43239307403564453, \"start\": 301, \"end\": 315}}, {\"mention_name\": {\"text\": \"Complete Civil Registration and Vital Statistics System\", \"confidence\": 0.5874813795089722, \"start\": 399, \"end\": 454, \"clean_text\": \"Complete Civil Registration and Vital Statistics System\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 399, \"end\": 454, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 399, \"end\": 454, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 399, \"end\": 454}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 399, \"end\": 454, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.821201503276825, \"start\": 399, \"end\": 454}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5874813795089722, \"start\": 399, \"end\": 454}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.43239307403564453, \"start\": 399, \"end\": 454}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.43239307403564453, \"start\": 399, \"end\": 454}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"literacy among researchers, low trust in data from the national statistical system, and lack of infrastructure to access and use the data, which may be particularly an issue of low- and middleincome countries.\\n\\nThe policies needed to boost data supply differ from those needed to boost data demand. Boosting the data supply of the national statistical system requires more financing for national statistical offices, better technical capacity of its staff, and at times better statistical laws to ensure their independence from other government bodies. Policies to boost data demand, on the other hand, involve financially cheap (but not necessarily politically cheap) wins, such as making data more accessible for researchers. It also involves increasing the data literacy of academics. Given the non-overlapping policy responses, it is useful for countries to know whether data demand or data supply is their weakest link in the data value chain. Below, we classify countries by whether they are primarily constrained by data supply, data demand, neither or both.\\n\\nBuilding on the semantics introduced by Porteous (2020) who classified countries into research deserts and research oases, we classify countries into four groups: _Data deserts_ have little data demand and little data supply, _data swamps_ have high data supply but little data demand, _data oases_ have high data demand but little data supply, and _data lakes_ have high data demand and high data supply. For assessing data supply, we again turn to the availability of various data sources, such as recent population and business censuses, household and health surveys, Civil Registration and Vital Statistics (CRVS), and sub-national geospatial data (for details see Dang et al., 2023). We measure data use as the residual from the regression of number of articles on the country using data on log population, log GDP per capita, and the average number of qualitative papers. The control for the number of qualitative papers once again serves as a proxy for general research interest and research environment unrelated to data.\\n\\nWe classify countries by whether they are below/above the median of our indicators of data supply and data use (Figure 7). Since we use median values in the two dimensions to create the categories, there are about as many countries belonging to each group. Yet there are large discrepancies by region and income group (Table 3). The poorest countries and poorest regions are more likely to be data deserts or data oases. These categories apply to more than 80% of lowand lower-middle income countries, 93% of countries in Sub-Saharan Africa, and only 7% of highincome countries. Thus, poorer countries and poorer countries tend to have lower data supply.\\n\\nLow-income countries and countries in Sub-Saharan Africa are twice as likely to be oases than deserts. In fact, when looking at the map, it is evident that a large part of the world’s data lakes, are to be found in Sub-Saharan Africa. This suggests that, despite data scarcity, a significant volume of quantitative articles is already produced on these countries, when factoring in their population size, income, and attention in non-quantitative scholarship. These countries may benefit most from bolstering their data supply. This applies for example to Uganda, Ghana, Nepal, and Malawi. By contrast, nearly all upper-middle income countries and countries in Latin America who have low data supply are deserts rather than oasis. That is, they face both data 16\", \"datasets\": [{\"mention_name\": {\"text\": \"national statistical system\", \"confidence\": 0.3105853199958801, \"start\": 55, \"end\": 82, \"clean_text\": \"national statistical system\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 55, \"end\": 82, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 55, \"end\": 82, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 55, \"end\": 82}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 55, \"end\": 82, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.6681492328643799, \"start\": 55, \"end\": 82}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3105853199958801, \"start\": 55, \"end\": 82}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4560018479824066, \"start\": 55, \"end\": 82}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4560018479824066, \"start\": 55, \"end\": 82}}, {\"mention_name\": {\"text\": \"population and business censuses\", \"confidence\": 0.7884879112243652, \"start\": 1574, \"end\": 1606, \"clean_text\": \"population and business censuses\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.37422987818717957, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1574, \"end\": 1606, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1574, \"end\": 1606}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1574, \"end\": 1606, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.5696210265159607, \"start\": 694, \"end\": 726}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7884879112243652, \"start\": 694, \"end\": 726}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5978413820266724, \"start\": 694, \"end\": 726}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5978413820266724, \"start\": 694, \"end\": 726}}, {\"mention_name\": {\"text\": \"household and health surveys\", \"confidence\": 0.6673457622528076, \"start\": 1608, \"end\": 1636, \"clean_text\": \"household and health surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1608, \"end\": 1636, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1608, \"end\": 1636, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1608, \"end\": 1636}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1608, \"end\": 1636, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.5696210265159607, \"start\": 728, \"end\": 756}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6673457622528076, \"start\": 728, \"end\": 756}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5978413820266724, \"start\": 728, \"end\": 756}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5978413820266724, \"start\": 728, \"end\": 756}}, {\"mention_name\": {\"text\": \"Civil Registration and Vital Statistics\", \"confidence\": 0.6031469702720642, \"start\": 1638, \"end\": 1677, \"clean_text\": \"Civil Registration and Vital Statistics\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1638, \"end\": 1677, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1638, \"end\": 1677, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1638, \"end\": 1677}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1638, \"end\": 1677, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.5696210265159607, \"start\": 758, \"end\": 797}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6031469702720642, \"start\": 758, \"end\": 797}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5978413820266724, \"start\": 758, \"end\": 797}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5978413820266724, \"start\": 758, \"end\": 797}}, {\"mention_name\": {\"text\": \"sub-national geospatial data\", \"confidence\": 0.3370954990386963, \"start\": 1690, \"end\": 1718, \"clean_text\": \"sub-national geospatial data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1690, \"end\": 1718, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1690, \"end\": 1718, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1690, \"end\": 1718}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1690, \"end\": 1718, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.5696210265159607, \"start\": 810, \"end\": 838}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3370954990386963, \"start\": 810, \"end\": 838}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5978413820266724, \"start\": 810, \"end\": 838}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5978413820266724, \"start\": 810, \"end\": 838}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"supply and data use constrains. This concerns countries such as Libya, Turkmenistan, and El Salvador.\\n\\nHigh-income countries and countries in East Asia & the Pacific are more likely to be a lake than a swamp. By contrast, upper-middle income countries and countries in Europe and North America are more likely to be a swamp than a lake. Hence, the latter countries may have successfully improved data supply but are lacking in data use. This concerns countries such as the Russian Federation, the UK, France, Germany, and the U.S. The Scandinavian countries are all classified as data lakes, meaning they have high supply and high use of their data.\\n\\n17\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"<!-- Start of picture text --> '<br>_a? Oases: ‘\\\\ Lakes:<br>g low data supply, ' MNG_ high data supply,<br>-“£9\\\" high data researchMDG LBR~_ BRASETH GMBche> *Mw LAO uePL RWA!&cd\\\\OMNRN “ziKWT FINADNKhigh data research<br>g SWZ“GINOOS° tog 8 AL og |e THA op) 1%, NOR<br>bd> NIC~BENT\\\\eQZ “SEN,zwe BGPwigan,ORT GEO SHEOR8 8+ |Soekor SHECITA<br>fo3 won ron noon pNg====SDN HT StEsalMARY AB - “paSC SKHM\\\"“Tune:SSNS“EGY?oe~~ geecoySs CAN -------<br>g Gag—oOD=— MRT NGA COL gp RERs* » MUS» °s* POL MEX<br>iaz | AFGBOIBDI \\\\TLS<CHNNise iigo ipn* 2 BIKSDZAARESALB°GBR® PHL”.<br>8 SOM TOD, Op.JI Lh SLV BOL“ aze\\\\N\\\\ KAZ IND USA<br>3 LBY acoUzB“7 MLI TJK. KGZT—-UKR BLR GTM, MDARUS<br>8'<br>3?©Doscrtel % '' Swamps:<br>2 low data supply, . ' high data supply,<br>& little data research GNB \\\\ little data research<br>3 J<br>025 50 75 100<br>SPI Pillar 4 (avg. 2017-2019)<br>© Highincome © Lowincome © Lower middie income ¢ Notclassified © Upper middie income<br>Red lines indicate median values.<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> = :EY 7, 4 Ss ”><br>;<br>ted ee =F<br>- = VY<br>kes - CAS t<br>Ranh poly x ’ ‘3<br>g@SS weae “aes<br>Classification ly, ¥ 1 :<br>BB vata Desert , ‘<br>| Data Lake ' 2<br>[0| vataData Swamp oasis \\\\Ym,3 Pe 7a<br>Source: SPI<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"_Table 3. Data Deserts, Oases, Lakes, and Swamps by Region and Income Group_\\n\\n||Data<br>Deserts|Data<br>Oases|Data<br>Swamps|Data<br>Lakes|Number of<br>countries|\\n|---|---|---|---|---|---|\\n|Region||||||\\n|East Asia & Pacific|29%|12%|6%|53%|17|\\n|Europe & Central Asia|11%|0%|58%|31%|45|\\n|Latin America & Caribbean|37%|11%|21%|32%|19|\\n|Middle East & North Africa|33%|22%|11%|33%|18|\\n|North America|0%|0%|100%|0%|2|\\n|South Asia|17%|50%|17%|17%|6|\\n|Sub-Saharan Africa|31%|62%|5%|3%|39|\\n|Income group||||||\\n|Low-income|35%|65%|0%|0%|20|\\n|Lower-middle income|31%|49%|11%|9%|45|\\n|Upper-middle income|33%|0%|44%|22%|36|\\n|High-income|7%|0%|38%|56%|45|\\n|Total|25%|24%|26%|25%|146|\\n\\n_Note_ : Share of countries within a region or income group that are classified as data deserts, oases, swamps, or lakes. For example, 29% of the 17 countries in East Asia & Pacific are classified as data deserts.\\n\\nIn Figure A.5 we show the classification when using number of quantitative articles per capita (unadjusted for GDP or number of qualitative articles) as the measure of data use. With this classification, high-income, often OECD, countries are doing well in generating articles using data. Many of the _data deserts_ are in low-income or conflict affected countries. While the per capita measure is useful for an overview of how countries are performing in producing articles using data, it is not always useful as a policy guide given that it is partially driven by GDP per capita and research interests, which are often outside the control of policy makers in the statistical system.\\n\\n# 6 **Robustness Checks**\\n\\nIn this section we offer some robustness checks to our main results. First, we estimate our main regressions using country fixed effects and year dummy variables rather than controlling for qualitative articles. Second, we investigate our results if we exclude medical papers.\\n\\n19\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"# **6.1 Panel Analysis**\\n\\nOur analysis may be subject to some omitted variable bias if countries with high statistical capacity share commonalities unrelated to their number qualitative articles. Below, rather than controlling for the number of qualitative articles, we add country fixed effects and year dummy variables to our OLS regressions, so that only within-country variation over time is used to identify our main relationships. Standard errors are clustered at the country level. Note that this specification may remove relevant variation if some countries have a constant high output of papers using data due to a constant high statistical capacity and vice versa.\\n\\nUsing this specification, the relationship between academic output using data and log GDP per capita is similar as previously estimated, with an elasticity between around 0.4 and 0.6 depending on the specification (Table 4). The relationship with population is not statistically significant under this specification in most cases. The SPI overall scores are not statistically significant either under this specification, but because the SPI overall scores only are available starting in 2016, the time series is relatively short. To overcome this shorter time series, an “extended SPI” index is created that incorporates data from the older Statistical Capacity Indicator (SCI), which the World Bank has produced since around 2004.<sup>4</sup> Using the longer time series in the extended SPI, a 10 unit increase in the extended SPI overall score is associated with an increase in academic output using data of around 0.1%. The estimated coefficient is statistically significant at the 5% level and similar in magnitude to the cross-sectional results that control for the number of qualitative articles. The extended SPI data sources score is also statistically significant at the 5% level.\\n\\n_Table 4. Longitudinal Relationships between Number of Papers Using Data and Statistical Performance_\\n\\n||(1)|(2)|(3)|(4)|(5)|(6)|\\n|---|---|---|---|---|---|---|\\n|Log GDP per capita|0.58**|||0.51|0.58**|0.62**|\\n||(0.20)|||(0.33)|(0.22)|(0.22)|\\n|Log Population||0.18||1.09|0.40|0.38|\\n|||(0.35)||(0.69)|(0.37)|(0.37)|\\n|SPI Overall Score|||0.00|0.00|||\\n||||(0.00)|(0.00)|||\\n|SPI Overall Score<br>(Extended Series)|||||0.01+<br>(0.00)||\\n|SPI Data Sources Score<br>(Extended Series)||||||0.00+|\\n\\n4 The SCI does not contain as many indicators as the SPI (25 indicators versus 51 in the SPI), and not all the indicators are similar. Eleven indicators in the SCI overlap with the SPI: national accounts base year, consumer price index, balance of payments manual in use, government finance accounting, population census, agriculture census, health surveys, household consumption/income survey, CRVS availability, and the IMF SDDS status of the country. Using these indicators, the extended SPI series contains 35 indicators with data back to 2004. These include 5 indicators covering data use, 1 covering data services, 16 covering data products, 8 covering data sources, and 4 covering data infrastructure. The correlation between the extended SPI and the official SPI overall score is around 0.90.\\n\\n20\", \"datasets\": [{\"mention_name\": {\"text\": \"Statistical Capacity Indicator\", \"confidence\": 0.5266559720039368, \"start\": 1317, \"end\": 1347, \"clean_text\": \"Statistical Capacity Indicator\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1317, \"end\": 1347, \"clean_text\": \"\"}, \"producer\": {\"text\": \"World Bank\", \"confidence\": 0.6347318887710571, \"start\": 1365, \"end\": 1375, \"clean_text\": \"World Bank\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1317, \"end\": 1347}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1317, \"end\": 1347, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9865007996559143, \"start\": 458, \"end\": 488}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5266559720039368, \"start\": 458, \"end\": 488}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.6749463081359863, \"start\": 458, \"end\": 488}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6749463081359863, \"start\": 458, \"end\": 488}}, {\"mention_name\": {\"text\": \"SPI Data Sources Score\", \"confidence\": 0.3295683264732361, \"start\": 2300, \"end\": 2322, \"clean_text\": \"SPI Data Sources Score\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2300, \"end\": 2322, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2300, \"end\": 2322, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2300, \"end\": 2322}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2300, \"end\": 2322, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.48686426877975464, \"start\": 204, \"end\": 226}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3295683264732361, \"start\": 204, \"end\": 226}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5806894302368164, \"start\": 204, \"end\": 226}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5806894302368164, \"start\": 204, \"end\": 226}}, {\"mention_name\": {\"text\": \"national accounts base year\", \"confidence\": 0.3434425890445709, \"start\": 2543, \"end\": 2570, \"clean_text\": \"national accounts base year\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2543, \"end\": 2570, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2543, \"end\": 2570, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2543, \"end\": 2570}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2543, \"end\": 2570, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.7248303294181824, \"start\": 184, \"end\": 211}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3434425890445709, \"start\": 184, \"end\": 211}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7418563961982727, \"start\": 184, \"end\": 211}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7418563961982727, \"start\": 184, \"end\": 211}}, {\"mention_name\": {\"text\": \"consumer price index\", \"confidence\": 0.4024520218372345, \"start\": 2572, \"end\": 2592, \"clean_text\": \"consumer price index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2572, \"end\": 2592, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2572, \"end\": 2592, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2572, \"end\": 2592}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2572, \"end\": 2592, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.7248303294181824, \"start\": 213, \"end\": 233}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4024520218372345, \"start\": 213, \"end\": 233}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7418563961982727, \"start\": 213, \"end\": 233}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7418563961982727, \"start\": 213, \"end\": 233}}, {\"mention_name\": {\"text\": \"balance of payments manual in use\", \"confidence\": 0.4064130187034607, \"start\": 2594, \"end\": 2627, \"clean_text\": \"balance of payments manual in use\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2594, \"end\": 2627, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2594, \"end\": 2627, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2594, \"end\": 2627}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2594, \"end\": 2627, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.7248303294181824, \"start\": 235, \"end\": 268}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4064130187034607, \"start\": 235, \"end\": 268}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7418563961982727, \"start\": 235, \"end\": 268}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7418563961982727, \"start\": 235, \"end\": 268}}, {\"mention_name\": {\"text\": \"population census\", \"confidence\": 0.5283282995223999, \"start\": 2660, \"end\": 2677, \"clean_text\": \"population census\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2660, \"end\": 2677, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2660, \"end\": 2677, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2660, \"end\": 2677}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2660, \"end\": 2677, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.7248303294181824, \"start\": 301, \"end\": 318}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5283282995223999, \"start\": 301, \"end\": 318}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7418563961982727, \"start\": 301, \"end\": 318}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7418563961982727, \"start\": 301, \"end\": 318}}, {\"mention_name\": {\"text\": \"agriculture census\", \"confidence\": 0.6513516306877136, \"start\": 2679, \"end\": 2697, \"clean_text\": \"agriculture census\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2679, \"end\": 2697, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2679, \"end\": 2697, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2679, \"end\": 2697}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2679, \"end\": 2697, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.7248303294181824, \"start\": 320, \"end\": 338}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6513516306877136, \"start\": 320, \"end\": 338}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7418563961982727, \"start\": 320, \"end\": 338}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7418563961982727, \"start\": 320, \"end\": 338}}, {\"mention_name\": {\"text\": \"health surveys\", \"confidence\": 0.6322857737541199, \"start\": 2699, \"end\": 2713, \"clean_text\": \"health surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2699, \"end\": 2713, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2699, \"end\": 2713, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2699, \"end\": 2713}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2699, \"end\": 2713, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.7248303294181824, \"start\": 340, \"end\": 354}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6322857737541199, \"start\": 340, \"end\": 354}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7418563961982727, \"start\": 340, \"end\": 354}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7418563961982727, \"start\": 340, \"end\": 354}}, {\"mention_name\": {\"text\": \"household consumption/income survey\", \"confidence\": 0.7857996821403503, \"start\": 2715, \"end\": 2750, \"clean_text\": \"household consumption/income survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2715, \"end\": 2750, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2715, \"end\": 2750, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2715, \"end\": 2750}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2715, \"end\": 2750, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.7248303294181824, \"start\": 356, \"end\": 391}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7857996821403503, \"start\": 356, \"end\": 391}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7418563961982727, \"start\": 356, \"end\": 391}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7418563961982727, \"start\": 356, \"end\": 391}}, {\"mention_name\": {\"text\": \"CRVS availability\", \"confidence\": 0.4394811689853668, \"start\": 2752, \"end\": 2769, \"clean_text\": \"CRVS availability\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2752, \"end\": 2769, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2752, \"end\": 2769, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2752, \"end\": 2769}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2752, \"end\": 2769, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.7248303294181824, \"start\": 393, \"end\": 410}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4394811689853668, \"start\": 393, \"end\": 410}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7418563961982727, \"start\": 393, \"end\": 410}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7418563961982727, \"start\": 393, \"end\": 410}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"<!-- Start of picture text --> Partial regression coefficients conditional on<br>log GDP per capita, log population, and<br>country and year fixed effects.<br>Population Census in past 10 years<br>Agriculture Census in past 10 years<br>Household surveys: 2 or more in 10 years<br>Agriculture Surveys: 2 or more in 10 years<br>Labor Force Surveys: 2 or more in 10 years<br>Health Surveys: 2 or more in 10 years<br>Business/Establishments: 2 or more in 10 years - 7?<br>Complete Civil Registration and Vital Statistics System TT<br>-0.2 0.0 0.2 o4<br>Coefficient estimates and 95% confidence intervals<br>Data from the World Bank's World Development Indicators (WDI) and SPI.<br>Papers include all papers using data years 2000-2020.<br>Confidence Intervals at 95% significance level<br><!-- End of picture text -->\", \"datasets\": [{\"mention_name\": {\"text\": \"Population Census\", \"confidence\": 0.598695695400238, \"start\": 159, \"end\": 176, \"clean_text\": \"Population Census\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 159, \"end\": 176, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 159, \"end\": 176, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 159, \"end\": 176}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 159, \"end\": 176, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.9078171253204346, \"start\": 159, \"end\": 176}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.598695695400238, \"start\": 159, \"end\": 176}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5290254354476929, \"start\": 159, \"end\": 176}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5290254354476929, \"start\": 159, \"end\": 176}}, {\"mention_name\": {\"text\": \"Agriculture Census\", \"confidence\": 0.6211783289909363, \"start\": 197, \"end\": 215, \"clean_text\": \"Agriculture Census\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 197, \"end\": 215, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 197, \"end\": 215, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 197, \"end\": 215}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 197, \"end\": 215, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.9078171253204346, \"start\": 197, \"end\": 215}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6211783289909363, \"start\": 197, \"end\": 215}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5290254354476929, \"start\": 197, \"end\": 215}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5290254354476929, \"start\": 197, \"end\": 215}}, {\"mention_name\": {\"text\": \"Household surveys\", \"confidence\": 0.5171104669570923, \"start\": 236, \"end\": 253, \"clean_text\": \"Household surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 236, \"end\": 253, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 236, \"end\": 253, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 236, \"end\": 253}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 236, \"end\": 253, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.9078171253204346, \"start\": 236, \"end\": 253}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5171104669570923, \"start\": 236, \"end\": 253}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5290254354476929, \"start\": 236, \"end\": 253}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5290254354476929, \"start\": 236, \"end\": 253}}, {\"mention_name\": {\"text\": \"Agriculture Surveys\", \"confidence\": 0.5518834590911865, \"start\": 280, \"end\": 299, \"clean_text\": \"Agriculture Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 280, \"end\": 299, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 280, \"end\": 299, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 280, \"end\": 299}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 280, \"end\": 299, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.9078171253204346, \"start\": 280, \"end\": 299}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5518834590911865, \"start\": 280, \"end\": 299}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5290254354476929, \"start\": 280, \"end\": 299}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5290254354476929, \"start\": 280, \"end\": 299}}, {\"mention_name\": {\"text\": \"Labor Force Surveys\", \"confidence\": 0.5594791173934937, \"start\": 326, \"end\": 345, \"clean_text\": \"Labor Force Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 326, \"end\": 345, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 326, \"end\": 345, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 326, \"end\": 345}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 326, \"end\": 345, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.9078171253204346, \"start\": 326, \"end\": 345}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5594791173934937, \"start\": 326, \"end\": 345}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5290254354476929, \"start\": 326, \"end\": 345}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5290254354476929, \"start\": 326, \"end\": 345}}, {\"mention_name\": {\"text\": \"Health Surveys\", \"confidence\": 0.5727434754371643, \"start\": 372, \"end\": 386, \"clean_text\": \"Health Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 372, \"end\": 386, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 372, \"end\": 386, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 372, \"end\": 386}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 372, \"end\": 386, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.9078171253204346, \"start\": 372, \"end\": 386}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5727434754371643, \"start\": 372, \"end\": 386}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5290254354476929, \"start\": 372, \"end\": 386}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5290254354476929, \"start\": 372, \"end\": 386}}, {\"mention_name\": {\"text\": \"Complete Civil Registration and Vital Statistics System TT\", \"confidence\": 0.4859752953052521, \"start\": 468, \"end\": 526, \"clean_text\": \"Complete Civil Registration and Vital Statistics System TT\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 468, \"end\": 526, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 468, \"end\": 526, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 468, \"end\": 526}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 468, \"end\": 526, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.3362158238887787, \"start\": 468, \"end\": 526}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4859752953052521, \"start\": 468, \"end\": 526}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4360317885875702, \"start\": 468, \"end\": 526}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4360317885875702, \"start\": 468, \"end\": 526}}, {\"mention_name\": {\"text\": \"World Development Indicators\", \"confidence\": 0.48890337347984314, \"start\": 630, \"end\": 658, \"clean_text\": \"World Development Indicators\"}, \"acronym\": {\"text\": \"WDI\", \"confidence\": 0.3784240782260895, \"start\": 660, \"end\": 663, \"clean_text\": \"WDI\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 630, \"end\": 658, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 630, \"end\": 658}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 630, \"end\": 658, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8214138150215149, \"start\": 630, \"end\": 658}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.48890337347984314, \"start\": 630, \"end\": 658}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6019178032875061, \"start\": 630, \"end\": 658}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6019178032875061, \"start\": 630, \"end\": 658}}, {\"mention_name\": {\"text\": \"SPI\", \"confidence\": 0.30787399411201477, \"start\": 669, \"end\": 672, \"clean_text\": \"SPI\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 669, \"end\": 672, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 669, \"end\": 672, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 669, \"end\": 672}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 669, \"end\": 672, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8214133381843567, \"start\": 669, \"end\": 672}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.30787399411201477, \"start\": 669, \"end\": 672}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6019169092178345, \"start\": 669, \"end\": 672}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6019169092178345, \"start\": 669, \"end\": 672}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"<!-- Start of picture text --> CH<br>°<br>MYS USA<br>1,000 IDN” oe AUS<br>2 286 UPN<br>g ROUz° one*. 29,\\\" BRA<br>a eee! ‘kor<br>D °3° ee<br>i= oe e<br>@ o ° e<br>> 100 &° °<br>2 ed a\\\"<br><g °\\\"35 4 e<br>3y = 18.34+ 0.29 x, R2=0.93<br>s tem (468) (0.01) G<br>=oD 10 ° ~~ ° ~~,CIV *,GIN,<br>5 DMA YCT v's .. CAP GAB TGO<br>2 DJl-*'* yop eve<br>TUV_STP.A 5 KIR'SSBMU<br>AND yw KNA FSMSGNQ<br>LiE*CywTM ASM “Nop<br>4 MNFATG: ME, MOQ.<br>N IMN 6NB\\\\. GIB PYF<br>TCAZSMR COM PLW<br>M cpv cuw \\\\MAE-FRO<br>10 100 1,000 10,000<br>All Papers Using Data<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"<!-- Start of picture text --> Logged number of Articles<br>]<br>1.00 All Medicine Economics Psychology Sociology Poli Sci<br>0.30 Corr Corr Corr Corr Corr ><br>0.10 0.994*** 0.960*** 0.816*** 0.898*** 0.919*** i<br>1000 =<br>40 Corr Corr Corr Corr<br>0.937 0.776 0.880 0.896 ¥<br>10 ®<br>1000 3 3<br>fe g<br>10 e 0.735 0.819 0.862 2<br>°<br>° » wR * .<br>11° .ee, 2 eee<br>400 °. .. tone~~ . °.e e . 5<br>* Corr. Corr: ri<br>10 ¥ 0.844 0.838 &<br>é . a<br>4 2? See S<br>100 °.e °e° °ee . ~~°e ] g<br>ome ° e % @-e. Corr 8<br>e ° ° 3 98s e . ).973°** &fs]<br>z ls 4 2<br>4 we ee oes oe Lseeee°See be {__']<br>100 * .. fea. ° .el aera ° ~<br>° 2, e % Q<br>101 * le bd °° - ie ° *oe ° rybte 3 ° 3 ; e 2on<br>10 100 10001000011 bp aa10 100 1000100001a10 100 1000r 1 ons,10 100r tg1 288. ©10 1001 141 10r 1001<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"accessible to researchers, for example through data sharing agreements or by boosting data literacy.\\n\\nOur results can help countries better understand if their statistical system is supporting the needs of researchers. Given that research is at the heart of evidence-based policy making, a statistical system oriented around making data available and easily accessible for these purposes is important. Countries could use the results to facilitate an understanding of the maturity of their statistical systems in supporting researchers, point to other countries that a particular country could learn from as it seeks to improve, and track progress or assess return on investment for funding related to statistical capacity building.\\n\\nOur results face some limitations. First, the s2orc database does not currently contain non-English language articles. While the dominance of the English language in academic publications has been well-documented (Altbach 2007), the database omits articles published in other languages. Second, because of computational limitations, only 1 million articles were analyzed out of a sampling frame of around 10 million articles. Future work could bring in these additional articles to improve precision. Third, while the natural language processing model was able to classify whether or not the articles made use of data with 87% accuracy, the model could be extended in ways to extract more information from the articles, including more details about the subject of study and which particular data sources may have been used (surveys, censuses, administrative data, geospatial information etc.).\\n\\n# **References**\\n\\nAltbach, Philip G. “The imperial tongue: English as the dominating academic language.” _Economic and political Weekly_ (2007): 3608-3611.\\n\\nCameron, Drew B, Anjini Mishra, and Annette N Brown. 2016. “The Growth of Impact Evaluation for International Development: How Much Have We Learned?” _Journal of Development Effectiveness_ 8 (1): 1–21.\\n\\nCameron, Grant J., Hai-Anh H. Dang, Mustafa Dinc, James Foster, and Michael M. Lokshin. 2021. “Measuring the Statistical Capacity of Nations.” _Oxford Bulletin of Economics and Statistics_ 83(4): 870-896.\\n\\nCourtioux, Pierre, François Métivier, and Antoine Rebérioux. 2022. “Nations Ranking in Scientific Competition: Countries Get What They Paid For.” _Economic Modelling_ 116: 105976.\\n\\nDang, Hai-Anh H, John Pullinger, Umar Serajuddin, and Brian Stacy. 2023. “Statistical Performance Indicators and Index: A New Tool to Measure Country Statistical Capacity.” _Scientific Data_ 10(1): 146.\\n\\nDas, Jishnu, Quy-Toan Do, Karen Shaines, and Sowmya Srikant. 2013. “US and Them: The Geography of Academic Research.” _Journal of Development Economics_ 105: 112–30.\\n\\n24\", \"datasets\": [{\"mention_name\": {\"text\": \"s2orc database\", \"confidence\": 0.7714194655418396, \"start\": 780, \"end\": 794, \"clean_text\": \"s2orc database\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 780, \"end\": 794, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 780, \"end\": 794, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 780, \"end\": 794}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 780, \"end\": 794, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9999040365219116, \"start\": 780, \"end\": 794}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7714194655418396, \"start\": 780, \"end\": 794}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.43992024660110474, \"start\": 780, \"end\": 794}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.43992024660110474, \"start\": 780, \"end\": 794}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"Devlin, Jacob, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2018. “BERT: Pre-Training of Deep Bidirectional Transformers for Language Understanding.” arXiv. . https://doi.org/10.48550/ARXIV.1810.04805 Hansen, Stephen, Michael McMahon, and Andrea Prat. 2018. “Transparency and Deliberation Within the FOMC: A Computational Linguistics Approach.” _The Quarterly Journal of Economics_ 133 (2): 801–70.\\n\\nHjort, Jonas, Diana Moreira, Gautam Rao, and Juan Francisco Santini. 2021. “How Research Affects Policy: Experimental Evidence from 2,150 Brazilian Municipalities.” _American Economic Review_ 111 (5): 1442-80.\\n\\n<mark>Jolliffe, Dean, Daniel Gerszon Mahler, Malarvizhi Veerappan, Talip Kilic, Philip Wollburg. 2023. “What Makes Public Sector Data Relevant for Development?”</mark> _<mark>World Bank Research Observer</mark>_ <mark>38 (2): 325-346.</mark> Kleinberg, Bennett, Maximilian Mozes, Arnoud Arntz, and Bruno Verschuere. 2018. “Using Named Entities for Computer-Automated Verbal Deception Detection.” _Journal of Forensic Sciences_ 63 (3): 714–23.\\n\\nLo, Kyle, Lucy Lu Wang, Mark Neumann, Rodney Kinney, and Daniel Weld. 2020. “S2ORC: The Semantic Scholar Open Research Corpus.” In _Proceedings of the 58_<sup>_th_</sup> _Annual Meeting of the Association for Computational Linguistics_ , 4969–83. Online: Association for Computational Linguistics. https://doi.org/10.18653/v1/2020.acl-main.447.\\n\\nNational Science Board, National Science Foundation. 2019. “Publications Output: US Trends and International Comparisons. Science & Engineering Indicators 2018.” _National Science Foundation_ .\\n\\nPaszke, Adam, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, et al. 2019. “PyTorch: An Imperative Style, High-Performance Deep Learning . Library.” arXiv. https://doi.org/10.48550/ARXIV.1912.01703 Phillips, Brian J., and Kevin T. Greene. 2022. “Where is Conflict Research? Western Bias in the Literature on Armed Violence.” _International Studies Review_ 24 (3): viac038.\\n\\nPorteous, Obie. 2020. “Research Deserts and Oases: Evidence from 27 Thousand Economics Journal Articles on Africa.” _Oxford Bulletin of Economics and Statistics_ .\\n\\nRobinson, Michael D, James E Hartley, and Patricia Higino Schneider. 2006. “Which Countries Are Studied Most by Economists? An Examination of the Regional Distribution of Economic Research.” _Kyklos_ 59 (4): 611–26.\\n\\nSabet, Shayda Mae, and Annette N Brown. 2018. “Is Impact Evaluation Still on the Rise? The New Trends in 2010–2015.” _Journal of Development Effectiveness_ 10 (3): 291–304.\\n\\nSanh, Victor, Lysandre Debut, Julien Chaumond, and Thomas Wolf. “DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter.” _arXiv preprint arXiv:1910.01108_ (2019).\\n\\n25\", \"datasets\": [{\"mention_name\": {\"text\": \"Public Sector Data\", \"confidence\": 0.4164694547653198, \"start\": 733, \"end\": 751, \"clean_text\": \"Public Sector Data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 733, \"end\": 751, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 733, \"end\": 751, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 733, \"end\": 751}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 733, \"end\": 751, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.39001938700675964, \"start\": 733, \"end\": 751}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4164694547653198, \"start\": 733, \"end\": 751}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6770160794258118, \"start\": 733, \"end\": 751}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6770160794258118, \"start\": 733, \"end\": 751}}, {\"mention_name\": {\"text\": \"Science & Engineering Indicators 2018\", \"confidence\": 0.39797312021255493, \"start\": 1530, \"end\": 1567, \"clean_text\": \"Science & Engineering Indicators 2018\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1530, \"end\": 1567, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1530, \"end\": 1567, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1530, \"end\": 1567}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1530, \"end\": 1567, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9994472861289978, \"start\": 41, \"end\": 78}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.39797312021255493, \"start\": 41, \"end\": 78}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8186172842979431, \"start\": 41, \"end\": 78}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8186172842979431, \"start\": 41, \"end\": 78}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"Shelar, Hemlata, Gagandeep Kaur, Neha Heda, and Poorva Agrawal. 2020. “Named Entity Recognition Approaches and Their Comparison for Custom Ner Model.” _Science & Technology Libraries_ 39 (3): 324–37.\\n\\nSmirnov, Anatoly A., and Irina V. Stukova. “Determinants of integration approach in the agrarian sphere development in contexts of transformation.” _Review of European studies_ 7, no. 8 (2015): 8.\\n\\nWorld Bank. 2021. “World Development Report 2021: Data for Better Lives.” Washington DC: World Bank. https://wdr2021.worldbank.org/ Yu Tian. 2020. “THE PARTNER REPORT ON SUPPORT TO STATISTICS PRESS 2020.” PARIS21. https://paris21.org/sites/default/files/inline-files/PRESS2020%20Final.pdf.\\n\\n26\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099503001172493286/pdf/IDU1ad0286b61aec014d8c19d3e1e651d3fe6286.pdf\", \"pages\": [27]}}]}"}}},{"rowIdx":80,"cells":{"doc_id":{"kind":"string","value":"34241066"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34241066\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"authors\": {\"0\": {\"author\": \"Najam,Rafiuddin\"}}, \"count\": \"Afghanistan\", \"count_key\": \"82662\", \"docna\": {\"0\": {\"docna\": \"Closing the gap : Effect of a gender quota\\n on women's access to education in Afghanistan\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"DIME Gender, Eco Opp & Fragility (DIME2)\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34241066\"}, \"repnb\": \"WPS10674\", \"docdt\": \"2024-01-18T00:00:00Z\", \"datestored\": \"2024-01-18T01:00:15Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"Affirmative action is a promising\\n solution to the crucial challenge of bridging the gap in\\n women's access to higher education in low- and\\n middle-income countries (LMICs). This paper uses public\\n universities' matriculation data from\\n 2013–2018 and difference-in-differences\\n estimators to examine the causal impact of a gender quota on\\n women's educational opportunities in Afghanistan. The\\n quota increased the proportion of women in the treated\\n concentration group by nine percentage points and the share\\n of women from low socio-economic status by three percentage\\n points. The expansion was associated with a 0.04-unit\\n decline in the average score ratio of female-to-male\\n applicants, driven by a reduction in the score threshold\\n needed for women's admission. The effects were\\n condensed in competitive concentrations, where the overall\\n share of women and women with low SES increased by 17 and\\n four percentage points, respectively. The findings suggest\\n that affirmative action is a viable option for addressing\\n the gender gap in fragile settings.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10674; RRR; Impact Evaluation series; PEOPLE\", \"display_title\": \"Closing the gap : Effect of a gender\\n quota on women's access to education in Afghanistan\", \"disclosure_date\": \"2024-01-18T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099300101182423297/text/IDU1617049971af801450e18c91117260363b929.txt\", \"listing_relative_url\": \"/research/2024/01/34241066/closing-gap-effect-gender-quota-womens-access-education-afghanistan\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099300101182423297/IDU1617049971af801450e18c91117260363b929\", \"new_url\": \"2024/01/34241066/IDU1617049971af801450e18c91117260363b929\", \"ext_pub_date\": \"2024-01-18T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Education and Gender m1326343 1545,Tertiary Education m1326349 482,Education Access and Equity m1327050 2127,Education in Post Conflict Situations\\n m1327743 1548\", \"docm_id\": \"IDU-61704997-af80-450e-8c91-17260363b929\", \"chronical_docm_id\": \"IDU-61704997-af80-450e-8c91-17260363b929\", \"publishtoextweb_dt\": \"2024-01-18T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10674\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"guid\": \"099300101182423297\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"57\", \"author_keys\": {\"0\": {\"author_key\": \"2342299\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099300101182423297/IDU1617049971af801450e18c91117260363b929\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10674\\n\\n# Closing the Gap\\n\\nEffect of a Gender Quota on Women’s Access to Education in Afghanistan _Rafiuddin Najam_ Development Economics Development Impact Group January 2024 A verified reproducibility package for this paper is available at http://reproducibility.worldbank.org, click **here** for direct access.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10674\\n\\n# **Abstract**\\n\\nAffirmative action is a promising solution to the crucial challenge of bridging the gap in women’s access to higher education in low- and middle-income countries (LMICs). This paper uses public universities’ matriculation data from 2013–2018 and difference-in-differences estimators to examine the causal impact of a gender quota on women’s educational opportunities in Afghanistan. The quota increased the proportion of women in the treated concentration group by nine percentage points and the share of women from low socio-economic status by three percentage points. The expansion was associated with a 0.04-unit decline in the average score ratio of female-tomale applicants, driven by a reduction in the score threshold needed for women’s admission. The effects were condensed in competitive concentrations, where the overall share of women and women with low SES increased by 17 and four percentage points, respectively. The findings suggest that affirmative action is a viable option for addressing the gender gap in fragile settings..\\n\\nThis paper is a product of the Development Impact Group, Development Economics. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The author may be contacted at rnajam@worldbank.org. A verified reproducibility package for this paper is available at http://reproducibility. worldbank.org, click **here** for direct access.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [{\"mention_name\": {\"text\": \"matriculation data\", \"confidence\": 0.5739385485649109, \"start\": 261, \"end\": 279, \"clean_text\": \"matriculation data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 261, \"end\": 279, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 261, \"end\": 279, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 261, \"end\": 279}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 261, \"end\": 279, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8796467781066895, \"start\": 261, \"end\": 279}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5739385485649109, \"start\": 261, \"end\": 279}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8341440558433533, \"start\": 261, \"end\": 279}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8341440558433533, \"start\": 261, \"end\": 279}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# Closing the gap: Effect of a gender quota on women’s access to education in Afghanistan\\n\\n## Rafiuddin Najam<sup>a∗</sup>\\n\\n> aPhD student in Public Policy, Oregon State University, Corvallis, USA; RSMFP Fellow Development Impact, World Bank, Washington DC, USA\\n\\n**Keywords:** Affirmative action, gender, women’s education, access, Afghanistan _JEL Classification codes: I24, I28, I23, O15_\\n\\n> ∗E-mail: najamr@oregonstate.edu\\n\\n> Disclaimer: The findings, interpretations, and conclusions expressed in this work are entirely those of the author and should not be attributed in any manner to the World Bank, its Board of Executive Directors, or the governments they represent.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **Abbreviations**\\n\\nGPI: gender parity index IHEs: institutions of higher education LMICs: low- and middle-income countries MoE: Ministry of Education MoHE: Ministry of Higher Education NEXA: National Examination Authority of Afghanistan PMB: private marginal benefits SDGs: sustainable development goals SES: socioeconomic status VIIRS: visible infrared imaging radiometer suite i\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"# **1 Introduction**\\n\\nIn low- and middle-income countries (LMICs), primarily in fragile and conflict-affected settings, achieving gender-equal access to institutions of higher education (IHEs) remains a priority. Such access is pivotal for meeting the United Nations Sustainable Development Goals (SDGs)—particularly SDG4, focusing on inclusive, quality education, and SDG5, advocating for gender equality—and amplifying socioeconomic benefits (Chaaban & Cunningham, 2011; Hill & King, 1995; Psacharopoulos & Patrinos, 2018). While the significance of post-secondary education is growing globally, LMICs struggle with lower enrollment level and a higher gender gap. Despite the vast expansion of IHEs in these regions, ensuring women’s access remains a substantial barrier. This obstacle impedes the reduction of gender inequality in IHEs (Ilie & Rose, 2016) and hampers women’s long-term labor market participation. In Afghanistan, fragility and cultural barriers have limited women’s access to post-secondary education (Baiza, 2013; Darwish & Wotipka, 2022; Hayward & Babury, 2015; Hayward & Karim, 2019; Shayan, 2015). Since the early 2000s, educational opportunities have expanded significantly, notably for girls in Afghan public institutions (Najam & Johnston, 2023). However, the gender parity index (GPI) of public post-secondary education remains one of the lowest globally. Following the Taliban’s ban on women’s education (since 2021), the GPI is expected to decline further.\\n\\nTo combat educational inequalities, policymakers and educational institutions have adopted interventions such as affirmative action to enhance the presence of underrepresented groups in IHEs. These policies are typically target-based, offering preferential treatment to specific applicant subgroups who face structural barriers to accessing education. Several types of affirmation action have been implemented: gender quotas (Bagde et al., 2016); racial-quotas, predominantly in the United States (Arcidiacono et al., 2015; Card & Krueger, 2005; Hinrichs, 2014; Long, 2004; Rose, 2005) and Brazil (Estevan et al., 2019; Francis-Tan & Tannuri-Pianto, 2015; Mello, 2022; Vieira & Arends-Kuenning, 2019); caste- and/or class1\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"based quotas, primarily in India (Basant & Sen, 2020; Bertrand et al., 2010; Cassan, 2019) and Israel (Alon & Malamud, 2014); and geographic quotas, in Sri Lanka (de Silva et al., 2021). Most of the literature indicates that affirmative action is instrumental in increasing access to education in historically underrepresented populations.\\n\\nPublic universities in Afghanistan operate under a stringent centralized system. Admission to a public university is determined solely by the applicants’ scores on the annual nationwide entrance exam for public universities, known as the _Kankor_ examination. Female applicants tended to score lower than their male counterparts in the Kankor. This disparity in performance is attributable to the systematic underinvestment and cultural barriers Afghan women face in accessing post-secondary education (Bamik, 2018; Hayward & Babury, 2015; Najam & Johnston, 2023), which is exacerbated by conflict (Darwish & Wotipka, 2022). In 2016, the Afghan government introduced a gender quota for admissions to address gender disparities and promote women’s long-term labor participation. The policy expanded until the Taliban’s return. This policy allocated seats at the enrollment unit level ( _concentrations_ ) by gender in particular universities and courses of study. These concentrations served as the _treated units_ . Consequently, the admissions competition was gender-specific—female applicants competed with other females, and male applicants competed with other males—for gender-designated seats in treated concentrations. Conversely, the other concentrations maintained open competition<sup>1</sup> (hereafter, _untreated units_ ).\\n\\nIn this paper, I examine the short-term efficacy of gender quotas on women’s matriculation<sup>2</sup> , especially those from low socioeconomic status—proxied by districts’ average nightlight data—in Afghan public universities. Specifically, I explore variations across study units and over time to assess the causal impact of gender quotas on the share of matriculated female students and score-related outcomes using the difference-in-differences (DD) methods. The causality of the DD estimates relies on the central parallel\\n\\n> 1Admission is solely based on applicants’ Kankor score.\\n\\n> 2Kankor participants receive a single admission offer from the administering authority. The terms ‘matriculation,’ ‘admission,’ and ‘enrollment’ are used interchangeably.\\n\\n2\", \"datasets\": [{\"mention_name\": {\"text\": \"districts’ average nightlight data\", \"confidence\": 0.5847768783569336, \"start\": 1840, \"end\": 1874, \"clean_text\": \"districts’ average nightlight data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1840, \"end\": 1874, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1840, \"end\": 1874, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1840, \"end\": 1874}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1840, \"end\": 1874, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.30026477575302124, \"start\": 983, \"end\": 1017}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.5847768783569336, \"start\": 983, \"end\": 1017}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.48808571696281433, \"start\": 983, \"end\": 1017}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.48808571696281433, \"start\": 983, \"end\": 1017}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"trend assumption. This implies that in the absence of the quota intervention when accounting for the fixed effects, the outcome trends for the treated and untreated groups would follow a similar pattern as the trends between the two groups in the pre-treatment period. I found evidence from an event study analysis that the parallel trends assumption holds.\\n\\nI use the universe of matriculation results of Kankor test-takers from the academic years 2013–2018. This dataset contains detailed applicant characteristics for 1.51 million participants, such as gender, test score, high school name, and the unit of study to which they were admitted. For this study, I focus _exclusively_ on a subset of 336,221 applicants admitted to public universities over six-years. I employ district-level monthly average luminosity data from 2014–2015 to indicate economic development<sup>3</sup> and identify applicants from areas of low socioeconomic status (SES). This categorization enables me to assess the impact of the quota intervention on the participation and score-related outcomes of female applicants with high and low<sup>4</sup> statuses.\\n\\nThe quota led to a nine-percentage point increase in the matriculation rate of female applicants in the treated concentrations, a 32% increase over the control group mean. Specifically, the quota raised the matriculation rate of female applicants from low-SES districts by three percentage points, a 30% increase compared to the control group mean. This expansion in women’s admissions to public IHEs has lessened the gender gap. However, it has also resulted in a reduced score threshold for female students’ admissions to treated units. This is revealed by a significant decrease of 0.04 units in the average score ratio of female-to-male applicants in the treated concentrations, a 4% reduction compared to the control group mean. This indicates that the preferential treatment enrolled female students with lower Kankor scores. This favored treatment might raise concerns about applicants’ academic performance post-admission,\\n\\n> 3I acknowledge the coarse nature of this classification. However, given the lack of granular metrics at the applicant and/or district-level in Afghanistan, the use of nightlight data provides a plausible indicator of economic development at the district-level. I am unable to differentiate between the socioeconomic statuses of individual applicants level and instead rely on district classifications.\\n\\n> 4I use the average of luminosity across district as the source of operationalization. Low-socioeconomic status indicates districts with an average luminosity value less than the mean.\\n\\n3\", \"datasets\": [{\"mention_name\": {\"text\": \"matriculation results of Kankor test-takers\", \"confidence\": 0.35716405510902405, \"start\": 381, \"end\": 424, \"clean_text\": \"matriculation results of Kankor test-takers\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 381, \"end\": 424, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 381, \"end\": 424, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 381, \"end\": 424}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 381, \"end\": 424, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.9231358170509338, \"start\": 381, \"end\": 424}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.35716405510902405, \"start\": 381, \"end\": 424}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9384443759918213, \"start\": 381, \"end\": 424}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9384443759918213, \"start\": 381, \"end\": 424}}, {\"mention_name\": {\"text\": \"district-level monthly average luminosity data\", \"confidence\": 0.82193922996521, \"start\": 774, \"end\": 820, \"clean_text\": \"district-level monthly average luminosity data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 774, \"end\": 820, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 774, \"end\": 820, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 774, \"end\": 820}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 774, \"end\": 820, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.6368247866630554, \"start\": 774, \"end\": 820}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.82193922996521, \"start\": 774, \"end\": 820}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8631160259246826, \"start\": 774, \"end\": 820}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8631160259246826, \"start\": 774, \"end\": 820}}, {\"mention_name\": {\"text\": \"nightlight data\", \"confidence\": 0.8563909530639648, \"start\": 2240, \"end\": 2255, \"clean_text\": \"nightlight data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2240, \"end\": 2255, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2240, \"end\": 2255, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2240, \"end\": 2255}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2240, \"end\": 2255, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9016677141189575, \"start\": 577, \"end\": 592}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8563909530639648, \"start\": 577, \"end\": 592}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.8465720415115356, \"start\": 577, \"end\": 592}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8465720415115356, \"start\": 577, \"end\": 592}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"which is commonly referred to as a mismatch between applicants’ preparation and the fields of study (Frisancho & Krishna, 2016; Rose, 2005).\\n\\nAdditionally, I assess the effects of the gender quota on women’s access and score-related outcomes based on the competitiveness of academic programs. Highly competitive academic programs encompass concentrations within computer science, economics, engineering, law and political science, and medical studies, collectively referred to as CEELM programs<sup>5</sup> , across public universities. Before 2016, female students constituted an average of 17% in CEELM programs, marking an 11% deficit compared with their representation in non-CEELM programs. I found that the quota led to a 17-percentage point rise in the share of female applicants and a four-percentage point increase in the share of low-SES female applicants in highly competitive concentrations. This nearly doubled the share of female students compared to the control group mean, suggesting a significant reduction in gender disparities in the selected concentrations. This is a crucial step towards enhancing women’s participation in the workforce in much-needed areas.\\n\\nThis study offers a pivotal contribution to empirical research assessing the effects of affirmative action on marginalized groups’ access to educational opportunities and the associated implications of qualityaccess trade-offs. First, to my knowledge, this is one of the first studies to examine the impact of affirmative action on women’s access to higher education in fragile and conflict-affected settings. Existing studies in this domain have primarily focused on more stable environments, such as the US, India, and Brazil. Second, this study contributes to the literature that investigates the causal effects of affirmative action (Basant & Sen, 2020; Bertrand et al., 2010; de Silva et al., 2021; Frisancho & Krishna, 2016; Mello, 2022), and/or a ban of affirmative action (Antonovics & Backes, 2014; Arcidiacono et al., 2015; Card & Krueger, 2005; Epple et al., 2008; Hinrichs, 2014; Long, 2004). The focus has been on marginalized groups’ access to educational opportunities, emphasizing varying impacts across different socioeconomic strata (Backes, 2012; Estevan et al., 2019; Francis & Tannuri-Pianto, 2012; Francis-Tan & Tannuri-Pianto,\\n\\n> 5Among the CEELM programs, no concentrations in economics and the law and political science are treated.\\n\\n4\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"2015), and the selective field of studies/colleges (Alon & Malamud, 2014; Bagde et al., 2016; Hinrichs, 2012; Howell, 2010; Vieira & Arends-Kuenning, 2019). Third, I assess the efficacy of an intervention targeting an increase in the potential supply of qualified women for the labor force. The lack of qualified female applicants—nationally and within refined sectors—presents a significant challenge to achieving employment quotas (Howard & Prakash, 2012; Myers, 2007; Peck, 2017; Prakash, 2020). Finally, this study examines a gender quota within Afghanistan’s unique centralized admission system, where applicants select their study fields without knowing their Kankor exam scores, highlighting the significant role of affirmative action in expanding the availability of educational opportunities at the national level for female applicants.\\n\\n# **2 Literature review**\\n\\nAffirmative action comprises target-based policies and practices. The primary objective is to enhance opportunities for specific applicant sub-groups facing systemic barriers to education and employment. This includes preferential consideration of historically underrepresented sub- groups based on factors such as gender, caste, class, region, and race, among others. An extensive body of literature—predominantly from India, the US, and Brazil—examines the direct and indirect impacts of adopting and eliminating affirmative action on (post-)matriculation outcomes.\\n\\nRegarding the efficacy of affirmative action in college admissions, the findings are consistent across various contexts. In India, the caste-based system creates social stratification, leading to structural inequalities in the education and employment of minorities (Bertrand et al., 2010; Howard & Prakash, 2012). Nevertheless, affirmative action mitigates the widening inequality. Bagde et al. (2016), Basant & Sen (2020),Bertrand et al. (2010), Lee (2021) and Cassan (2019) find that caste-based quota policies increase the representation of disadvantaged castes in post-secondary education. In Brazil, Estevan et al.\\n\\n5\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"(2019), Francis & Tannuri-Pianto (2012), Francis-Tan & Tannuri-Pianto (2015), Mello (2022), and Vieira & Arends-Kuenning (2019) indicate that racial and group-specific policies boost the enrollment of the targeted population. In Israel, Alon & Malamud (2014) reveals that affirmative action policies enhance the probability of admission and enrollment, particularly in selective majors. Similarly, in Sri Lanka, de Silva et al. (2021) find that geographically-targeted policies improve the representation of students from disadvantaged regions.\\n\\nIn the US, affirmative action in higher education involves race-based admission selections—a divisive subject (Arcidiacono et al., 2015). Several universities and states have chosen to ban affirmative action due to controversies<sup>6</sup> and rulings from the US Supreme Court regarding race-based considerations in college admissions (Baker, 2019). Card & Krueger (2005) indicate that the ban on affirmative action reduced the overall admission rate for African-American and Hispanic students. However, highly qualified minority applicants continue to send their test scores to selective state institutions. Hinrichs (2012) shows that an affirmative action ban results in decreased enrollment of underrepresented minorities at selective colleges, although there is no effect at typical colleges. Similarly, Backes (2012) reports a decline in minority student enrollment. Hinrichs (2014) asserts that banning affirmative action decreases the number of minority graduates. Nevertheless, the graduation rate for minorities has increased because of the shifting composition of minority students in selective colleges. Furthermore, Epple et al. (2008) and Howell (2010) emphasize that an affirmative action ban can significantly reduce minority representation in toptier institutions. Klasik & Cortes (2022), Long (2004; 2015), and Antonovics & Backes (2014) reveal that alternative policies fall short of achieving minority representation levels comparable to those achieved under affirmative action.\\n\\nOne subset of the literature assesses the effects of affirmative action’s policies on human capital\\n\\n> 6For a detailed discussion about the trade-offs imposed by affirmative action and an extensive review of the literature, see (Arcidiacono & Lovenheim, 2016), (Fryer & Loury, 2005), and (Holzer & Neumark, 2000).\\n\\n6\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"development. In India, Khanna (2020) and Cassan (2019) demonstrate that affirmative action improves years of schooling and educational attainment, respectively. In the US, Akhtari et al. (ming) find that affirmative action reduces the racial gap and encourages pre-college human capital investment.\\n\\nMixed evidence exists on the impacts of affirmative action on minorities’ academic performance. Rose (2005) and Frisancho & Krishna (2016) find that students benefiting from affirmative action lag behind their peers academically, particularly in selective majors. Conversely, Alon & Malamud (2014) suggest that students admitted based on preferential treatment do not fall behind academically, regardless of their selection of majors.\\n\\nThe primary focus of affirmative action policies in education is to address educational disparities. However, they also play a pivotal role in shaping the employment outcomes of minorities. Howard & Prakash (2012) posits that the impact of the employment quota is linked to the schooling years of minorities, highlighting the interlinkages between educational progression and economic mobility. Myers (2007) finds that abolishing employment-based affirmative action is correlated with a decline in minority labor force participation, further illustrating the broader ramification of these policies. Thus, policies to rectify educational disparities could also serve as a mechanism to alleviate employment inequalities. This is exemplified by the gender quota in public IHEs introduced by the Afghan government, which sought to expand the supply of qualified women in the labor force.\\n\\n7\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"# **3 Background**\\n\\n## **3.1 Admission to Afghan public universities**\\n\\nThe admission to public IHEs is strictly based on applicants’ Kankor scores. Kankor is a nationwide entrance exam<sup>7</sup> , held annually countrywide for applicants, predominantly new high school graduates interested in pursuing undergraduate studies at public universities. This is the only mechanism through which applicants can attain admission into public universities’ daytime undergraduate programs, as shown in Figure A1.1. The examination has been administered by the National Examination Authority of Afghanistan (NEXA) since 2018 and was previously overseen by the General Directorate of Examinations within the Ministry of Higher Education (MoHE). The administering authority is also responsible for grading all test-takers and finalizing the Kankor results before the start of each academic year. Applicants receive only one admission offer from the administrating authority, and it is almost impossible to transfer from one admitted concentration to another within and/or across public IHEs.\\n\\nThe Kankor exam comprises 160 multiple-choice questions derived from all subjects<sup>8</sup> taught in Afghan public high schools in grades 10–12<sup>9</sup> . The Kankor score range is 0–360 points; on average, each question is worth 2.25 points. From 2013–2018, the exam was held nationwide over several weeks, and the administering agency finalized admission decisions after all the exams were completed. The questions are drawn from a large question bank; thus, the questions vary across and within exam regions<sup>10</sup> , although the pro-\\n\\n> 7An open competition among all participants; it serves as a filtering mechanism to admit qualified students— based on their performance on the exam, measured by their score—into public universities.\\n\\n> 8The subjects taught in grades 10–12 in public schools are uniform across the country. They are developed by the central curriculum directorate at the Ministry of Education (MoE) in Kabul, Afghanistan, ensuring no variability in the topics covered across regions. However, there are disparities in school and teacher quality. Students in public schools do not have the option to enroll in specific subjects or tracks of study; instead, they must undertake all the courses offered for each grade.\\n\\n> 9While students in private schools receive instruction in additional coursework/topics, these schools are required to offer the fundamental subjects for each grade as designated by the MoE.\\n\\n> 10There is a slight chance that some questions from an exam taken earlier in region _K_ will appear in a later exam taken in region _L_ .\\n\\n8\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"portion of questions from each subject topic remains the same, regardless of the exam location and date. The exam lasts two hours and 40 minutes<sup>11</sup> , after which test-takers receive a selection sheet from the proctors<sup>12</sup> . The selection sheet contains a list of concentrations and their enrollment capacity, grouped by major or university department. After the first adoption of the gender quota in 2016, the treated concentrations’ enrollment capacity was separated by gender with a unique concentration gender-specific code, as seen in Figure A1.2.\\n\\nAt the end of the Kankor exam, applicants select and rank up to five concentrations despite not knowing their Kankor scores. The applicants enter the unique code for each concentration into the answer sheet. The administering authority uses this to finalize the matriculation decisions. The exam is open to all. Irrespective of the concentration type (e.g., medical or literature), applicants must achieve a qualifying score to ensure admission to their chosen concentrations. There is no fixed score threshold, as it can vary annually depending on the number of applicants for a particular concentration. After the nationwide Kankor examinations are concluded, the administering authority begins grading and finalizing admission decisions. They enroll high-scoring students in concentrations based on their scores until all available spots are filled. The gender quota adjusted the enrollment criteria by changing the focus from scores in treated concentrations to considering both scores and sex.\\n\\nTable 1 shows the descriptive statistics of annual high school graduates, Kankor exam participants, and the matriculation rate in Afghanistan for 2013–2018. On average, 220,000 students graduated from high school, among which 36% were girls, as shown in Panel A. On average, approximately 190,000 applicants participated in the Kankor exam in the between the 2013–2018 academic years (Panel B), of whom 33% were women. Female applicants were less likely to participate in the Kankor exam as a share of girls who graduated from high school, as shown in Column 4 of Panel B. Additionally, regarding\\n\\n> 11The NEXA authority has recently extended the exam duration to over 3 hours.\\n\\n> 12Public university lecturers/professors serve as the official proctors. The administering authority rotates proctors across regions to oversee the exam, although most are local lecturers and/or professors.\\n\\n9\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"matriculation into public and private IHEs, the GPI was 1:3 and 1:4, respectively. The share of women admitted was considerably worse than the share of female high school graduates, as seen in Column 4 of Panels C and D. Finally, female applicants were significantly more likely to score lower than male test-takers among all applicants, as seen in Panel E, and among matriculated students, as seen in Panel\\n\\n## F of Table 1.\\n\\nTable 1: Descriptive statistics for Kankor exam participants between 2013-2018\\n\\n|Mean|Standard|% female|/ female to|As a share of high|\\n|---|---|---|---|---|\\n|(1)|Dev.<br>(2)|male avg. <br>(3)|score ratio|school graduates<br>(Panel A) (4)|\\n|Panel A: High school graduates in _thousands_|||||\\n|Female<br>78.53|(9.77)||||\\n|Male<br>141.19|(15.13)||||\\n|Total<br>219.72|(24.49)|36%|||\\n|Panel B: Kankor test-takers/Applicants in _thous_<br><br>|_ands_<br>||||\\n|Female<br>62.24|(10.31)|||79%|\\n|Male<br>127.78|(19.22)|||90%|\\n|Total<br>190.02|(28.83)|33%||86%|\\n|Panel C: Newly admitted in public IHEs in _thou_<br>|_sands_<br>||||\\n|Female<br>14.64|(3.03)|||19%|\\n|Male<br>41.39|(1.35)|||29%|\\n|Total<br>56.04|(3.37)|26%||25%|\\n|Panel D: Newly admitted in private IHEs in _tho_|_usands_<br>||||\\n|Female<br>9.23|(1.83)|||12%|\\n|Male<br>35.99|(5.65)|||25%|\\n|Total<br>45.22|(7.47)|20%||20%|\\n|Panel E: Average score for Kankor applicants (0|-360)<br>||||\\n|Female<br>186.81|(44.66)||||\\n|Male<br>193.23|(48.43)||||\\n|Total<br>189.41|(50.44)|0.97|||\\n|Panel F: Average score for admitted students in|public IHE|(0-360)|||\\n|Female<br>235.66|(47.04)||||\\n|Male<br>236.47|(46.05)||||\\n|Total<br>236.26|(46.31)|0.99|||\\n\\nNote: the raw data is compiled from multiple sources, mainly Afghanistan yearly statistics books published by the National Statistics and Information Authority (NSIA), Ministry of Education (MoE), and the Kankor results for 2013–2018.\\n\\n# **3.2 Gender quota**\\n\\nThe gender quota for admission to public universities was first introduced in the 2016 academic year and gradually expanded until the Taliban returned to power in 2021. Female test-takers typically scored 10\", \"datasets\": [{\"mention_name\": {\"text\": \"Afghanistan yearly statistics books\", \"confidence\": 0.3489247262477875, \"start\": 1679, \"end\": 1714, \"clean_text\": \"Afghanistan yearly statistics books\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1679, \"end\": 1714, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1679, \"end\": 1714, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1679, \"end\": 1714}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1679, \"end\": 1714, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.4456219971179962, \"start\": 138, \"end\": 173}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3489247262477875, \"start\": 138, \"end\": 173}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7702587246894836, \"start\": 138, \"end\": 173}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7702587246894836, \"start\": 138, \"end\": 173}}, {\"mention_name\": {\"text\": \"Kankor results\", \"confidence\": 0.7904833555221558, \"start\": 1823, \"end\": 1837, \"clean_text\": \"Kankor results\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1823, \"end\": 1837, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1823, \"end\": 1837, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1823, \"end\": 1837}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1823, \"end\": 1837, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.4456219971179962, \"start\": 282, \"end\": 296}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7904833555221558, \"start\": 282, \"end\": 296}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7702587246894836, \"start\": 282, \"end\": 296}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7702587246894836, \"start\": 282, \"end\": 296}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"significantly lower than male participants, both among all Kankor participants and among those admitted to public IHEs, as shown in Panels E and F of Table 1. Najam & Johnston (2023) show that the proportion of female student enrollees in public IHEs has risen over time. Despite the increase in women’s access to higher education, public universities’ strict score-based admission criteria exacerbate gender inequality, especially in highly selective programs. This disparity in performance might stem from the systematic underinvestment and cultural barriers Afghan women face in accessing post-secondary education (Bamik, 2018; Hayward & Babury, 2015; Najam & Johnston, 2023), which is exacerbated by conflict (Darwish & Wotipka, 2022).\\n\\nThe MoHE has prioritized improving women’s access to higher education, as outlined in the Higher Education Strategic Plan II 2010–2014 (Ministry of Higher Education, 2009) and the HE Strategic Plan III 2016–2020 (Ministry of Higher Education, 2015). The introduction of a gender quota in 2016 aimed to reduce gender disparities in higher education and boost women’s participation in the country’s workforce, particularly in leadership roles in public offices. Yari & Schafer (2021) show that in 2018, Afghan women constituted approximately 22% of the Afghan civil service and held only 7.5% of the decision-making roles. Gender quotas are a crucial step in developing the human capital of the targeted group—in this case, enhancing the supply of qualified, educated women. The gender quota allocated, on average, 37% of seats for female test-takers in the treated concentrations at 12 out of 37 public universities<sup>13</sup> . Treated concentrations were selected based on the country’s need for professionals in specific fields of study, such as medicine, computer science, and public administration. Another consideration was whether female test-takers were less likely to secure admission in fields such as engineering, computer science, law, and political science due to intense competition in a score-based admission approach. Additionally, as shown in Table A3.1, the treated concentrations were primarily located at regional universities. Thus, the policy expanded educational opportunities for women in highly competitive and/or respected public\\n\\n> 13See Table A3.1 for the share of ever-treated concentrations in a given university.\\n\\n11\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"IHEs nationwide.\\n\\nIn 2016, the policy was implemented for Kankor exam test-takers applying for their freshman year in that academic year. The MoHE introduced a biometric system for applicant registration and verification<sup>14</sup> in the same year to address allegations of fraud and corruption against the Kankor exam in previous years. The MoHE publicly announced the biometric registration system, but not gender-specific changes, in the enrollment approach of treated concentrations in the first-year of implementation<sup>15</sup> . Therefore, the level of awareness among applicants regarding the changes in the enrollment method—from open competition to gender-specific considerations—remained ambiguous. The absence of a public announcement suggests that this information was not evenly distributed across applicants and regions, at least during the first year of adoption. This lack of advertisement has profound implications on students’ academic decisions. First, a lack of complete information beforehand might have led to misaligned expectations regarding the chances of acceptance. Second, students might have been excluded from the concentrations for which there was a higher likelihood of acceptance due to the gender quota allocation. Finally, students from privileged backgrounds might have accessed this information through informal channels, while other applicants remained uninformed, raising equity concerns.\\n\\nAdditionally, I report the t-balance test for the primary outcomes before the introduction of the gender quota in Table B3.1. The results show that the overall share of female applicants and the share of female applicants from low-SES groups who matriculated in treated concentrations were significantly lower than the control group mean. The quota policy, which serves as a quantity intervention, aims to improve the rate of female students by distributing seats to gender-specific proportions in treated concentrations. On average, this rate was greater than the actual share of female students in the period\\n\\n> 14This universal change does not threaten the identification strategy as it affects all applicants equally. However, the introduction of the biometric system might discourage certain groups from participating in the Kankor exam.\\n\\n> 15I verify that concentrations received official communication from the MoHE about the gender quota implementation. However, I found no evidence indicating that this decision was communicated to test-takers across the country, at least before the Kankor exam for the 2016 academic year.\\n\\n12\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"<!-- Start of picture text --> 80.00 —<br>eo ° @r @<br>60.00 — ®o °<br>g @ )<br>$s r) @<br>3z° ° © e@eoo” © wo @ee5 oe@ : @<br>2 ee o ®<br>8os 40.00 r) ® 8e %e @ow @ o® @ @<br>5 °. ®, e oe ® C) @<br>3 @ %®e U e ©<br>€ hd o& @e @<br>@x C82 @ ea)Co é @<br>20.00- @ 8 e @<br>e<br>@C)C © @ @ @ %female ;in concentration.<br>—— Fitted value<br>oo<br>I I I I I<br>0.00 20.00 40.00 60.00 80.00<br>% female enrolled pre-quota<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"details, see Appendix A.4) and by the expansion of academic programs<sup>16</sup> . Academic programs refer to the number of concentrations across universities that offer a particular academic degree<sup>17</sup> , such as economics, engineering, and other programs. I account for these changes in the empirical specifications by including various fixed effects. Finally, in principle, there may be discrepancies between matriculated applicants and those eventually attending public IHEs. Nevertheless, the gap is likely negligible considering the institutional features (e.g., limited transfer possibility, free tuition, and dormitories) and future return on education. It should not pose a significant threat to the consistency of the findings.\\n\\n# **4 Data**\\n\\nI compiled three datasets to assess the gender quota impact on outcomes of interest. First, I attained the Kankor matriculation dataset for 2013–2018, from publicly available sources to compute variables of interest. Second, I obtained district-level nightlights data from the Visible Infrared Imaging Radiometer Suite (VIIRS) Stray Light Corrected Nighttime Day/Night Band Composites Version 1 via Google Earth at the second administrative unit level. Finally, I used the school database in Afghanistan available on the Ministry of Education (MoE) website. It contains information on school names and district locations, allowing me to match test-takers to their school districts.\\n\\nFirst, in the Kankor dataset, I observed applicants’ gender, final test scores, high school name, high school graduation year, the province where the applicants took the exam, the institution’s name for those who matriculated into 2-year associate degrees, and university-unique-concentration identifier for applicants enrolled into 4-year public IHEs on 1,151,042 Kankor test-takers for 2013–2018. I limited the data to those applicants enrolled into the 4-year public universities (29% of all test-takers). The\\n\\n> 16See Table A2.1 for the conceptual difference between concentrations, academic programs, and universities.\\n\\n> 17See Table A3.2 for the share of ever-treated concentrations across academic programs.\\n\\n14\", \"datasets\": [{\"mention_name\": {\"text\": \"Kankor matriculation dataset\", \"confidence\": 0.9451686143875122, \"start\": 869, \"end\": 897, \"clean_text\": \"Kankor matriculation dataset\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 869, \"end\": 897, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 869, \"end\": 897, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"2013–2018\", \"confidence\": 0.5474706292152405, \"start\": 902, \"end\": 911}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 869, \"end\": 897, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.968558669090271, \"start\": 869, \"end\": 897}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.9451686143875122, \"start\": 869, \"end\": 897}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.6150999069213867, \"start\": 869, \"end\": 897}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6150999069213867, \"start\": 869, \"end\": 897}}, {\"mention_name\": {\"text\": \"district-level nightlights data\", \"confidence\": 0.6241025328636169, \"start\": 998, \"end\": 1029, \"clean_text\": \"district-level nightlights data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 998, \"end\": 1029, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 998, \"end\": 1029, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 998, \"end\": 1029}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 998, \"end\": 1029, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.5730166435241699, \"start\": 998, \"end\": 1029}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6241025328636169, \"start\": 998, \"end\": 1029}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5595988035202026, \"start\": 998, \"end\": 1029}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5595988035202026, \"start\": 998, \"end\": 1029}}, {\"mention_name\": {\"text\": \"Visible Infrared Imaging Radiometer Suite\", \"confidence\": 0.38219475746154785, \"start\": 1039, \"end\": 1080, \"clean_text\": \"Visible Infrared Imaging Radiometer Suite\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1039, \"end\": 1080, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1039, \"end\": 1080, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1039, \"end\": 1080}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1039, \"end\": 1080, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.5730166435241699, \"start\": 1039, \"end\": 1080}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.38219475746154785, \"start\": 1039, \"end\": 1080}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5595988035202026, \"start\": 1039, \"end\": 1080}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5595988035202026, \"start\": 1039, \"end\": 1080}}, {\"mention_name\": {\"text\": \"school database in Afghanistan\", \"confidence\": 0.7709285616874695, \"start\": 1235, \"end\": 1265, \"clean_text\": \"school database in Afghanistan\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1235, \"end\": 1265, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1235, \"end\": 1265, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1235, \"end\": 1265}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1235, \"end\": 1265, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9995489716529846, \"start\": 456, \"end\": 486}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.7709285616874695, \"start\": 456, \"end\": 486}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9378213882446289, \"start\": 456, \"end\": 486}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9378213882446289, \"start\": 456, \"end\": 486}}, {\"mention_name\": {\"text\": \"Kankor dataset\", \"confidence\": 0.680361807346344, \"start\": 1459, \"end\": 1473, \"clean_text\": \"Kankor dataset\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1459, \"end\": 1473, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1459, \"end\": 1473, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1459, \"end\": 1473}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1459, \"end\": 1473, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.6066467761993408, \"start\": 680, \"end\": 694}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.680361807346344, \"start\": 680, \"end\": 694}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7102290391921997, \"start\": 680, \"end\": 694}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7102290391921997, \"start\": 680, \"end\": 694}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"treatment status for a concentration was jointly identified using two primary sources: [1] the Kankor exam selection sheet produced by the administering authority as an example, is shown in Figure A1.2, and [2] the Kankor matriculation dataset, which explicitly denoted whether an applicant was admitted into the treated units. If there were any discrepancies between the two sources, the author contacted a specific university and the MoHE to identify the treatment status for these concentrations. For example, four religious’ studies units were admitting applicants based on gender before 2016 in the matriculation dataset. Nonetheless, these units historically enrolled students in a gender-specific manner for cultural reasons, and were not subject to the gender quota. Hence, these concentrations were not considered treated units because of the gender quota policy in the analytical sample.\\n\\nThe yearly descriptive statistics of Kankor participants—the share of female students and that of female students admitted into 4-year public IHEs are reported in Table B1.1. The number of concentrations increase over time, as shown in Table B1.1, which does not threaten the identification strategy, as I controlled for the concentration fixed effects. This increase was driven by the splitting of enrollment levels from general admission into several concentrations in some fields of study, the enlargement of academic programs, and the establishment of universities for further details (see Appendix A.4).\\n\\nMoreover, I obtained monthly nightlight data as an average of the Day/Night Band radiance values in nanoWatts/cm2/sr units from Stray Light Corrected VIIRS composited at the district level for 2014– 2015<sup>18</sup> . I excluded pixels with zero values across all images to increase the accuracy of the zonal statistics computations at district level. The use of nightlight as a proxy for economic development and measure of economic activities (Basihos, 2016; Beyer et al., 2018; Galdo et al., 2020; McCord & Rodriguez-Heredia, 2022; S ̈anger et al., 2022) is becoming popular. High economic development, which correlates with enhanced service provision, will enable women to access better educational services. Conversely, a low level\\n\\n> 18The Visible Infrared Imaging Radiometer Suite (VIIRS) Stray Light Corrected Nighttime Day/Night Band Composites Version 1 data were not available prior to 2014.\\n\\n15\", \"datasets\": [{\"mention_name\": {\"text\": \"Kankor matriculation dataset\", \"confidence\": 0.9008150696754456, \"start\": 215, \"end\": 243, \"clean_text\": \"Kankor matriculation dataset\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 215, \"end\": 243, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 215, \"end\": 243, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 215, \"end\": 243}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 215, \"end\": 243, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.972892701625824, \"start\": 215, \"end\": 243}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.9008150696754456, \"start\": 215, \"end\": 243}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8692434430122375, \"start\": 215, \"end\": 243}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8692434430122375, \"start\": 215, \"end\": 243}}, {\"mention_name\": {\"text\": \"yearly descriptive statistics of Kankor participants\", \"confidence\": 0.48913028836250305, \"start\": 903, \"end\": 955, \"clean_text\": \"yearly descriptive statistics of Kankor participants\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 903, \"end\": 955, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 903, \"end\": 955, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 903, \"end\": 955}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 903, \"end\": 955, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.5911816954612732, \"start\": 903, \"end\": 955}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.48913028836250305, \"start\": 903, \"end\": 955}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6147241592407227, \"start\": 903, \"end\": 955}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6147241592407227, \"start\": 903, \"end\": 955}}, {\"mention_name\": {\"text\": \"monthly nightlight data\", \"confidence\": 0.6846283078193665, \"start\": 1530, \"end\": 1553, \"clean_text\": \"monthly nightlight data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1530, \"end\": 1553, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1530, \"end\": 1553, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1530, \"end\": 1553}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1530, \"end\": 1553, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.28853967785835266, \"start\": 651, \"end\": 674}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6846283078193665, \"start\": 651, \"end\": 674}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.41625165939331055, \"start\": 651, \"end\": 674}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.41625165939331055, \"start\": 651, \"end\": 674}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"of nightlight is indicative of dire economic situations and disruptions in public services (S ̈anger et al., 2022). Nightlight provides a rough indicator. However, given the lack of granular metrics at the applicant and/or district levels in Afghanistan, nightlight data provides a plausible indicator of economic development at the district level. Therefore, I cannot differentiate between applicants’ socioeconomic statuses at the individual level and instead rely on district classifications. I used the two-year pre-treatment average luminosity by district as a binary indicator to identify low-SES applicants, operationalized as districts with luminosity lower than the mean (0.2678 nanoWatts/cm2/sr) in the pre-treatment period. I used this indicator to examine the impact of quotas on the share of female students from low SES. To achieve this objective, I successfully matched 95.4%<sup>19</sup> of all matriculated applicants (336,151) to their districts of origin. I accomplished this by merging student records at the district level based on school and province names (i.e., the province from which they participated in the Kankor exam) with the applicant’s high school districts using the comprehensive school database from the MoE.\\n\\n# **5 Empirical Strategy**\\n\\nThe gender quota implementation started in 2016 and expanded in the following years. The quota distributed the available seats at some concentrations based on gender but maintained open competition at other concentrations. Thus, I exploited the variation across concentrations and over time to study the causal impact of the gender quota on the share of matriculated female students and score-related outcomes in treated units using the following DD model shown in Equation 1:\\n\\n> 19The reasons for the 4.6% of unmatched applicants include multiple schools located in different districts within the same province. However, labels for these schools are entered identically without any unique district information, making it difficult to identify the exact school from which an applicant graduated in the matriculation dataset. Additionally, applicant school names are missing, either because they were not entered, or the applicants completed their high school studies in neighboring countries (e.g., Pakistan and Iran).\\n\\n16\", \"datasets\": [{\"mention_name\": {\"text\": \"nightlight data\", \"confidence\": 0.5170931816101074, \"start\": 255, \"end\": 270, \"clean_text\": \"nightlight data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 255, \"end\": 270, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 255, \"end\": 270, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 255, \"end\": 270}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 255, \"end\": 270, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.8782165050506592, \"start\": 255, \"end\": 270}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5170931816101074, \"start\": 255, \"end\": 270}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.7296268343925476, \"start\": 255, \"end\": 270}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7296268343925476, \"start\": 255, \"end\": 270}}, {\"mention_name\": {\"text\": \"two-year pre-treatment average luminosity by district\", \"confidence\": 0.4211374521255493, \"start\": 507, \"end\": 560, \"clean_text\": \"two-year pre-treatment average luminosity by district\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 507, \"end\": 560, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 507, \"end\": 560, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 507, \"end\": 560}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 507, \"end\": 560, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9915724992752075, \"start\": 507, \"end\": 560}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.4211374521255493, \"start\": 507, \"end\": 560}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9373306632041931, \"start\": 507, \"end\": 560}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9373306632041931, \"start\": 507, \"end\": 560}}, {\"mention_name\": {\"text\": \"comprehensive school database from the MoE\", \"confidence\": 0.5416931509971619, \"start\": 1201, \"end\": 1243, \"clean_text\": \"comprehensive school database from the MoE\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1201, \"end\": 1243, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1201, \"end\": 1243, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1201, \"end\": 1243}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1201, \"end\": 1243, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9989142417907715, \"start\": 341, \"end\": 383}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.5416931509971619, \"start\": 341, \"end\": 383}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8448651432991028, \"start\": 341, \"end\": 383}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8448651432991028, \"start\": 341, \"end\": 383}}, {\"mention_name\": {\"text\": \"matriculation dataset\", \"confidence\": 0.5648964047431946, \"start\": 2076, \"end\": 2097, \"clean_text\": \"matriculation dataset\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2076, \"end\": 2097, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2076, \"end\": 2097, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2076, \"end\": 2097}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2076, \"end\": 2097, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.8923074007034302, \"start\": 428, \"end\": 449}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5648964047431946, \"start\": 428, \"end\": 449}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7568444609642029, \"start\": 428, \"end\": 449}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7568444609642029, \"start\": 428, \"end\": 449}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"The _Yisut_ are the outcomes of interest for a concentration _i_ at university _s_ , academic program _u_ , and time _t_ . _β_ is the estimate of interest or the quota policy effects, and _λ_ is the fixed effect for the concentration _i_ at university _s_ and academic program _u_ ; _ρ_ is the fixed effect for university _s_ in time _t_ , _γ_ is the fixed effect of academic programs _u_ in time _t_ , where _θ_ is the year fixed effect<sup>20</sup> , and _εisut_ is the error term, and models are estimated with the cluster standard error at the concentration level.\\n\\nConcentrations, academic programs, and institutional-level selectivity are critical aspects that must be considered when studying matriculation in post-secondary education at the national level using a centralized matriculation system. This issue is more prominent in developing nations, where applicants are more likely to choose prestigious fields of study, causing the admission process for specific concentrations to become more competitive than others, _λisu_ , (the concentration-university-academic program fixed effect) controls for systematic differences across concentrations.\\n\\nThe university-by-year fixed effect _ρst_ captures any factor affecting an entire university in a particular year, such as changes in security conditions in the area where the university is located or universitylevel infrastructure development (e.g., the expansion of girls’ dormitories). For instance, the gradual withdrawal of the International Security Assistance Force from across the country to pre-designated locations under the Resolute Support Mission by the end of 2014, led to deteriorating security conditions in regions where security transitions occurred (Fetzer et al., 2021). Consequently, some universities become less favorable to applicants due to such regional shocks, which are likely to change the composition of applicants choosing a given university in year _t_ . Additionally, university-level infrastructures development (e.g., buildings and dormitories) is another factor that is likely to change the composition of students selecting a given university. The _ρst_ captures any regional shocks, the effects of which vary over time and across universities.\\n\\nThe academic program by year fixed effect _γut_ controls for any unobservable factor affecting an\\n\\n> 20This can be dropped in the model as I control for university-year and academic program-year fixed effects.\\n\\n17\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"academic program over time and is not university-specific. There are 17 academic programs across the 37 public universities, as shown in Table A3.2. For instance, if there is an expansion of a particular academic program (e.g., engineering or medicine) in time _t_ , such expansion increases the availability of choices for applicants. Thus, an applicant who is determined to study a specific concentration (e.g., civil engineering, econometrics) will likely utilize the expansion and select their preferred concentration in newly established academic programs across universities. This factor changes the composition of applicants enrolled in specific academic programs over time and is controlled by _γut_ .\\n\\nThe inclusion of these fixed effects is important for the DD estimator to differentiate trends and account for common shocks to attain a credible estimate of the quota policy’s impact on the outcomes of interest. The identification strategy provides a causal effect as long as the central assumption of parallel trends between the treated and untreated units in the absence of the policy (Angrist & Pischke, 2009; Cunningham, 2021; Huntington-Klein, 2022; Morgan & Winship, 2014), conditional on fixed effects, is not violated. If this assumption holds, we expect that the outcome trends between the treated and comparison groups, conditional on fixed effects, would follow a similar pattern to that between the two groups in the pre-treatment period. To check for parallel pre-treatment trends, I ran the following event study model: In Equation 2, the baseline year is the year before to the policy adoption in concentration _isu_ , I allowed for dynamic leads and lags, and the remaining terms are described in Equation 1. I find plausible evidence from the event study analysis reported in Appendix B.4, that the trends between the treated and controls units, conditional on fixed effects, are not statistically different in the pre-treatment period. The policy was implemented without advance announcements before its enactment in 2016, limiting any anticipation effect.\\n\\n18\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"Another concern was the staggered implementation of quota adoption in some concentrations in the post-2016 period, as discussed in recent literature (Borusyak et al., 2022; Callaway & Sant’Anna, 2021; de Chaisemartin & D’Haultfœuille, 2020; Goodman-Bacon, 2021; Sun & Abraham, 2021), which suggest the conventional two-way fixed effect may fail to provide unbiased estimates. In this study setting, differential treatment timing does not pose a significant threat to the estimates because most concentrations adopted a quota policy in 2016, with a few late adopters, which were mostly newly established concentrations with no variation in the treatment status and absorbed by the concentrationuniversity-academic program fixed effect. Additionally, once a concentration fixes the allocated slots by gender, it remains treated, and the allocation level does not change, reducing the risk posed to the TWFE estimate stated in recent literature. Nevertheless, I report the findings and event study results using the estimators of Borusyak et al. (2022) and Sun & Abraham (2021).\\n\\n# **6 Results**\\n\\n## **6.1 Main results**\\n\\nTable 2 reports the overall impact of gender quotas on women’s access to educational opportunities and score-related outcomes. The quota increased the share of female students at the treated concentrations by nine percentage points, a 32% increase relative to the comparison group mean, as seen in Column 1. Specifically, the quota improved the share of female students from low-SES backgrounds by 3 percentage points, a 30% increase compared to the control group mean, as shown in Column 2. These findings highlight the effectiveness of the quota in closing the gender gap in higher education by amplifying women’s educational opportunities.\\n\\nNonetheless, as seen in Columns 3–8, score-related outcomes indicate that such an expansion is asso19\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"ciated with a reduction in the score threshold required for women’s admissions in treated concentrations. As result of the quota, I found that the average score of matriculated female students in treated units declined by 11 points, a 5% decrease relative to the comparison group mean. This decrease was driven by a substantial decline of 14 points in the average score of female applicants from the high-SES group, representing a 6% decrease compared to the control group mean. The average score of female students admitted from low SES decreased by 5 points or 2% (Column 6 in Table 2).\\n\\nThere were no significant changes in the average scores of the admitted male applicants in the treated concentrations. This result is unexpected because, owing to the allocation of seats at these concentrations<sup>21</sup> , the competition among male applicants should have increased, increasing the average score of admitted male students. There are two potential reasons for this. First, while the policy allocated, on average, 37% of seats across treated units to female students, the average proportion of female students in the post-treatment period was 33%. This indicates that not all seats were actually filled by female applicants<sup>22</sup> . Systematic errors in grading and finalizing admission decisions led to the admission of male instead of female test-takers. Second, it is possible that the displaced male applicants might have subsequently been admitted to the control concentrations, which might have changed the average score of the untreated concentrations in a direction similar to that of the treated units. Consequently, I could not detect significant changes in the average scores of male students at the treated concentrations. The findings show that the average gender-specific score threshold across treatment statuses was closer to the\\n\\n> 21I find no significant changes in the total number of students enrolled in the treated concentrations (results not reported here). Therefore, as the number of seats available for male applicants in these concentrations is more limited than it would have been in the absence of the gender quota, one might expect the average score of male students to increase. However, this is not the case.\\n\\n> 22In fact, there were systematic errors in finalizing admission decisions. In the post-quota period, in some cases, male test-takers either mistakenly or intentionally entered concentration-female specific codes in lieu of concentration-male unique codes, as seen in Figure A1.2 on their selection sheet. The administering authority used a code reader system that matched the applicants’ answer sheet with the pre-programmed answer key and admitted applicants to their preferred concentration, as long as they met the score threshold. In the first few years of the quota policy, the administering authority admitted applicants without verifying their actual gender in the database; consequently, male applicants were enrolled instead of female test-takers. Nevertheless, this issue has been resolved by the NEXA authority in recent years.\\n\\n20\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"minimum score, indicating a higher density of applicants in the lower tail, as shown in Table B2.1.\\n\\nThe strict score-based approach suggests that female applicants must score almost the same points as male Kankor test-takers to secure admission into public IHEs. Table B3.1 shows that the female-to-male students’ average score ratio in the pre-treatment period was almost the same across treatment statuses. To determine whether the quota changed the composition of students within the treated concentrations, I assessed the changes in the female-to-male students’ average score ratio and report the results in Column 8 of Table 2. The negative coefficient indicates that, on average, female applicants admitted into the treated concentrations had scores that were 0.04 units lower than their male counterparts, representing a 4% drop compared to the control group mean. A decrease in the score ratio suggests that the quota reduced the score threshold required for the admission of female applicants. Despite a 4% decrease in the average female-to-male student average score ratio in the short term, it may rebound in the long term. This is likely due to the indirect effects of affirmative action on improving pre-college human capital investments, as shown by Khanna (2020) and Akhtari et al. (ming). Finally, the OLS findings are robust to staggered DD estimators, as reported in Table 2, and alternative treatment specifications (major level<sup>23</sup> ), as shown in Table B5.1.\\n\\n> 23For further details about how major level analysis differ from concentration level analysis please see Data Appendix in A.4\\n\\n21\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"Table 2: Effects of gender quota on outcomes at the concentration level\\n\\n||Acce|ss to IHEs|||Score related|outcomes|||\\n|---|---|---|---|---|---|---|---|---|\\n||Share<br>females|Share females<br>low-SES|Total avg.<br>score|Female avg.<br>score|High-SES fe-<br>male avg. score|Low-SES<br>female<br>avg. score|Male avg.<br>score|Female to male<br>avg. score ratio|\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|\\n|Estimators|||||||||\\n|OLS|0.09***|0.03***|-3.21*|-10.92***|-14.01***|-4.98**|-0.52|-0.04***|\\n||(0.02)|(0.01)|(1.66)|(2.11)|(2.48)|(2.40)|(1.79)|(0.01)|\\n|Borusyak et al. (2021)<sup>_ζ_</sup>|0.09***|0.04***|-1.15|-11.00***|-10.26***|-7.36***|1.88|-0.05***|\\n||(0.02)|(0.01)|(1.96)|(1.98)|(2.10)|(2.01)|(2.05)|(0.01)|\\n|Sun-Abraham (2020)|0.08***|0.025***|-2.32|-9.43***|-12.53***|-3.98|0.10|-0.04***|\\n||(0.018)|(0.006)|(1.81)|(2.19)|(2.39)|(2.87)|(1.99)|(0.007)|\\n|Concentration FE|✓|✓|✓|✓|✓|✓|✓|✓|\\n|University*year FE|✓|✓|✓|✓|✓|✓|✓|✓|\\n|Academic pro*year FE|✓|✓|✓|✓|✓|✓|✓|✓|\\n|Year FE|✓|✓|✓|✓|✓|✓|✓|✓|\\n|_Control group mean_|0.28|0.10|225|224|228|221|225|0.99|\\n|N|3119|3119|3119|3119|2723|2699|3119|3119|\\n\\nNote: Each coefficient is from a separate regression. The dependent variables in columns 1 and 2 are the share of females and share of females from low socioeconomic backgrounds, respectively, which can be interpreted as average percentage point increase in the share of female in treated concentrations due to the quota introduction. In columns 3–7, the dependent variables are overall average score, average score for females, average score for high- and low-SES females, and male students average score at concentration level, respectively; it can be interpreted as points decrease/increase. Column 8 denotes female to male average score ratio at concentration level. All the models include stated fixed effects. Standard errors are clustered at concentration level and reported in parentheses. Significance level: *p _<_ 0.1; **p _<_ 0.05; ***p _<_ 0.01 _ζ_ Owing to the imputing nature of the estimator by Borusyak et al. (2021), the actual number of observations used to estimate the result is slightly lower than that with the OLS and Sun-Abraham (2020). The _N_ represents the number of observations used to compute the OLS and Sun-Abraham (2020) estimates.\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"# **6.2 Heterogeneity result**\\n\\nIn this section, I report on the effects of the gender quota on women’s access and score-related outcomes based on the competitiveness of academic programs. Table 3 presents the heterogeneity results. I classified concentrations into two groups based on the marketability and social value ascribed to these fields of study in Afghan society. Highly selective academic programs include computer science, economics, engineering, law and political science, and medicine. These are collectively known as CEELM programs<sup>24</sup> across public universities. The t-test differences in the main outcomes of interest between the CEELM and non-CEELM programs in the pre-2016 period is shown in Table B3.2. Notably, the share of female students and female students from low-SES backgrounds was 10 and six percentage points lower, respectively, in CEELM than in non-CEELM programs. Similarly, the average scores of matriculated applicants in CEELM programs surpassed the average scores of enrolled students in non-CEELM programs. Given the competitive nature of CEELM programs and the lower share of female students admitted to them, it is crucial to highlight the effectiveness of the gender quota on women’s access to highly selective programs. Enhancing women’s educational opportunities in these marketable programs is essential to close the gender gap in competitive academic programs and improve the long-term supply of qualified women in the labor force.\\n\\nThe impact of the gender quota on the outcomes of interest is reported in Panel A for the CEELM concentrations and Panel B for the non-CEELM concentrations in Table 3. Column 1 shows that the quota increased the share of female students in the treated CEELM concentrations by 17 percentage points, a 94% increase relative to the control group mean. Conversely, the gender quota raised the share of female students in non-CEELM concentrations by six percentage points, a 19% increase relative to the comparison group mean.\\n\\n> 24Among the CEELM programs, concentrations in economics, and law and political science are not treated.\\n\\n23\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"Table 3: Heterogeneous effects of gender quota on outcomes at the concentration level\\n\\n||Acc|ess to IHEs|||Score related|outcomes|||\\n|---|---|---|---|---|---|---|---|---|\\n||Share<br>females|Share females<br>low-SES|Total avg.<br>score|Female avg.<br>score|High-SES fe-<br>male avg. score|Low-SES<br>female<br>avg. score|Male avg.<br>score|Female to male<br>avg. score ratio|\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|\\n|Estimators|||||||||\\n|Panel A: CEELM|||||||||\\n|OLS|0.17***|0.04***|-4.36|-17.94***|-21.12***|-3.65|0.88|-0.07***|\\n||(0.03)|(0.01)|(3.46)|(5.19)|(6.43)|(7.77)|(3.23)|(0.02)|\\n|Borusyak et al. (2021)<sup>_ζ_</sup>|0.17***|0.04***|-0.35|-18.12***|-18.06***|-9.09*|5.57|-0.09***|\\n||(0.02)|(0.01)|(4.17)|(4.44)|(4.61)|(4.78)|(4.18)|(0.01)|\\n|Sun-Abraham (2020)|0.186***|0.045***|-1.69|-13.11**|-17.72***|-6.37|4.48|-0.06***|\\n||(0.036)|(0.008)|(3.44)|(5.22)|(6.27)|(9.93)|(3.17)|(0.01)|\\n|_Control group mean_|0.18|0.05|260|259|260|256|260|1|\\n|N|748|748|748|748|688|528|748|748|\\n|Panel B: Non-CEELM|||||||||\\n|OLS|0.06***|0.02***|-2.27|-7.84***|-10.99***|-4.88*|-0.52|-0.03***|\\n||(0.02)|(0.01)|(1.84)|(2.14)|(2.57)|(2.53)|(2.08)|(0.01)|\\n|Borusyak et al. (2021)<sup>_ζ_</sup>|0.05***|0.04***|-2.10|-7.80***|-6.75***|-6.64***|-0.48|-0.03***|\\n||(0.02)|(0.01)|(1.75)|(1.87)|(2.01)|(1.87)|(1.84)|(0.01)|\\n|Sun-Abraham (2020)|0.046**|0.019**|-1.47|-6.98***|-10.61***|-3.128|-0.03|-0.03***|\\n||(0.02)|(0.008)|(2.02)|(2.32)|(2.60)|(2.97)|(2.33)|(0.008)|\\n|_Control group mean_|0.31|0.12|216|215|219|214|217|0.99|\\n|N|2339|2339|2339|2339|2007|2140|2339|2339|\\n|Concentration FE|✓|✓|✓|✓|✓|✓|✓|✓|\\n|University*year FE|✓|✓|✓|✓|✓|✓|✓|✓|\\n|Academic pro*year FE|✓|✓|✓|✓|✓|✓|✓|✓|\\n|Year FE|✓|✓|✓|✓|✓|✓|✓|✓|\\n\\nNote: Highly competitive academic programs encompass concentrations within computer science, economics, engineering, law and political science, and medical studies, collectively referred to as CEELM programs. Each coefficient is from a separate regression. The dependent variables in columns 1 and 2 are the share of females and share of females from low socioeconomic backgrounds, which can be interpreted as average percentage point increase in the share of females in treated concentrations owing to the quota introduction. In columns 3–7, the dependent variables are overall average score, average score for females, average score for high- and low-SES females, and male students average score at concentration level, respectively; it can be interpreted as points decrease/increase. Column 8 denotes female to male average score ratio at concentration level. All the models in Panels A and B include the above stated fixed effects. Standard errors are clustered at concentration level and reported in parentheses. Significance level: *p _<_ 0.1; **p _<_ 0.05; ***p _<_ 0.01 _ζ_ Owing to the imputing nature of the estimator by Borusyak et al. (2021), the actual number of observations used to estimate the result is slightly lower than that with the OLS and Sun-Abraham (2020). The _N_ represents the number of observations used to compute the OLS and Sun-Abraham (2020) estimates.\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [27]}}, {\"page\": 28, \"chunk\": 28, \"input_text\": \"The gender quota increased the admission rate of female students from low SES by four percentage points or an 80% rise relative to the control mean in CEELM concentrations, as seen in Column 2. Meanwhile, the increase in the matriculation rate of female students from low SES in non-CEELM concentrations was two percentage points, a 16% increase compared with the control group mean.\\n\\nThe findings for the score-related outcomes indicate that expanding educational opportunities for women in CEELM concentrations correspond to a greater decrease in the score threshold required for admission compared to non-CEELM concentrations, as seen in Column 8 in Table 3. This trend is primarily attributed to a significant reduction in the average score of matriculated high-SES female students, as shown in Column 5. These results suggest that the gender quota is crucial to improving women’s access to highly competitive concentrations, benefiting the female student population, especially those from low-SES backgrounds.\\n\\nI examine the effect of the gender quota on the admissions of male applicants and their average scores from low- and high-SES backgrounds for CEELM and non-CEELM concentrations. These findings are summarized in Table 4. The quota policy allocated seats by gender without expanding the total enrollment. The overall increase in female applicants is associated with a significant displacement of male students in (non)CEELM concentrations. However, the rise in the rate of low-SES female students reported in Table 3 does not imply a significant drop in the admission rate of male applicants from lowSES backgrounds, specifically in CEELM concentrations, as shown in Column 2 of Table 4. Additionally, I do not observe significant changes in the average scores of male applicants from low- or high-SES backgrounds across CEELM and non-CEELM concentrations, as shown in Columns 3 and 4 of Table 4.\\n\\n25\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [28]}}, {\"page\": 29, \"chunk\": 29, \"input_text\": \"Table 4: Effects of gender quota on male students’ outcomes at the concentration level\\n\\n||Acc|ess to IHEs|Score relat|ed outcomes|\\n|---|---|---|---|---|\\n||Share<br>males|Share males<br>low-SES|High-SES male<br>avg. score|Low-SES male<br>avg. score|\\n||(1)|(2)|(3)|(4)|\\n|Estimators|||||\\n|Panel A: CEELM|||||\\n|OLS|-0.17***|-0.04|1.20|-0.80|\\n||(0.03)|(0.03)|(2.91)|(4.08)|\\n|Borusyak et al. (2021)<sup>_ζ_</sup>|-0.17***|-0.11***|5.64|5.76|\\n||(0.02)|(0.02)|(4.24)|(4.07)|\\n|Sun-Abraham (2020)|-0.18***|-0.05|3.82|4.08|\\n||(0.03)|(0.03)|(2.87)|(4.50)|\\n|_Control group mean_|0.82|0.39|260|259|\\n|N|748|748|748|746|\\n|Panel B: Non-CEELM|||||\\n|OLS|-0.06***|-0.03*|-1.24|-0.52|\\n||(0.02)|(0.01)|(2.07)|(2.38)|\\n|Borusyak et al. (2021)<sup>_ζ_</sup>|-0.05***|-0.04**|-1.18|-0.53|\\n||(0.02)|(0.02)|(1.78)|(1.93)|\\n|Sun-Abraham (2020)|-0.046**|-0.02|-0.96|0.19|\\n||(0.02)|(0.015)|(2.27)|(2.73)|\\n|_Control group mean_|0.69|0.41|217|216|\\n|N|2339|2339|2299|2334|\\n|Concentration FE|✓|✓|✓|✓|\\n|University*year FE|✓|✓|✓|✓|\\n|Academic pro*year FE|✓|✓|✓|✓|\\n|Year FE|✓|✓|✓|✓|\\n\\nNote: Highly competitive academic programs encompass concentrations within computer science, economics, engineering, law and political science, and medical studies, collectively referred to as CEELM programs. Each coefficient is from a separate regression. The dependent variables in columns 1 and 2 are the share of males and share of males from low socioeconomic backgrounds, which can be interpreted as average percentage point increase in the share of males students in treated concentrations owing to the quota. In columns 3–4, the dependent variables are average score for high- and low-SES male students, respectively; it can be interpreted as points decrease/increase. All the models in Panels A and B include the above stated fixed effects. Standard errors are clustered at concentration level and reported in parentheses. Significance level: *p _<_ 0.1; **p _<_ 0.05; ***p _<_ 0.01 _ζ_ Owing to the imputing nature of the estimator by Borusyak et al. (2021), the actual number of observations used to estimate the result is slightly lower than that with the OLS and Sun-Abraham (2020). The _N_ represents the number of observations used to compute the OLS and Sun-Abraham (2020) estimates.\\n\\n# **7 Rationale for Quota Impact**\\n\\nIn this section, I use a supply-and-demand framework to explain why the findings correspond with the anticipated impact of a gender quota on education. This allows us to understand the underlying 26\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [29]}}, {\"page\": 30, \"chunk\": 30, \"input_text\": \"mechanisms driving overall and socioeconomic status-specific impacts. In Figure 2, the horizontal axis denotes the fixed number of slots available at concentrations. The left and right vertical axes indicate the required score thresholds for the admission of female and male applicants, respectively ( _Sf_ and _Sm_ denote the supply curves for female and male applicants, respectively). In an open competition, outcome _A_ indicates the average gender-specific proportion in a concentration, where the share of female students equals _Q_<sup>_∗_</sup> _c_<sup>,andtheshareofmalestudentsequals1</sup><sup>_−Q∗_</sup> _c_<sup>.Thescorethresholdforadmissionissetto</sup> _P_<sup>_∗_</sup> . Hence, applicants must score at least as high as _P_<sup>_∗_</sup> to secure admission into a concentration. The gender quota acts as a quantity restriction, distributing the number of seats available according to gender. Specifically, the gender quota, on average, fixes the proportion of seats in treated concentrations for female students to _Q_<sup> ̄</sup> _s_<sup>25</sup> , represented by a dashed vertical line in Figure 2.\\n\\nFigure 2: Effects of gender quota on admissions and score thresholds <!-- Start of picture text --> scoref scorem<br>E Pm<br>A<br>P ∗<br>B<br>Pf<br>Sm Sf<br> ̄<br>Q ∗ c Qs<br>Female 0 Q ̄<br>Male Q ̄ 0<br><!-- End of picture text --> Therefore, the quota expands the share of female students from _Q_<sup>_∗_</sup> _c_<sup>to</sup> _Q_<sup> ̄</sup> _s_ . The distribution of seats can change the score threshold required for gender-specific admissions in treated concentrations. The 25On average, 37% of the available seats are reserved for female applicants in treated units.\\n\\n27\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [30]}}, {\"page\": 31, \"chunk\": 31, \"input_text\": \"gender quota increases the number of seats for female applicants by reducing the availability of seats for male applicants. As a result of these changes, the competition for women might be reduced, lowering the score threshold required for admission from _P_<sup>_∗_</sup> to _Pf_ . Conversely, the score for male applicants rises from _P_<sup>_∗_</sup> to _Pm_ . However, not all reserved slots for female applicants were actually filled by women, and male applicants matriculated in numbers greater than their allocated shares under the gender quota. These discrepancies might explain the insignificant changes in the average scores of male applicants.\\n\\nOne may argue that quota intervention distorts the competitive equilibrium by admitting female applicants with lower test scores who would not otherwise been enrolled. This drop in the score might lead to concerns, including a mismatch, which indicates that admitted students might not thrive academically because of preferential treatment. However, I argue that the share of female students under the competitive equilibrium _Q_<sup>_∗_</sup> _c_ is below the socially optimal level. The low likelihood of admission and substantial cultural barriers can discourage female applicants from investing in human capital during the pre-Kankor period, specifically, female applicants from low SES. In Figure 3, the y-axis denotes the admission cost for female students (i.e., the time and financial investment to prepare for the Kankor exam), and the x-axis shows the number of female students matriculated in a concentration. In a competitive equilibrium, the share of female students enrolled (denoted by _A_ ) is given by the intersection of supply curve _S_ 1 and the vertical solid line of _Qc_ , in which the average admission cost equals _C_<sup>_∗_</sup> .\\n\\nIn this state, female applicants _Cl_ , who face higher admission costs with lower private marginal benefits (PMB), might have reduced incentives to invest in pre-Kankor human capital development. Decreased investment in pre-Kankor human capital by female applicants from low SES can exacerbate existing educational and economic inequalities. The gender quota reserved seats based on gender, moving the supply curve to _S_ 2, in which the number of available seats for female applicants is fixed at _Qs_ . This level is considered a socially optimal output, with the outcomes of _L_ and _H_ at the extensive and intensive 28\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [31]}}, {\"page\": 32, \"chunk\": 32, \"input_text\": \"margins, respectively. The extensive margin enhances the enrollment prospect for female applicants from low-SES backgrounds, who face higher admission costs (e.g., limited access to Kankor preparation, social barriers, and household responsibilities) by lowering the competition level. By reducing the competition level, the gender quota potentially levels the playing field, providing applicants with a better opportunity to enroll in treated concentrations that might not previously have been considered (Fryer & Loury, 2005). Furthermore, this can encourage pre-Kankor human capital development, as shown by Khanna (2020) and Akhtari et al. (ming).\\n\\nFigure 3: Low-SES female students as marginally admitted applicants <!-- Start of picture text --> Admission cost<br>female students S 1 S 2<br>Cl L<br>A<br>C ∗<br>Ch H<br>PMB<br>Qc Qs Qf<br><!-- End of picture text --> Conversely, on the intensive margin, the gender quota may likely deter incentives for a subgroup of female applicants who would have pursued these concentrations, even without the quota. This subgroup of students likely came from high-SES backgrounds and tended to have lower admission costs (e.g., greater access to Kankor preparation and fewer social barriers as they might reside in urban areas) and a higher probability of admission (e.g., receiving a better quality of education in high school). In the post-quota period, the calculation of the PMB may have been lower than that in the pre-quota period. This might be driven by the expectation that more qualified women will potentially graduate and enter the job market, 29\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [32]}}, {\"page\": 33, \"chunk\": 33, \"input_text\": \"increasing competition among women in the labor force. Therefore, the changes in score-related outcomes for female applicants from high SES backgrounds shed light on their responses to the adoption of the gender quota.\\n\\nThe gender quota in Afghanistan was introduced to broaden educational opportunities for women. This affirmative action was adopted in the anticipation that educated women would bolster economic equality, contribute to national development, and improve social welfare. Numerous studies highlight that investment in women’s education can positively contribute to a country’s gross domestic product (Chaaban & Cunningham, 2011), women’s roles in intrahousehold financial and non-financial decisions (Le & Nguyen, 2021), and overall family and child well-being (Hill & King, 1995; Hobcraft, 1993). Additionally, such investments tend to have a higher return (Asadullah, 2006). Figure B6.1 shows that improving educational opportunities for women, especially those from low-SES backgrounds, can lead to greater social benefits, particularly in fragile settings where systematic underinvestment in women’s education limits their economic and social contributions to the overall well-being of society.\\n\\n# **8 Discussion and Conclusion**\\n\\nI assessed the impact of gender admission quota on women’s access to Afghan public IHEs. The findings show that the quota increased gender parity in public higher education by expanding educational opportunities for women, particularly by increasing the participation rate of women from low-SES backgrounds. Moreover, the results highlight that this intervention, at least in the short term, reduced the admission score threshold for women, predominantly for applicants from high SES backgrounds. A simple supply-and-demand model can explain the mechanisms behind the expected and observed impacts of the gender quota. These findings are of significant policy importance, as this is one of the first studies to rigorously examine the impact of affirmative action on women’s access to higher education in fragile and 30\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [33]}}, {\"page\": 34, \"chunk\": 34, \"input_text\": \"conflict-affected settings.\\n\\nThe gender quota policy increased the share of female applicants in the treated concentration group by 32% and the share of female students matriculated from low-SES backgrounds by 30%, relative to the control group mean. These results are consistent with the results of previous studies (Alon & Malamud, 2014; Bagde et al., 2016; Bertrand et al., 2010; de Silva et al., 2021; Mello, 2022). Additionally, I found that the gender quota almost doubled the share of female students admitted (94%) and that of female applicants enrolled from low-SES backgrounds (80%) in highly competitive academic programs. These findings are consistent with those of Bagde et al. (2016) and Vieira & Arends-Kuenning (2019).\\n\\nReducing the score threshold required for women’s admission might be a source of concern, particularly regarding the academic preparation and success of matriculated applicants. This is an avenue for future research to assess the impact of the gender quota on post-matriculation outcomes<sup>26</sup> . However, Khanna (2020) and Akhtari et al. (ming) show that affirmative action creates prospects for marginalized groups, which, in turn, can increase minorities’ schooling years and pre-college human capital investment via the enlargement of opportunities, as suggested by Fryer & Loury (2005). Therefore, expanding the matriculation rate of female students in public IHEs could have increased pre-Kankor investment among female applicants, mainly those from low-SES backgrounds. Nevertheless, if the attrition rate among matriculated female students due to the quota is higher than expected, affirmative action could be costly, as it distributes scarce educational resources without significant societal impact, as indicated by Arcidiacono & Lovenheim (2016) and Fryer & Loury (2005).\\n\\nThis study highlights the vital role of gender quotas in increasing educational opportunities for women in fragile and conflict-affected settings such as Afghanistan. Owing to the lack of data and recent political changes, I could not examine educational outcomes beyond the matriculation phase. Nonetheless, this\\n\\n> 26I contacted various officials and entities to access student-level or concentrations-level post-matriculation outcomes with no success. The significant changes in Afghanistan since the Fall of 2021 have introduced substantial disruption in research activities.\\n\\n31\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [34]}}, {\"page\": 35, \"chunk\": 35, \"input_text\": \"study contributes to the growing body of literature on the role of affirmative action in reducing gender inequality. This study demonstrates that even in conflict-affected settings and heavily patriarchal societies with limited resources for human capital development, affirmative action might be a potential pathway for reducing gender inequality. However, one must recognize that preferential admission to IHEs will likely produce unintended consequences. For instance, lowering the score threshold required for admission, at least in the short term.\\n\\nThe return of the Taliban to power and its current ban on women’s post-secondary education are reversing the progress made in women’s access to post-secondary education since the beginning of the twenty-first century, especially after the implementation of the gender quota policy. Furthermore, the Taliban’s ban on girls’ education in secondary schools and women’s participation in the Kankor exam, as of August 2023, will have a severe negative impact in the coming years. This disrupts the pipeline of applicants graduating from high school, leading to a lower rate of women eligible for the Kankor exam if the Taliban lifts the existing restrictions. Such interruptions have catastrophic consequences, as they decrease the availability of women in vital professions, especially in the health and education sectors, primarily in rural areas.<sup>27</sup>\\n\\n> 27This is a significant setback for the country’s self-reliance, as it increases dependency on international humanitarian aid. Additionally, if the Taliban pursues a gender-segregated education system but does not limit women’s access to academic programs, this might increase women’s participation in post-secondary education. However, this could also result in additional administrative costs. The public higher education sector has been facing significant funding issues since 2014 (Najam & Johnston, 2023), and this has been intensified by the political shocks of 2021.\\n\\n32\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [35]}}, {\"page\": 36, \"chunk\": 36, \"input_text\": \"# **Acknowledgements**\\n\\nI would like thank Todd Pugatch, Alison Johnston, Paul Thompson, Patrick Emerson, Jennifer AlixGarcia, and Olga N. Shemyakina for their invaluable comments, guidance, and support. I am thankful to John Ashton Loeser, Marcus Holmlund, Daniel Rogger, and Arianna Legovini from the World Bank for their support, insightful feedback, and suggestions. I am also very grateful to DIME Analytics at the World Bank for their incredible support in verifying the computational reproducibility package of this study. I am equally thankful to the workshop/seminar participants at the OSU SPP Brownbag, OSU Dev Econ group, AMIE Workshop in Applied Microeconomics 2022, AEFP Doctoral Students Series 2022, PNW Labor Day Workshop 2022, NWDW 2023, PacDev23, Development Econ Series at University of Georgia, SEA 2023, SEEDS 2023, and the anonymous reviewers for insightful comments. All errors lie solely with the author.\\n\\n# **Funding**\\n\\nThis research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.\\n\\n# **Data availability**\\n\\nThe dataset used in this study comprises the Kankor exam results for all test-takers from 2013–2018, collected from publicly available online sources. The author compiled a six-year individual-level dataset and constructed concentration- and major-level datasets. The author is willing to make the aggregate data and the Stata codes used for analysis available online upon publication. Individual-level de-identified data can be shared on a case-by-case basis. Computational reproducibility verified by DIME Analytics. Please see details of the reproducibility in the online appendix or this LINK.\\n\\n# **Disclosure statement**\\n\\nNone\\n\\n# **Declaration of Generative AI and AI-assisted technologies in the writing process**\\n\\nDuring the preparation of this work, the author used ChatGPT-4 in order to identify grammatical errors in some sections of the draft version of the paper before rigorous proofreading. After using this tool/service, the author reviewed and edited the content as needed and takes full responsibility for the content of the publication.\\n\\n33\", \"datasets\": [{\"mention_name\": {\"text\": \"Kankor exam results\", \"confidence\": 0.5118803381919861, \"start\": 1142, \"end\": 1161, \"clean_text\": \"Kankor exam results\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1142, \"end\": 1161, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1142, \"end\": 1161, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"2013–2018\", \"confidence\": 0.48342734575271606, \"start\": 1187, \"end\": 1196}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1142, \"end\": 1161, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.815382182598114, \"start\": 326, \"end\": 345}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5118803381919861, \"start\": 326, \"end\": 345}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9559482336044312, \"start\": 326, \"end\": 345}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9559482336044312, \"start\": 326, \"end\": 345}}, {\"mention_name\": {\"text\": \"six-year individual-level dataset\", \"confidence\": 0.38580232858657837, \"start\": 1270, \"end\": 1303, \"clean_text\": \"six-year individual-level dataset\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1270, \"end\": 1303, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1270, \"end\": 1303, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1270, \"end\": 1303}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1270, \"end\": 1303, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"microdata\", \"confidence\": 0.6474589109420776, \"start\": 454, \"end\": 487}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.38580232858657837, \"start\": 454, \"end\": 487}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5671157836914062, \"start\": 454, \"end\": 487}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5671157836914062, \"start\": 454, \"end\": 487}}, {\"mention_name\": {\"text\": \"None\", \"confidence\": 0.30776119232177734, \"start\": 1724, \"end\": 1728, \"clean_text\": \"None\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1724, \"end\": 1728, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1724, \"end\": 1728, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1724, \"end\": 1728}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1724, \"end\": 1728, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.8280037641525269, \"start\": 28, \"end\": 32}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.30776119232177734, \"start\": 28, \"end\": 32}, \"usage_context\": {\"text\": \"background\", \"confidence\": 0.5652562975883484, \"start\": 28, \"end\": 32}, \"is_used\": {\"text\": \"False\", \"confidence\": 0.5652562975883484, \"start\": 28, \"end\": 32}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099300101182423297/pdf/IDU1617049971af801450e18c91117260363b929.pdf\", \"pages\": [36]}}]}"}}},{"rowIdx":81,"cells":{"doc_id":{"kind":"string","value":"34241163"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34241163\", \"last_modified_date\": \"2024-03-08T00:00:00Z\", \"admreg\": \"Latin America & Caribbean,Latin America & Caribbean\", \"admreg_key\": \"119228,119228\", \"authors\": {\"0\": {\"author\": \"Fietz,Katharina Maria\"}, \"1\": {\"author\": \"Morgandi,Matteo\"}, \"2\": {\"author\": \"Lyrio De Oliveira,Gabriel\"}, \"3\": {\"author\": \"Ferreira Cruz E Superti,Luiz Henrique\"}}, \"count\": \"Brazil\", \"count_key\": \"82649\", \"docna\": {\"0\": {\"docna\": \"Does Local Employment Growth Accelerate\\n Exits from Social Assistance ? Evidence From Brazil’s\\n Conditional Cash Transfer Bolsa Familia\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"Social Protection & Labor LCR (HLCSP),Social Protection & Labor Global (HSPGE)\", \"projn\": \"BR-Social Protection Reforms For\\n Restructuring And Recovery During -- P174836\", \"subsc\": \"Public Administration - Social Protection\", \"trustfund\": \"TF0B0931-R16 RSR BR Promoting Gender Equity,TF0B3106-Understanding Challenges in\\n Access to Labor Market and Social Services,TF0B4068-Brazil - Adaptation of\\n Brazil's parenting program to provide remote sup,TF0B7109-BR - SIEF IE Optimal Design of\\n Digital Early Childhood Development Inte,TF0B7490-R19 Womenâs Economic\\n Inclusion through Labor and Social Assistance\", \"trustfund_key\": \"1806729,1923427,1961148,2097356,2112265\", \"theme\": \"Social Safety Nets,Gender,Human Development and Gender,Economic Policy,Social Protection,Social Development and Protection,Fiscal Policy,Disease Control,Pandemic Response,Labor Market Policy and Programs,Active Labor Market Programs,Fiscal sustainability,Social Insurance and Pensions,Social protection delivery systems\", \"prdln\": \"Advisory Services & Analytics\", \"prdln_key\": \"1416466\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34241163\"}, \"repnb\": \"WPS10676\", \"docdt\": \"2024-01-18T00:00:00Z\", \"datestored\": \"2024-01-18T03:08:01Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"A substantial literature studied the\\n capacity of safety nets to expand automatically during labor\\n market shocks; however, less is known about the dynamics of\\n social assistance when labor market conditions improve, and\\n who may benefit from positive changes, especially in\\n developing countries. This paper studies how rising formal\\n employment at the municipal level affects the likelihood of\\n beneficiary families to exit Bolsa Familia,\\n Brazil’s dynamic means-tested cash transfer. The\\n analysis exploits panel data from Brazil’s vast\\n social registry, matched with seven years of Bolsa Familia\\n payroll information and formal employment records. It finds\\n that the Bolsa Familia program displays significant and\\n heterogeneous dynamism, with beneficiaries with higher\\n levels of education and fewer constraints to labor supply\\n taking fewer years to exit. The analysis then uses\\n fixed-effects estimates, combined with a shift shares\\n instrument, to identify the effects of exogenous changes in\\n the local labor market on exits. The findings show that the\\n increase in local employment leads to a small, statistically\\n significant rise in the probability of exiting from Bolsa\\n Familia, associated with rising formal employment among\\n social assistance recipients. These effects are concentrated\\n in households with spare labor supply and those with medium\\n levels of education.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10676; PEOPLE\", \"display_title\": \"Does Local Employment Growth\\n Accelerate Exits from Social Assistance ? Evidence From\\n Brazil’s Conditional Cash Transfer Bolsa Familia\", \"disclosure_date\": \"2024-01-18T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099508001182418663/text/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.txt\", \"listing_relative_url\": \"/research/2024/01/34241163/local-employment-growth-accelerate-exits-social-assistance-evidence-brazil’s-conditional-cash-transfer-bolsa-familia\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099508001182418663/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c\", \"new_url\": \"2024/01/34241163/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c\", \"ext_pub_date\": \"2024-01-18T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Cash Transfers m1327437 1608,Labor Incomes and Poverty m1503852 2676,Labor Supply m1504055 2751,Social Assistance Design and Evaluation\\n m1582326 4407,Graduation Strategy m1583884 4414\", \"docm_id\": \"IDU-527571dc-f7d1-4c8a-ae59-7c1b8225ec9c\", \"chronical_docm_id\": \"IDU-527571dc-f7d1-4c8a-ae59-7c1b8225ec9c\", \"publishtoextweb_dt\": \"2024-01-18T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10676\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"projectid\": \"P174836\", \"guid\": \"099508001182418663\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"43\", \"author_keys\": {\"0\": {\"author_key\": \"1815794\"}, \"1\": {\"author_key\": \"943637\"}, \"2\": {\"author_key\": \"1987661\"}, \"3\": {\"author_key\": \"1988406\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099508001182418663/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10676\\n\\n# Does Local Employment Growth Accelerate Exits From Social Assistance?\\n\\nEvidence From Brazil’s Conditional Cash Transfer Bolsa Familia _Katharina Fietz Matteo Morgandi Gabriel Lyrio de Oliveira Luiz Henrique Superti_ Social Protection and Jobs Global Practice January 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10676\\n\\n# **Abstract**\\n\\nA large literature highlights the need for low- and middle-income countries to establish adaptive social safety nets responsive to labor market shifts. While much is known about safety nets’ expansion during downturns, less is understood about their response to labor market recoveries. This paper examines how municipal formal employment affects the likelihood of families exiting Bolsa Familia, Brazil’s means-tested cash transfer program. Survival analysis reveals that families exiting the program tend to have better-educated household heads, live in urban areas, and have fewer children. Fixed-effects estimation, combined with a shift-share instrument, shows that local employment growth leads to a small but statistically significant increase in the probability of exiting Bolsa Familia. A 10% growth in local employment raises the exit probability by 0.342 percentage points, increasing the average transition rate from 8.61% to 8.69%. These effects are concentrated in households with spare labor supply.\\n\\nThis paper is a product of the Social Protection and Jobs Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at katharina.fietz@giga-hamburg.de, mmorgandi@worldbank.org, gabriel_ldo@usp.br, and luizhfcs@gmail.com.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# **Does Local Employment Growth Accelerate Exits From Social Assistance?**\\n\\nEvidence From Brazil’s Conditional Cash Transfer Bolsa Familia<sup>∗</sup>\\n\\n**Katharina Fietz**<sup>†</sup> **Matteo Morgandi**<sup>‡</sup> **Gabriel Lyrio de Oliveira**<sup>§</sup> **Luiz Henrique Superti**<sup>¶</sup>\\n\\nOriginally published in the Policy Research Working Paper Series on _January 2024._ This version is updated on _April 2025._ To obtain the originally published version, please email prwp@worldbank.org.\\n\\n**JEL Classification:** H53, I38, J01, J21, J24, O12\\n\\n**Keywords:** Conditional Cash Transfer, Labor Market, Employment, Graduation, Bolsa Familia\\n\\n> ∗We would like to thank Pablo Ariel Acosta, Tiago Falcao Silva, Renate Hartwig, Marcela Iba ̃nez, Krisztina Kis-Katos, Jann Lay, Josefina Posadas, Joana Silva, Jan Stuhler, the Brazilian Ministry of Social Development, and participants of the 28th Annual LACEA-LAMES Meeting, the German and Nordic Development Conference, the BeNA Winter Workshop, the 14th Interdisciplinary Ph.D. Workshop Perspectives on (Un-)Employment, the Development Economics Seminar G ̈ottingen and the GIGA Internal Seminar for their valuable comments and suggestions to this paper. In addition, we thank the Brazilian Ministries of Social Development and Labor for data access. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. The authors declare that they have no relevant or material financial interest related to the research described in this paper. All remaining errors are ours.\\n\\n> †University of G ̈ottingen, GIGA; `katharina.fietz@giga-hamburg.de` (Corresponding author; Neuer Jungfernstieg 21, 20354 Hamburg)\\n\\n> ‡The World Bank; `mmorgandi@worldbank.org`\\n\\n> §University of S ̃ao Paulo; `gabriel.ldo@gmail.com`\\n\\n> ¶University of S ̃ao Paulo; `luizhfcs@gmail.com`\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **1 Introduction**\\n\\nSocial safety nets – especially in the form of conditional cash transfers (CCTs) – are essential for reducing poverty and enhancing the resilience of vulnerable populations. There is growing consensus on the need for adaptive social assistance programs that can dynamically respond to economic shocks in low- and middle-income countries (Bowen _et al._ , 2020). While substantial research has examined pathways into these programs (Bah _et al._ , 2019; Baird _et al._ , 2013; Bowen _et al._ , 2020; Coady and Parker, 2009; Grosh _et al._ , 2022) as well as the positive effects of conditionalities on beneficiaries (Baird _et al._ , 2014; Fiszbein _et al._ , 2009; Glewwe and Kassouf, 2012; Kabeer and Waddington, 2015; Lagarde _et al._ , 2007; Mill ́an _et al._ , 2019; Oconnor, 2024), less attention has been given to the factors that facilitate beneficiaries’ exits. As governments worldwide face growing budget constraints, understanding the conditions that support program graduation is critical for designing effective, dynamic, and sustainable social assistance systems.\\n\\nThis paper seeks to address this knowledge gap concerning program exits. First, it examines the factors that are correlated with exits from Brazil’s CCT program, Bolsa Familia (BF) and finds that families exiting the program tend to have better-educated household heads, live in urban areas, and have fewer children. Second, it explores how improved labor market conditions influence the duration of beneficiaries in BF. Specifically, we analyze the impact of municipal employment changes on the likelihood of families exiting the program within the same locality.<sup>1</sup> Understanding household dynamics, including the individual or municipality-level factors that contribute to better incomes and the capacity of programs to respond to such changes, is critical for informing the sustainable design, operation, and support of CCTs.\\n\\nOur identification strategy relies on combining panel data analysis with an instrumental variable (IV) approach. Gerard _et al._ (2021) show a positive effect of BF beneficiaries on formal employment at the local level in Brazil and attribute it to an economic multiplier effect. While those findings underscore the importance of accounting for local heterogeneity in the labor market we acknowledge the challenge of potential endogenous employment growth and reverse causality. We thus follow Bartik (1991) and instrument local employment growth with the national employment growth of over 600 industries and the share of the respective industry at the local level. The exogeneity of the instrument is given by the fact that national employ-\\n\\n> 1Exits prompted by an increase in local employment may result from voluntary program exits or from official cross-checks revealing that families’ per capita incomes exceed the eligibility threshold. In the case of voluntary exits, families can take advantage of the _retorno garantido_ (guaranteed return), allowing them to re-enter the program within two years post-exit without being subject to a waiting list, provided they meet eligibility criteria.\\n\\n1\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"ment growth across over 600 industries is plausibly uncorrelated with local economic shocks or unobserved local factors, affecting local employment growth only through the industry composition at the local level.<sup>2</sup> Our data – administrative data on BF beneficiaries as well as matched employer-employee data on formal employment – enables us to follow movements in and out of BF as well as formal dependent employment without interruption from 2012 up to 2019.<sup>3</sup> Our primary results suggest that local employment growth leads to a small – yet statistically significant – increase in the probability of beneficiary families leaving the BF program. A 10% growth in local employment raises the exit probability by 0.342 percentage points, increasing the average transition rate of program exit from 8.61% to 8.69%. In addition, we find that these exogenous “rising tides” at the local level cause an increase in the probability of having at least one adult family member in a formal dependent job. Several robustness checks, such as replicating the analysis at the municipality level and checking for administrative capacities of the local authorities confirm our results.\\n\\nThose findings indicate that the poor are responsive to local improvement in the labor market and confirm that a program can be operated in ways that are responsive to such income changes.<sup>4</sup> Nevertheless, the small size of the effects also highlights the relevance of policies that strengthen the capabilities and support of the poor to seize labor market opportunities.<sup>5</sup> Heterogeneity analysis of our results, in fact, shows that exits are concentrated among those households that are less likely to face labor force participation constraints, with fewer young children and more spare adult labor. This is consistent with the descriptive findings of another recent study on the family-level correlates of the BF benefit duration (Morgandi _et al._ , 2023).\\n\\nThe question is highly relevant for governments in many developing countries. After deploying CCTs on a large scale, often as a response to economic downturns, policymakers are now facing new challenges connected with their institutionalization. For programs to include new cohorts of young eligible beneficiaries, policymakers need to establish exit criteria that do not reverse the program’s gains. The prevailing practice in high-income countries has been to link\\n\\n> 2Using the leave-one-out estimator we calculate the national growth rate by excluding the local growth rate of the municipality we are calculating the instrument for. Thus, we rule out that the national growth rate is driven by the local growth rate. Given the large number of municipalities the scenario is unlikely, however, the strategy increases the credibility of our instrument.\\n\\n> 3Our analysis stops in 2019 since re-certification requirements were suspended in 2020-2021 due to the COVID19 pandemic.\\n\\n> 4Exiting BF voluntarily enables the option of a guaranteed return to the program. If people exit through cross-checks they need to reapply to the program and might be subject to a waiting list. When individuals graduate from the program obtaining formal dependent employment they are eligible to other in-work benefits such as the family benefit _Salario Familiar_ .\\n\\n> 5Mourelo and Escudero (2017) show that active labor market policies integrated in cash transfer can increase the probability of having a formal job.\\n\\n2\", \"datasets\": [{\"mention_name\": {\"text\": \"administrative data on BF beneficiaries\", \"confidence\": 0.7076869606971741, \"start\": 236, \"end\": 275, \"clean_text\": \"administrative data on BF beneficiaries\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 236, \"end\": 275, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 236, \"end\": 275, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 236, \"end\": 275}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 236, \"end\": 275, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9950697422027588, \"start\": 236, \"end\": 275}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.7076869606971741, \"start\": 236, \"end\": 275}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5008956789970398, \"start\": 236, \"end\": 275}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5008956789970398, \"start\": 236, \"end\": 275}}, {\"mention_name\": {\"text\": \"matched employer-employee data on formal employment\", \"confidence\": 0.3703249990940094, \"start\": 287, \"end\": 338, \"clean_text\": \"matched employer-employee data on formal employment\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 287, \"end\": 338, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 287, \"end\": 338, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 287, \"end\": 338}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 287, \"end\": 338, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9950697422027588, \"start\": 287, \"end\": 338}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3703249990940094, \"start\": 287, \"end\": 338}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5008956789970398, \"start\": 287, \"end\": 338}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5008956789970398, \"start\": 287, \"end\": 338}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"program exit to improvements in household income or living conditions. However, little has been researched on the income and labor market dynamics of CCT beneficiaries in developing countries.\\n\\nBeing among the few CCTs in developing countries that frequently reassess eligibility via household income tests, the BF program is ideal for exploring this question.<sup>6</sup> A rich set of administrative data on formal employment at the municipal level allows us to capture the relationship between “rising tides” in the local labor market and program exits, while individual data on job matches allows us to identify one of the likely channels for exits, the increase in formal employment of its beneficiaries.\\n\\nIn addition, the program itself has several optimal features to explore our research question. First, while much of the literature implicitly assumes that duration in CCTs programs is either exogenously determined by time limitation rules, or, absent these, largely indefinite, with families lasting in CCTs until children reach adulthood<sup>7</sup> , the basic benefit of BF allows families and individuals to stay in the program as long as they meet the eligibility criteria of a per capita income below the eligibility threshold. In other words, not all families have to exit when children become adults. Second, the eligibility determination in BF depends on a centralized means test, instead of the more common proxy means test targeting. Families at the time of application self-report their income from all sources: this information then undergoes cross-verification with several other registries. The same verification process is repeated at each re-certification episode. As such, the program is sensitive to income fluctuations at entry and exit (Gazola Hellmann, 2015). These design features make BF an ideal case study on the factors that affect exit from CCTs, a question that, to the authors’ knowledge, has not been studied before for CCTs or for programs in developing countries.\\n\\nOur paper contributes to different strands of the literature. First, we add to the literature by examining the reasons that explain exits from social assistance. So far, the evidence originates mainly on programs in high-income countries.<sup>8</sup> Ayala and Rodr ́ıguez (2007) and B ̈ackman\\n\\n> 6Beneficiaries are obliged to report any change in their income situation as soon as it happens. In addition, they need to update / re-verify their provided information every two years independent of changes.\\n\\n> 7For instance a series of new studies estimates the long-term impacts of CCT programs on former beneficiaries who reached adulthood. Government efforts to deploy complementary interventions for families to increase chances of exit, or ‘graduation’, are also underpinned by similar assumptions about families’ long-term stagnation in the programs (De Oliveira and Chagas, 2020; Fietz _et al._ , 2021).\\n\\n> 8There are several explanations for this gap in the research. First, in a large share of CCT programs in developing countries, re-certification processes remain ad-hoc; second, programs adopting proxy means testing are more sensitive to changes in scores when household structural conditions change and less when incomes change, so fluctuations are few. Third, while beneficiaries in Brazil are well balanced between urban and rural areas, in many countries CCTs are mostly focusing on rural poverty, where formal income opportunities are rare. For all these reasons, several countries resort to making their programs time-bound or tied to beneficiary death or\\n\\n3\", \"datasets\": [{\"mention_name\": {\"text\": \"administrative data\", \"confidence\": 0.6924088597297668, \"start\": 387, \"end\": 406, \"clean_text\": \"administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 387, \"end\": 406, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 387, \"end\": 406, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 387, \"end\": 406}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 387, \"end\": 406, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9850483536720276, \"start\": 387, \"end\": 406}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6924088597297668, \"start\": 387, \"end\": 406}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5011414289474487, \"start\": 387, \"end\": 406}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5011414289474487, \"start\": 387, \"end\": 406}}, {\"mention_name\": {\"text\": \"individual data on job matches\", \"confidence\": 0.6999797224998474, \"start\": 563, \"end\": 593, \"clean_text\": \"individual data on job matches\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 563, \"end\": 593, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 563, \"end\": 593, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 563, \"end\": 593}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 563, \"end\": 593, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9850483536720276, \"start\": 563, \"end\": 593}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6999797224998474, \"start\": 563, \"end\": 593}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5011414289474487, \"start\": 563, \"end\": 593}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5011414289474487, \"start\": 563, \"end\": 593}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"and Bergmark (2011), for example, examined the impact of individual characteristics on the probability of exit in Spain and Sweden, finding that gender and prior work experience influenced exit probabilities. A recent policy note, on which the current research builds, looks at the duration of stay of families in BF more descriptively and identifies the level of education of adult members, labor supply constraints and prior work experience as significant predictors (Morgandi _et al._ , 2023).\\n\\nSecond, this paper also contributes indirectly to the vast literature on cash transfers and labor supply. Economic theory suggests that cash transfer may potentially reduce labor supply, especially in the formal sector, if the transfer generates a high marginal tax rate on labor earnings due to benefit withdrawal rules. In the case of BF, Fietz _et al._ (2021) show that from a theoretical _de jure_ perspective, the program has an incentive-compatible design. The authors find that BF has one of the lowest personal and marginal tax rates on earned labor incomes when compared with other programs in the OECD.<sup>9</sup> Moreover, the BF program contains safeguards that allow beneficiaries to remain for additional time in the program when their income rises above the threshold at entry. Yet, it is unclear if families are entirely aware of these rules and act based on such incentives (Fietz _et al._ , 2021). For instance, Levy and Cruces (2021) argue that beneficiaries might prefer informal employment out of concern of not being able to re-enter the program if they leave.\\n\\nThe empirical evidence on work incentives and cash transfer programs generally indicates limited or no disincentive effects on labor supply.<sup>10</sup> For Brazil specifically, the evidence suggests that BF did not discourage (formal) labor market participation, although some studies find increases in informal hours worked.<sup>11</sup> Closely related to this is the research looking at labor children reaching adulthood.\\n\\n> 9Becker (1965) shows that individuals should increase their demand for leisure as non-labor income rises. Following this reasoning cash transfer recipients would be expected to reduce their labor supply as soon as they receive the monetary transfer. Additionally, beneficiaries might be influenced through various other channels – such as the health productivity effect, the self-employed liquidity effect, the insurance effect, and the investment in labor search effect (Baird _et al._ , 2018). Those effects, which may increase or decrease the labor supply of cash transfer beneficiaries, challenge the simplistic theory of the lazy welfare recipient.\\n\\n> 10In a systematic review, Kabeer and Waddington (2015) finds negative effects on labor force participation in only one program (in Pakistan), though some studies document reduced formality (e.g., Garganta and Gasparini (2015)).\\n\\n> 11Using the Brazilian household survey and exploiting a change in the program’s eligibility line, Barbosa and Corseuil (2014) do not find evidence of BF distorting occupational choices toward informal jobs. Fruttero _et al._ (2020) use Brazilian data and an unannounced change in the program’s eligibility rules. Restricting their sample to families that had just updated their information before the policy change was announced – hence, those families were unable to update information untruthfully due to the policy change – the authors find that BF beneficiaries are “4.7 percentage points more likely to be found at least once in a formal job in 2010 or 2011 and 3.3 percentage points more likely to be found in a formal job throughout these two years.” The results are especially strong for younger cohorts (between 25 and 35 years old and 35 and 45 years old). De Brauw _et al._ (2015) find no overall impact on labor force participation of BF beneficiaries using longitudinal household-level panel data collected in 2005 and 2009. The authors, however, do find an increase in the intensive margin of informality (through\\n\\n4\", \"datasets\": [{\"mention_name\": {\"text\": \"Brazilian household survey\", \"confidence\": 0.8116316795349121, \"start\": 2913, \"end\": 2939, \"clean_text\": \"Brazilian household survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2913, \"end\": 2939, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2913, \"end\": 2939, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2913, \"end\": 2939}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2913, \"end\": 2939, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9874080419540405, \"start\": 452, \"end\": 478}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8116316795349121, \"start\": 452, \"end\": 478}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8865142464637756, \"start\": 452, \"end\": 478}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8865142464637756, \"start\": 452, \"end\": 478}}, {\"mention_name\": {\"text\": \"Brazilian data\", \"confidence\": 0.3679012656211853, \"start\": 3137, \"end\": 3151, \"clean_text\": \"Brazilian data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3137, \"end\": 3151, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3137, \"end\": 3151, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3137, \"end\": 3151}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3137, \"end\": 3151, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.2334970086812973, \"start\": 676, \"end\": 690}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3679012656211853, \"start\": 676, \"end\": 690}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6102450489997864, \"start\": 676, \"end\": 690}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6102450489997864, \"start\": 676, \"end\": 690}}, {\"mention_name\": {\"text\": \"longitudinal household-level panel data\", \"confidence\": 0.7325826287269592, \"start\": 3871, \"end\": 3910, \"clean_text\": \"longitudinal household-level panel data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3871, \"end\": 3910, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3871, \"end\": 3910, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3871, \"end\": 3910}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3871, \"end\": 3910, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.39132869243621826, \"start\": 639, \"end\": 678}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.7325826287269592, \"start\": 639, \"end\": 678}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.756013035774231, \"start\": 639, \"end\": 678}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.756013035774231, \"start\": 639, \"end\": 678}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"market constraints and childcare duties. Blundell _et al._ (2018), for example, show that childcare duties play an important role in labor supply decisions – especially for mothers.<sup>12</sup> Our research findings add to the existing literature by examining program exits and their relationship with local employment growth. We find that families exposed to “rising tides” in the labor market are not only more likely to exit BF but also more likely to have members transition into formal employment. This suggests that BF’s design does not create strong disincentives against formal work participation even after being exposed to the program for some time.\\n\\nThe rest of the paper is structured as follows. Section 2 describes the institutional design of BF, its interaction with the rest of the tax benefit system and the data used. Section 3 presents some descriptive statistics followed by Section 4, where the methods used are explained. Section 5 shows results and robustness checks before we conclude in Section 6.\\n\\n# **2 Institutional Design and Data**\\n\\n## **2.1 Bolsa Familia Program**\\n\\nBrazil was among the first countries in the world to embrace CCTs. Launched in 2003, BF aimed to alleviate both temporary and chronic poverty and fostered human capital development for children. When BF was first introduced, it also sought to consolidate several sub-national programs, such as _Bolsa Escola_ and a gas subsidy (Lindert _et al._ , 2007; Bastagli, 2011; Fietz _et al._ , 2021). This section describes the rules and structure of BF as they applied in 2019. Since then, the program has undergone significant changes, including its replacement by _Aux ́ılio Brasil_ in 2021 and further modifications in 2023.\\n\\nTo receive assistance under the 2019 framework, families had to be registered in the single registry Cadastro Unico<sup> ́</sup> (CadUnico).<sup>13</sup> BF consisted of two benefits: (i) the basic benefit and (ii) the variable benefit. As of 2019, the final year of our panel dataset, the monthly basic benefit was paid to families with a monthly per capita income of less than R$ 89 (USD 21.4). Families whose per capita income remained below R$ 89 (USD 21.4) also received a supplement so that the household would reach at least a monthly per capita income of R$ 89 (USD 21.4). The variable benefit was granted to individuals with a monthly per capita income below the\\n\\n> additional hours of informal work).\\n\\n> 12Other references are Aguilar-Gomez _et al._ (2019) for Mexico, Berniell _et al._ (2021) for Chile, and Heath (2017) for theoretical evidence from Ghana.\\n\\n> 13Individuals with a per capita income below 0.5 of the minimum wage (MW) were eligible to register in CadUnico. For more information about the registry, see Section 2.2.\\n\\n5\", \"datasets\": [{\"mention_name\": {\"text\": \"panel dataset\", \"confidence\": 0.5225768685340881, \"start\": 1991, \"end\": 2004, \"clean_text\": \"panel dataset\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1991, \"end\": 2004, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1991, \"end\": 2004, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1991, \"end\": 2004}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1991, \"end\": 2004, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.5024022459983826, \"start\": 447, \"end\": 460}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5225768685340881, \"start\": 447, \"end\": 460}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.44038501381874084, \"start\": 447, \"end\": 460}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.44038501381874084, \"start\": 447, \"end\": 460}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"poverty line (in 2019, R$ 178, USD 42.7), and the amount depended on the number of children or pregnant women in the family. In 2019, the program paid R$ 41 (USD 9.8) per month per child or pregnant woman, with a maximum of five variable benefits per family. The variable benefit was linked to education and health conditions (see Table 1 for an overview of the rules).\\n\\nTable 1: Bolsa Familia rules as of 2019\\n\\n|Extreme poor families (monthly p.c. in-<br>come below 89 R$/21.4 USD)|Poor families (monthly p.c. income below<br>178 R$/42.7 USD)|\\n|---|---|\\n|• Basic beneft (89 R$/21.4 USD per<br>family)<br>• Variable beneft (41 R$/9.8 USD<br>per child or pregnant women, max.<br>5 variable benefts per family)|• Variable beneft (41 R$/9.8 USD<br>per child or pregnant women, max.<br>5 variable benefts per family)|\\n|• Top-up beneft after all other ben-<br>efts (top-up to reach 89 R$/21.4<br>USD p.c. income)||\\n\\n**Notes:** The table describes the rules of Bolsa Familia as of 2019.\\n\\nBF was not – and still is not – a constitutional right but a budgetary program, meaning that not everyone who was eligible for the program during our study period necessarily received it; the program had a waiting list. Each municipality registered vulnerable families and was allocated a budget based on poverty estimates from the Brazilian Institute of Statistics (Instituto Brasileiro de Geografia e Estat ́ıstica – IBGE). The process of including new beneficiaries from the waiting list as spots became available followed a centralized procedure led by the Ministry of Social Development (MDS).\\n\\nThere were several reasons why families exited the program under the 2019 regulations and many of those still apply under the new program rules. The first related to changing household demographics: once children reached adulthood, families could only receive the basic benefit, and if their income exceeded the extreme poverty threshold, they were excluded from the program altogether. Second, if families failed to meet health and education conditions without sufficient justification, they were removed from the program after being offered opportunities to comply. Third, households were required to withdraw the cash benefit within a certain period; failure to do so resulted in exclusion from the program. Finally, household income had to remain below the eligibility threshold, though with one important exception: the _Regra de Permanencia_ (permanence rule). This rule allowed beneficiaries to remain in the program for an additional two years and maintain their income above the eligibility line (up to certain limits), provided 6\", \"datasets\": [{\"mention_name\": {\"text\": \"poverty estimates from the Brazilian Institute of Statistics\", \"confidence\": 0.3539894223213196, \"start\": 1287, \"end\": 1347, \"clean_text\": \"poverty estimates from the Brazilian Institute of Statistics\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1287, \"end\": 1347, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1287, \"end\": 1347, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1287, \"end\": 1347}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1287, \"end\": 1347, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9436942934989929, \"start\": 520, \"end\": 580}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3539894223213196, \"start\": 520, \"end\": 580}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7822064757347107, \"start\": 520, \"end\": 580}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7822064757347107, \"start\": 520, \"end\": 580}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"that they voluntarily reported their new income at a local social assistance office.\\n\\nAdministrative data showed that under the 2019 system, the vast majority of exits from BF occurred because families’ per capita income rose above the eligibility threshold. The so-called _Averigua ̧c ̃ao e Revis ̃ao Cadastral_ – inquiry and registration review – of existing beneficiaries cross-checked administrative records (such as those on formal employment) against beneficiaries’ self-reported data in CadUnico. This process was carried out centrally in Bras ́ılia and operated independently of municipalities’ economic conditions (Ministerio da Cidadania, 2013). BF beneficiaries were aware of the implications of taking up formal employment for eligibility, as it often led to either program exit or a transition into the permanence rule. Once family income exceeded the threshold, households were removed from the program. While some families left voluntarily, exclusion through the inquiry and registration review could also be considered voluntary since people aimed for higher income despite being aware of the consequences – namely, exclusion from the program (Ministerio da Cidadania, 2008).\\n\\nBF’s design and its interaction with the broader tax-benefit system created an incentivecompatible framework that should not have discouraged individuals from pursuing formal employment. The program had low marginal tax rates on acquired labor income compared to most social assistance programs (Fietz _et al._ , 2021). This was because the value of the BF benefit remained relatively low, fluctuating between a fourth and a fifth of the MW during the observation period. Additionally, Brazil extended automatic benefits to low-wage workers with children ( _Sal ́ario Fam ́ılia_ ), which partially offset reductions in BF payments and social insurance contributions. Furthermore, personal income tax exemptions applied to workers at entry wage levels. Lastly, BF’s permanence rule, _Regra de Permanencia_ , acted as a temporary income disregard. Figure 1 visually illustrates the design of the BF program as it operated in 2019, including the implications of its permanence rule for a family in which an individual transitioned into fulltime formal employment. Thus, even though eligibility was primarily determined by a family income test, nearly all family types experienced an increase in disposable income as wages rose, with minimal discontinuities. Given these factors, we hypothesize that beneficiaries might have favored formal employment over continued program participation if such an opportunity had arisen.\\n\\nSince 2019, the BF program has undergone multiple changes. In 2021, it was replaced by _Aux ́ılio Brasil_ , which introduced modifications to benefit structures and eligibility rules. In 2023, BF was reinstated with further adjustments, including an increase in benefit levels. While the 7\", \"datasets\": [{\"mention_name\": {\"text\": \"Administrative data\", \"confidence\": 0.40285733342170715, \"start\": 86, \"end\": 105, \"clean_text\": \"Administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 86, \"end\": 105, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 86, \"end\": 105, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 86, \"end\": 105}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 86, \"end\": 105, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9970033764839172, \"start\": 86, \"end\": 105}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.40285733342170715, \"start\": 86, \"end\": 105}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6321868300437927, \"start\": 86, \"end\": 105}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6321868300437927, \"start\": 86, \"end\": 105}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"<!-- Start of picture text --> Two adults, two children (between 0 and 15 years old)<br>700 Threshold Cadastro<br>Unico and RP<br>600<br>500 Extreme poverty line (R$ 89 1mMw<br>per capita). Poverty_ line (R$ 178<br>400 per capita)<br>wn<br>c<br>a300 =<br>€<br>Z8 200 ssaeeassteaas9006400555323ae,<br>x aA eens i<br>= 100 - Oo<br>ra3 - —_——— prerETETTTTTTTTTTTTTTTTTTTTtr eS5$}5OOO55OOOO6O0O6004 PRAAEAAAEARERAERAE! SSOSOSOOESOSSSOONOSSOSESSSOSSSSSSOSSOOSESESEDESOSSSSSSSSSESE66O646O646450554400OOO$OO6O0O65Ob55Ob50000O006606066606060660004 KRRRRRARRAAERARAR RASA RSA R RRS ASKeReKeeKeeeereeeeeeSEES EE EEEEEEEESS es<br>(100)<br>(200)<br>(300)<br>COTSSRSSSRRSSLSSSPSRARSAKRSRSRARCSSSSAREARFARKRASSSSARHEA AMT OHONMHADHANTYDHSOSOHAMHEDRONTOSHMAHNMHEBONTERAAMHOEBROATH<br>SSRR88R RB FFRRARSSRRRSERSHKSF S SSHHnNRtRSSAARA A RAansPseRgRRSSCRRRSS<RARRAAAAASASATSAARAASHER SEARSSSRRR R SSSRAASARARAAATGA<br>Declared labor income in RS<br>34464 RP Group 2Variable Benefit Mmm BF Variable Benefit lam Salario Familia<br>lm BF Basic Benefit \\\"rz RP Group 1 Basic Benefit Benefit to overcome extreme poverty<br>lam Social Security Contributions Net Income per capita (no RP) ===» Net income per capita (RP Group 1)<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"**Folha de Pagamentos do Bolsa Familia (BF payment records)** The payroll dataset of BF contains information on the monthly payments received by every family. It does not allow for distinguishing between the different benefits but gives the total sum of transfers received. In addition, datasets on the exit reasons and people in the benefit graduation/transition rules ( _Regra de Permanencia_ ) are available from 2016 onwards.\\n\\n**Rela ̧c ̃ao Anual de Informa ̧c ̃oes Sociais (RAIS)** RAIS is an administrative dataset that contains information on formal dependent employment in Brazil. Every company needs to fill in the information on its employees’ occupation, monthly compensation, contract type, date of entering the company, date of exiting the company, and dismissal reason. In addition, information on the firm, including the economic sector, the location, and the size of the establishment (defined by the number of employees), is available.\\n\\n# **2.3 Sample Construction and Key Variables**\\n\\nWe construct a panel to longitudinally track all individuals aged 18 to 59 who were newly registered in CadUnico and either belonged to a BF recipient family as of December 2012 or joined an already recipient BF family at some point during the observation period. BF participation is determined based on payment records, which are registered at the household head level. To identify all family recipients, we expand the payment information using the family identification number, ensuring coverage of all household members. Individual records in CadUnico are then matched to RAIS using the anonymized _Programa Integra ̧c ̃ao Social_ (PIS) number. The initial set of individuals is followed monthly until 2019. We then further collapse the data to a semester basis – enabling us to track individuals over 14 semesters and examine their transitions into and out of the formal labor market – and at the family level – to examine family-level movements in and out of the BF program. Figure A.1 in the Annex provides an overview of the data merging process.\\n\\n**Exiting Bolsa Familia** Exiting BF is defined by using the program payment records ( _folha de pagamento_ ). We define exiting BF as not being in the program for at least four consecutive months. This threshold has been chosen since beneficiaries who do not comply with program rules might get a grace period of three months in which the benefit is suspended, though people are not expelled from the program immediately. Additionally, we consider beneficiary families that enter the permanence rule as if they have left. The reasoning is that we want to identify the impact of employment changes at the municipality level on exit probabilities. The permanence\\n\\n9\", \"datasets\": [{\"mention_name\": {\"text\": \"Folha de Pagamentos do Bolsa Familia\", \"confidence\": 0.7448747158050537, \"start\": 2, \"end\": 38, \"clean_text\": \"Folha de Pagamentos do Bolsa Familia\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.3713306486606598, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2, \"end\": 38, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2, \"end\": 38}, \"geography\": {\"text\": \"Brazil\", \"confidence\": 0.35640624165534973, \"start\": 581, \"end\": 587, \"clean_text\": \"Brazil\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.8287556171417236, \"start\": 2, \"end\": 38}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7448747158050537, \"start\": 2, \"end\": 38}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5049724578857422, \"start\": 2, \"end\": 38}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5049724578857422, \"start\": 2, \"end\": 38}}, {\"mention_name\": {\"text\": \"payroll dataset of BF\", \"confidence\": 0.36855292320251465, \"start\": 66, \"end\": 87, \"clean_text\": \"payroll dataset of BF\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 66, \"end\": 87, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 66, \"end\": 87, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 66, \"end\": 87}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 66, \"end\": 87, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.8287556171417236, \"start\": 66, \"end\": 87}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.36855292320251465, \"start\": 66, \"end\": 87}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5049724578857422, \"start\": 66, \"end\": 87}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5049724578857422, \"start\": 66, \"end\": 87}}, {\"mention_name\": {\"text\": \"RAIS\", \"confidence\": 0.4524819850921631, \"start\": 487, \"end\": 491, \"clean_text\": \"RAIS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 487, \"end\": 491, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 487, \"end\": 491, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 487, \"end\": 491}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 487, \"end\": 491, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9981682300567627, \"start\": 487, \"end\": 491}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4524819850921631, \"start\": 487, \"end\": 491}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5061612725257874, \"start\": 487, \"end\": 491}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5061612725257874, \"start\": 487, \"end\": 491}}, {\"mention_name\": {\"text\": \"payment records\", \"confidence\": 0.43831759691238403, \"start\": 1307, \"end\": 1322, \"clean_text\": \"payment records\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1307, \"end\": 1322, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1307, \"end\": 1322, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1307, \"end\": 1322}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1307, \"end\": 1322, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.0, \"start\": 520, \"end\": 535}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.43831759691238403, \"start\": 520, \"end\": 535}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8753048181533813, \"start\": 520, \"end\": 535}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8753048181533813, \"start\": 520, \"end\": 535}}, {\"mention_name\": {\"text\": \"program payment records\", \"confidence\": 0.31232017278671265, \"start\": 2119, \"end\": 2142, \"clean_text\": \"program payment records\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.30426937341690063, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2119, \"end\": 2142, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2119, \"end\": 2142}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2119, \"end\": 2142, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.0, \"start\": 494, \"end\": 517}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.31232017278671265, \"start\": 494, \"end\": 517}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9651430249214172, \"start\": 494, \"end\": 517}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9651430249214172, \"start\": 494, \"end\": 517}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"rule, however, is a mechanism to delay, but not to prevent such exits. We only consider the first exit of a family since we do not want to estimate any effects on re-entry.<sup>14</sup> If the head leaves the social registry CadUnico, but a different person from the same family receives the benefits and thus _converts_ to being the head, we classify this family as still _in BF_ .\\n\\n**Formal labor market participation** We identify formal labor market participation by merging information from CadUnico and RAIS. We classify families as having at least one member in formal dependent employment if any adult family member can be found in RAIS in a given semester.\\n\\n**Municipality-level employment growth** We calculate employment growth at the municipality level using RAIS data. RAIS reports the number of job links, which slightly differs from the number of workers since a small share of individuals holds more than one formal job. To calculate the current stock of employment, we use job link data for the private sector, additionally excluding public firms within it, as hiring policies in the public sector are expected to follow different mechanisms. Employment growth is measured as the difference between the current period’s number of private-sector employment links and the previous period’s number of private-sector employment links, divided by the number of private-sector employment links in the previous period.\\n\\n# **3 Descriptive Statistics**\\n\\nThis section presents descriptive statistics of our sample. We start by laying out some characteristics of the individual data before describing some of the exit dynamics using the family data. In December 2012, the average age of individuals was 35. Individuals between 25 and 34 years old are the prevalent age group (32%), followed by individuals between 35 and 44 years old. In addition, 66% of individuals were female and 34% male, consistent with the targeting approach of the program (which privileges female household heads). The mean family size in 2012 was 3.6, and most families had one or two children. Most individuals are Afro descendent and have a middle school education (Annex Table A.1).\\n\\nWhen registering in CadUnico, all adult members are also asked to report basic information regarding their employment status. Out of all individuals, 43% reported that they worked within the last 12 months. Most of them worked as self-employed, followed by formal dependent\\n\\n> 14BF has a waiting list. Hence, not everyone eligible will get the benefits. These circumstances make estimation on entry difficult.\\n\\n10\", \"datasets\": [{\"mention_name\": {\"text\": \"RAIS data\", \"confidence\": 0.3659762442111969, \"start\": 771, \"end\": 780, \"clean_text\": \"RAIS data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 771, \"end\": 780, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 771, \"end\": 780, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 771, \"end\": 780}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 771, \"end\": 780, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8650704622268677, \"start\": 771, \"end\": 780}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3659762442111969, \"start\": 771, \"end\": 780}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7737526893615723, \"start\": 771, \"end\": 780}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7737526893615723, \"start\": 771, \"end\": 780}}, {\"mention_name\": {\"text\": \"job link data\", \"confidence\": 0.45288869738578796, \"start\": 990, \"end\": 1003, \"clean_text\": \"job link data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 990, \"end\": 1003, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 990, \"end\": 1003, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 990, \"end\": 1003}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 990, \"end\": 1003, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.5357794165611267, \"start\": 990, \"end\": 1003}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.45288869738578796, \"start\": 990, \"end\": 1003}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6595078706741333, \"start\": 990, \"end\": 1003}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6595078706741333, \"start\": 990, \"end\": 1003}}, {\"mention_name\": {\"text\": \"family data\", \"confidence\": 0.6069057583808899, \"start\": 1643, \"end\": 1654, \"clean_text\": \"family data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1643, \"end\": 1654, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1643, \"end\": 1654, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1643, \"end\": 1654}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1643, \"end\": 1654, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6956656575202942, \"start\": 867, \"end\": 878}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6069057583808899, \"start\": 867, \"end\": 878}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6433581709861755, \"start\": 867, \"end\": 878}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6433581709861755, \"start\": 867, \"end\": 878}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"<!-- Start of picture text --> =<br>a<br>e<br>+<br>“<br>RO>\\\\ OYVv weN mk”Vv DY\\\\ OOOOYVv >N OYVv AYXN NDAYVv .@LS DOE4”Vv iN<br>ee FFE FCS SF YF SF SF FH<br>EEE Probability of staying MEE Probability of leaving<br><!-- End of picture text --> <!-- Start of picture text --> »: .<br>Exit. rate<br>40<br>10°S<br>30<br>ii 20<br>20°S 10<br>a °<br>><br>-\\\"<br>7O°W 60°W 50°wW 40°W 30°W<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"Table 2: Family characteristics\\n\\n||All families|\\n|---|---|\\n|Number of Families in BF|6,582,657|\\n|Number of Families that left during the observation period|4,602,881|\\n|Number of families that came back to the program|630,549|\\n|Number of families that moved to diferent municipalities|457,079|\\n|Number of families that moved to a diferent municipality prior to exit|80,629|\\n|Average number of years in the program|3.8|\\n|% of families having one member in formal dependent employment at entry|28.02|\\n|% of families having one member in formal dependent employment at exit|35.48|\\n|% of families having head in formal dependent employment at entry|15.12|\\n|% of families having head in formal dependent employment at exit|21.98|\\n\\n**Notes:** The table presents summary statistics for families registered in CadUnico in December 2012 for the first time and being in BF at the beginning of the first semester of 2013. Information on RAIS are obtained through a merge of CadUnico data, BF payroll information and administrative matched employer-employee data (RAIS).\\n\\ncharacteristics of adults, as reported at the time of registration, and their probabilities to exit BF. This serves as useful descriptive evidence in support of the identification strategy for the effect of local employment changes on the probability of exit, which will be described in the next section. Following prior literature and Morgandi _et al._ (2023), we employ the Kaplan–Meier (KM) estimator and Cox regressions.<sup>15</sup> Table 3 shows that exit probabilities are correlated with some of the beneficiaries’ observable characteristics at entry.<sup>16</sup> The speed and probability of exit from the program increase with age and education levels of the adults, particularly those with tertiary diplomas. Exit is also higher in urban areas. Families with 1 or 2 children have a higher probability of leaving the program while families having more than 3 children are less likely to leave compared to families without children.\\n\\n# **4 Empirical Strategy**\\n\\n## **4.1 The Basic Model**\\n\\nTo identify the effect of local employment growth on our two outcomes of interest – the probability of having at least one family member in formal dependent employment and on families’\\n\\n> 15The KM estimator is a non-parametric statistic, which is used to estimate the survival function (in our case time in BF). The estimator can be written as _S_<sup>�</sup> ( _t_ ) = _i_ :� _ti≤t_ �1 _− n_<sup>_di_</sup> _i_ �, where _ti_ represents a time when at least one event (exit) happened, _di_ the number of events (exits), which happened at time _ti_ , and _ni_ individuals known to have survived (have stayed in the program) up to time _ti_ . Cox regression is a method of investigating the effect of one or more variables upon an event takes to happen and can be written as _λ_ 0( _t_ ) = _λ_ 0( _t_ ) _e_<sup>_Xβ_</sup> , where _λ_ 0( _t_ ) is a non-parametric baseline hazard function and _X_ are matrices for the independent variables.\\n\\n> 16A positive coefficient is associated with an increased likelihood of leaving the program. A hazard ratio (shown in column (2)) greater than 1 corresponds to a shorter time in the BF program, whereas a hazard ratio smaller than 1 indicates being for a longer time in the BF program.\\n\\n12\", \"datasets\": [{\"mention_name\": {\"text\": \"RAIS\", \"confidence\": 0.5673883557319641, \"start\": 1051, \"end\": 1055, \"clean_text\": \"RAIS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1051, \"end\": 1055, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1051, \"end\": 1055, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1051, \"end\": 1055}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1051, \"end\": 1055, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.742299497127533, \"start\": 452, \"end\": 456}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5673883557319641, \"start\": 452, \"end\": 456}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.49116262793540955, \"start\": 452, \"end\": 456}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.49116262793540955, \"start\": 452, \"end\": 456}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"Table 3: Cox regression results\\n\\n||(1)<br>Coefcient|(2)<br>Hazard Ratio|\\n|---|---|---|\\n|Education head: Primary or less|0.0817***<br>(0.0020)|1.085***<br>(0.0021)|\\n|Education head: Middle school|0.2490***<br>(0.0019)|1.2828***<br>(0.0025)|\\n|Education head: Secondary school|0.3101***<br>(0.0043)|1.3636***<br>(0.0059)|\\n|Education head: Tertiary|0.8949***<br>(0.0136)|2.4472***<br>(0.0335)|\\n|Head 26 - 40 years old|0.2994***<br>(0.0024)|1.3490***<br>(0.0032)|\\n|Head 41 - 50 years old|0.5134***<br>(0.0025)|1.671***<br>(0.0042)|\\n|Head 51 - 69 years old|0.9329***<br>(0.0027)|2.5419**<br>(0.0070)|\\n|Urban|0.3502***|1.4194***|\\n||(0.0013)|(0.0018)|\\n|1 children|0.0695***<br>(0.0019)|1.0720***<br>(0.0020)|\\n|2 children|0.0216***<br>(0.0019)|1.0219***<br>(0.0020)|\\n|3 children|-0.0547***<br>(0.0021)|0.9466***<br>(0.0020)|\\n|4+ children|-0.2064***<br>(0.0024)|0.8134***<br>(0.0019)|\\n|N|4,456,917|4,456,917|\\n\\n_∗ ∗∗ ∗∗∗ p <_ 0 _._ 10, _p <_ 0 _._ 05, _p <_ 0 _._ 01\\n\\n**Notes:** The table shows Cox regression results. A positive coefficient is associated with an increased likelihood of leaving the program. A negative coefficient indicates a lower chance of leaving the program over time. A hazard ratio (shown in column (2)) greater than 1 corresponds to a shorter time in the BF program, whereas a hazard ratio smaller than 1 indicates being for a longer time in the BF program. Reference groups for education classifications and age groups are _illiterate_ and _18-24 years old_ , respectively.\\n\\npropensity to leave BF – we employ panel data analysis with family-level and time-fixed effects. This approach allows us to control for unobserved time-invariant family characteristics that could be endogenous to our outcome variables. As it likely takes time for a local positive employment shock to translate into positive effects on the incomes of poor households, typically more detached from the labor market, and for the program to detect and react to any such changes, we estimate the effects on the outcomes of interest in the following period defined as one semester<sup>17</sup> :\\n\\n> 17 While our data structure would allow to observe monthly movements, the time that it takes for the program to react is unlikely to be at a monthly level.\\n\\n13\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"where _Yiτt_ represents our outcomes of interest: A dummy variable which takes the value of 1 (i) if family _i_ in municipality _τ_ has a family member in formal dependent employment,<sup>18</sup> and (ii) our main outcome of interest: if a family _i_ in municipality _τ_ left BF.<sup>19</sup> Given the nature of the dependent variable, we use a linear probability model.<sup>20</sup> Employment Growth _τt−_ 1 is defined as the change in the total number of formal private sector jobs in the municipality _τ_ and semester _t −_ 1.<sup>21,22,23</sup> _Xit_ is a vector of time-varying co-variates (age groups of the household head and the number of children living in a household).<sup>24</sup> _αi_ are family fixed effects that also capture in which trimester the family left, _pt_ are time-fixed effects. We use clustered standard errors at the municipality level since our treatment (employment growth) is happening at the municipality level.\\n\\n# **4.2 Endogeneity Problem and Instrumental Variable Approach**\\n\\nTo answer our research question we need to address several methodological challenges and potential sources of endogeneity.\\n\\n**Municipality-level population growth** The first challenge is the identification of exogenous positive changes in local employment opportunities. While we can observe the total formal job growth by municipality per year/semester/month, we cannot control precisely for municipalitylevel population growth since the last census in Brazil was in 2010 (as of 2019). Yet, the population effect appears to be negligible (IBGE’s annual estimates of population growth between 2011 and 2019 averaged 1% (IBGE, 2020)) compared to the role of local economic cycles. Formal employment has been closely following economic cycles, with a variation of about 12% between the lowest and highest levels in the same period. In fact, the Brazilian economy had more formal wage jobs in 2012 than in 2019 (MTE, 2019). In addition, large parts of the employed population in Brazil work informally (IBGE, 2019), so excess population growth, in the absence\\n\\n> 18The dummy dependent variable _having a family member in formal dependent employment_ takes the value of 1 if a family member is in formal dependent employment and 0 if no family member is in formal dependent employment.\\n\\n> 19The dummy dependent variable (family _i_ leaves BF) takes the value of 1 if the family leaves the program – based on the official payroll information of BF containing information on monthly benefit payouts of the benefit –, 0 if it stays, and is missing if the family is not in BF in the time-point.\\n\\n> 20A conditional fixed-effect model could be used as well, but the linear probability model makes an IV approach and the interpretation of the results more straightforward.\\n\\n> 21To obtain this information, we use RAIS data, which reports the number of job links, which slightly differs from the number of workers since a small share has more than one formal job.\\n\\n> 22In the case that families moved between municipalities we consider the employment growth of the origin municipality as common in the migration literature (Jaeger _et al._ , 2018).\\n\\n> 23In order to not have our results driven by extreme outliers, we winsorize employment growth at the 99.5 percentile. By doing so we change less than 1% of the municipalities’ employment growth.\\n\\n> 24The CadUnico data provides only a few variables that vary over time.\\n\\n14\", \"datasets\": [{\"mention_name\": {\"text\": \"official payroll information of BF\", \"confidence\": 0.5053799152374268, \"start\": 2424, \"end\": 2458, \"clean_text\": \"official payroll information of BF\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2424, \"end\": 2458, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2424, \"end\": 2458, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2424, \"end\": 2458}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2424, \"end\": 2458, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.4271093010902405, \"start\": 943, \"end\": 977}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.5053799152374268, \"start\": 943, \"end\": 977}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.48188942670822144, \"start\": 943, \"end\": 977}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.48188942670822144, \"start\": 943, \"end\": 977}}, {\"mention_name\": {\"text\": \"RAIS data\", \"confidence\": 0.7731646299362183, \"start\": 2818, \"end\": 2827, \"clean_text\": \"RAIS data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2818, \"end\": 2827, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2818, \"end\": 2827, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2818, \"end\": 2827}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2818, \"end\": 2827, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.491523414850235, \"start\": 592, \"end\": 601}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7731646299362183, \"start\": 592, \"end\": 601}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6890921592712402, \"start\": 592, \"end\": 601}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6890921592712402, \"start\": 592, \"end\": 601}}, {\"mention_name\": {\"text\": \"CadUnico data\", \"confidence\": 0.8659934401512146, \"start\": 3360, \"end\": 3373, \"clean_text\": \"CadUnico data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3360, \"end\": 3373, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3360, \"end\": 3373, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3360, \"end\": 3373}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3360, \"end\": 3373, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.5192055702209473, \"start\": 332, \"end\": 345}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8659934401512146, \"start\": 332, \"end\": 345}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.45519229769706726, \"start\": 332, \"end\": 345}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.45519229769706726, \"start\": 332, \"end\": 345}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"of formal dependent employment creation, tends to flow into unemployment, self-employment (often informal), or informality, which has been observed in Brazil during the last crises (Firpo and de Pieri, 2018; World Bank, 2020).\\n\\n**Multiplier effect caused by BF coverage** The second methodological challenge relates to potential reverse causality, namely that the changes in coverage of BF cash transfers might affect levels of local employment through a local multipliers effect (Gerard _et al._ , 2021). At the national level, the number of beneficiary families remained more or less stable in our observation period (in December 2012, 13.9 million families were registered in BF compared to 13.2 million families in December 2019); however, we do observe variation at the local level, which might cause biases in our results (Ministerio da Cidadania, 2022).\\n\\n**Shift-share instrument** For both challenges, we propose to use a shift-share instrument after Bartik (1991). In particular, we instrument local employment growth using national employment growth across more than 600 industries, interacted with the local share of each respective industry in the previous period. The instrument is plausibly exogenous, as national employment growth at the industry level is unlikely to be correlated with local economic shocks or unobserved local factors, affecting local employment growth only through the area’s industry composition. The Bartik instrument can be expressed as:\\n\\nwhere _zλτt−_ 1 is the share of private sector employment in industry _λ_ in municipality _τ_ in _t−_ 1 and _gλ_ is the national growth rate of private sector employment in industry _λ_ at time t.<sup>25,26</sup> We compile different variations of the shift-share instrument. We start by using annual growth rates. However, since we have monthly information on exits, we switch to semester growth rates. In an additional step, we exclude the growth rate of the municipality for which the industry is calculated. In other words, if we calculate the instrument for municipality _τ_ in industry _i_ , we exclude employment levels of industry _i_ in municipality _τ_ when computing the national growth rate (leave-one-out estimator, illustrated by _−τ_ in model (2)). This gives additional credibility to the instrument since we can exclude concerns that national growth rates of a specific industry\\n\\n> 25We follow the same procedure as above and winsorize the national employment growth by industry at the 99.5 percentile.\\n\\n> 26We acknowledge that we could also work with initial shares at time t = 0, however, since we are looking at a long time period during which there were changing economic situations, looking at the shares of the previous semester seems more reasonable (Borusyak _et al._ , 2022).\\n\\n15\", \"datasets\": [{\"mention_name\": {\"text\": \"national employment growth across more than 600 industries\", \"confidence\": 0.42936426401138306, \"start\": 1033, \"end\": 1091, \"clean_text\": \"national employment growth across more than 600 industries\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1033, \"end\": 1091, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1033, \"end\": 1091, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1033, \"end\": 1091}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1033, \"end\": 1091, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.5682860016822815, \"start\": 252, \"end\": 310}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.42936426401138306, \"start\": 252, \"end\": 310}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9608144760131836, \"start\": 252, \"end\": 310}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9608144760131836, \"start\": 252, \"end\": 310}}, {\"mention_name\": {\"text\": \"national employment growth\", \"confidence\": 0.7298507690429688, \"start\": 1219, \"end\": 1245, \"clean_text\": \"national employment growth\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1219, \"end\": 1245, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1219, \"end\": 1245, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1219, \"end\": 1245}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1219, \"end\": 1245, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.3589921295642853, \"start\": 438, \"end\": 464}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7298507690429688, \"start\": 438, \"end\": 464}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6456022262573242, \"start\": 438, \"end\": 464}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6456022262573242, \"start\": 438, \"end\": 464}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"are driven by the municipality for which the instrument is calculated. Excluding growth rates of the respective municipality is in line with Goldsmith-Pinkham _et al._ (2020) who state that the leave-one-out method addresses “the finite sample bias that comes from using own-observation information”.\\n\\nOur instrument also addresses any potential biases originating from the unobserved growth in informal labor income. The national labor demand shocks from formal industries cause changes in the local employment of formal industries which, in turn, might cause the BF beneficiaries to exit the program through some mechanisms. The most probable one is that the beneficiary herself gets a formal job. Informal labor income remains an unlikely reason for households to exit the program. Instead, formal jobs are more likely to be a cause for exit from BF – because such earnings are observable, and regularly cross-checked by authorities in Brazil. However, we also note that any other causes of informal sector growth which could correlate with the BF exits are offset by our IV approach.\\n\\nWhile the relevance of the instrument can be observed from the data and will be discussed below, the validity of the instrument cannot be observed from the regression outputs. We are aware that there is a lively discussion in the literature on share (Goldsmith-Pinkham _et al._ , 2020) and shift (Borusyak _et al._ , 2022) exogeneity. In our case, we argue that national employment growth rates are independent of local employment shares – hence arguing for shift exogeneity. This is the case due to our calculation of the instrument (leaving-one-out method). The only way through which industry growth of a certain municipality would influence the national employment growth rate would be through a major shock in this given municipality, which would then, in turn, influence other municipalities, which would accordingly influence the national growth rate. This scenario is likely in cases with a small number of municipalities. However, due to our high number of industries (over 600) and municipalities (over 5,000), we argue that our shifts are exogenous.<sup>27</sup> In addition, we acknowledge the fact that municipalities with similar industry structures might be subject to similar general trends as pointed out by Adao _et al._ (2019). While we follow the tests proposed by Borusyak _et al._ (2022) and Adao _et al._ (2019) in Section 5.2, we argue that even if similar industry structures are affected in similar ways, it is not necessarily worrisome in our setting. We are not trying to isolate a certain factor that affects employment but rather look at general employment trends. This is confirmed by Adao _et al._ (2019) who find that in their\\n\\n> 27Shares might be endogenous since we have a frequent panel at the municipality level.\\n\\n16\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"application the first stage differs less when looking at the application of Bartik (1991) since “the Bartik IV absorbs the bulk of the shift-share covariates that affect the change in the employment rate and wages across commuting zones.” Section 5.2 shows further robustness checks to the instrument.\\n\\n# **5 Results**\\n\\n## **5.1 Main Results**\\n\\n**Formal labor market participation** As a first step, we aim to understand if employment growth is directly impacting labor market participation of BF beneficiaries. Table 4 shows the results on formal labor market participation for the basic model and the IV approach in column (1) and (2) respectively. As can be seen, the basic model (column (1)) shows a small effect for our outcome of interest _labor market participation of at least one family member_ . However, due to the above-mentioned endogeneity challenges, which mainly relate to reverse causality, we expect a downward bias for our results. This is confirmed by looking at the results using the IV approach. Column (2) of Table 4 indicates that employment growth causes a small and significant increase in the probability of having at least one family member in formal dependent employment. An increase of 10% employment growth at the local level increases the probability of having at least one family member in formal dependent employment by 0.259 percentage points. While the increase is small compared to the average formal labor market participation of 21.58%, is it not negligible compared to the average transition probability of 2.65%. The results suggest that employment growth is leading to a small – however statistically significant – increase in the probability of having a family member in formal dependent employment and thus motivates us to look at the impact of employment growth on exit dynamics.\\n\\n**Exit probabilities** In the second step, we thus look at the impact of employment growth on the probability of leaving BF. Results of the basic model suggest the probability of exit changes significantly, however, to a small extent if employment at the municipality level increases (columns (1) and (2) of Table 5). As before, we interpret these results with caution, as potential endogeneity – particularly the risk of reverse causality – may bias the estimates downward. To address this concern, we turn to the IV approach (columns (3) and (4) of Table 5). The firststage results confirm the relevance of the instrument, indicated by its statistical significance and a sufficiently high F-statistic. We find that the effect of employment growth is positive\\n\\n17\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"Table 4: Formal labor market participation\\n\\n||(1)<br>In FLM<br>Basic Model|(2)<br>In FLM<br>IV Approach|\\n|---|---|---|\\n|Employment Growth_t−_1|0.0020**<br>(0.0008)|0.0259**<br>(0.0105)|\\n|Time-variant controls|✓|✓|\\n|Family-fxed efects|✓|✓|\\n|Time-fxed efects|✓|✓|\\n|F-Statistic||25.32|\\n|Mean FLMP|0.2158|0.2158|\\n|Average transition probability|0.0265|0.0265|\\n|N|52,020,809|51,354,456|\\n\\n_∗ ∗∗ ∗∗∗ p <_ 0 _._ 10, _p <_ 0 _._ 05, _p <_ 0 _._ 01 **Notes:** The table presents the main results from model (1) in column (1) and from the IV approach as specified in model (2) in column (2). The dependent variable is _a family has at least one member in formal dependent employment_ . The regression controls for family- and timefixed effects. Robust clustered standard errors at the municipality level are shown in parentheses.\\n\\nand significant. Hence, an increase of 10% employment growth at the local level increases the probability that a family leaves BF by 0.342 percentage points (column (4) Table 5). The average employment growth in our study period was 2.4%, which then, in turn, caused an increase in the average transition probability by 0.08 percentage points: from 8.61% to 8.69%. It should be remembered that Brazil went through different economic cycles in our study period. While there were economic upturns, there were also economic downturns, which in part explain the small results.\\n\\nTable 5: Exit probability\\n\\n||(1)<br>Exit BF<br>Basic Model|(2)<br>Exit BF<br>Basic Model|(3)<br>Exit BF<br>IV Approach|(4)<br>Exit BF<br>IV Approach|\\n|---|---|---|---|---|\\n|Employment Growth_t−_1|0.0036***|0.0036***|0.0334**|0.0342**|\\n||(0.0013)|(0.0013)|(0.0148)|(0.0151)|\\n|Time-variant controls||✓||✓|\\n|Family-fxed efects|✓|✓|✓|✓|\\n|Time-fxed efects|✓|✓|✓|✓|\\n|F-Statistic|||25.34|25.32|\\n|Mean exit probability|0.0861|0.0861|0.0861|0.0861|\\n|N|52,319,618|52,020,809|51,655,314|51,354,456|\\n\\n_∗ ∗∗ ∗∗∗ p <_ 0 _._ 10, _p <_ 0 _._ 05, _p <_ 0 _._ 01 **Notes:** The table presents the main results from model (1) in columns (1) and (2) and from the IV approach as specified in model (2) in columns (3) and (4). The dependent variable is _a family left BF_ . The regression controls for family- and time-fixed effects. Robust clustered standard errors at the municipality level are shown in parentheses.\\n\\n18\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"Those results can be understood as complementary to the results of Gerard _et al._ (2021). BF acts as both a social safety net and an economic stimulus that creates pathways to formal employment. The resulting labor market improvements facilitate beneficial exits that potentially free up resources for new beneficiaries.\\n\\n# **5.2 Robustness Checks**\\n\\n**Municipality level** As a first robustness check, we replicated the analysis at the municipality level. Instead of using the individual probability of leaving the program, we here use _exit rates_ as our outcome of interest.<sup>28</sup> Table A.3 shows that when replicating the analysis at the municipality level, we can see a similar effect compared to the individual level. Column (2) shows the outcome of the 2nd stage IV approach. If employment grows by 10%, the exit rate at the municipality increases by 0.245 percentage points.\\n\\n**Beneficiaries receiving the basic benefit** As a second robustness check, we restrict our sample to those beneficiaries that receive either only the basic benefit or the basic and variable benefit (based on predictions derived from family composition and the total benefit amount). We thus exclude all families that _only_ receive the variable benefit which is conditioned on having children. By doing so we are aiming to capture potential mechanical reasons – such as children approaching adulthood. The results shown in Panel A of Table A.4 are similar in significance level and magnitude as in our main specification.\\n\\n**Fraud** In addition, we introduce a variable to identify individuals who secure formal dependent employment subsequently exiting the CadUnico, while their family remains enrolled. This approach is designed to address the possibility of individuals being incentivized to remove themselves from the registry once they secure formal employment, thereby enabling them to access both labor income and the BF benefit concurrently. Including such a variable does not change our point estimates, as shown in Panel B of Table A.4.\\n\\n**CadUnico actualization** Further, we control for the share of CadUnico registries that have been updated.<sup>29</sup> We do so by including a variable that indicates the average value of updates by municipality. As seen in Panel C of Table A.4 including the actualization score does not alter the effect of employment growth on the probability of leaving BF.\\n\\n> 28This approach allows us to also consider entries in the program. Thus, we calculate exit rates by looking at the number of exits per semester divided by the total number of families in BF.\\n\\n> 29The MDS provides information on the actualization score (MDS, 2020).\\n\\n19\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"**Different instrument specification** Next, we re-estimate our main specification by using a different shift-share instrument. Instead of using industries to calculate the instrument, we use 2,527 different occupations based on RAIS to calculate the national growth rates (shifters) and the local occupational shares. As in our main specification of the instrument, we exclude occupations related to the public sector as well as public firms. Results shown in Panel D of Table A.4 remain positive and significant and even have slightly higher point estimates. Nevertheless, in our main estimation strategy we prefer to follow Bartik (1991) to calculate the instrument and to use industries instead of occupations.\\n\\n**Validity shift-share instrument** The recent literature has addressed the topic of the validity of shift-share instrument (see Goldsmith-Pinkham _et al._ (2020) and Borusyak _et al._ (2022)). As mentioned in Section 4 we follow the approach by Borusyak _et al._ (2022) to test our shift-share instrument for shock exogeneity.<sup>30</sup>\\n\\nThe authors suggest aggregating the data at the industry level to examine the conditional quasi-random shock assignment and the lack of correlation among shock residuals. Summary statistics for the shocks are presented in Table A.5. With a substantial number of shocks (8,579) and industries (660 at the 4-digit level and 278 at the 3-digit level), the statistics indicate that no single industry is significantly driving our observed effects. These findings demonstrate that we can mitigate concerns regarding the influence of any particular industry.\\n\\nAs a second step, we redo our analysis at the industry level as proposed by Borusyak _et al._ (2022). The authors show that the shift-share IV estimator equals the “second-stage coefficient from a _sn_ -weighted shock-level IV regression that uses the shocks _gn_ as the instrument”:<sup>31</sup> where _y_ ̄ _λ_<sup>_⊥_istheoriginal(residualized)outcome(exitrate),and</sup><sup>_x_ ̄</sup><sup>_⊥_</sup> _λ_<sup>istheoriginal(residual-</sup> ized) treatment (employment growth). As pointed out by the authors _y_ ̄ _λ_<sup>_⊥_“reflectstheaverage</sup> residualized outcome of the observations most exposed to the nth shock” and _x_ ̄<sup>_⊥_</sup> _λ_<sup>“isthesame</sup> weighted average of residualized treatment”. _x_ ̄<sup>_⊥_</sup> _λ_<sup>isinstrumentedby</sup><sup>_gλ_.</sup> Borusyak _et al._ (2022) show that their exposure-robust standard errors are asymptotically equivalent to the ones proposed by Adao _et al._ (2019), who note that locations with similar\\n\\n> 30 While our main analysis is done at the family level, we do the robustness checks at the municipality level since our treatment is happening at the municipality level.\\n\\n> 31In our case _sn_ equals _sλ_ .\\n\\n20\", \"datasets\": [{\"mention_name\": {\"text\": \"RAIS\", \"confidence\": 0.7397903203964233, \"start\": 229, \"end\": 233, \"clean_text\": \"RAIS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 229, \"end\": 233, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 229, \"end\": 233, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 229, \"end\": 233}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 229, \"end\": 233, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.5049911141395569, \"start\": 229, \"end\": 233}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7397903203964233, \"start\": 229, \"end\": 233}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8475356698036194, \"start\": 229, \"end\": 233}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8475356698036194, \"start\": 229, \"end\": 233}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"industry structures might be subject to similar general trends. The equivalence is achieved by including exposure-weighted averages as controls in model (3).<sup>32</sup> Nevertheless, in our setting, similar industry structures which are affected in similar ways, are not necessarily worrisome as mentioned in Section 4.<sup>33</sup> Column (1) in Table A.6 shows the results at the municipality level and using robust standard errors at the municipality level. Column (2) redoes the analysis at the industry level and uses robust standard errors at the exposure level. While our first stage F-Statistic stays significant, the significant level of our coefficient decreases; however, is still significant at the 10 % level. We acknowledge the fact that our coefficients differ slightly, however, we suspect that the difference arises due to the use of fixed effects in the panel data setting. Overall, we conclude that our instrument is still relevant at the industry level; however, the second-stage results have a lower significance level.\\n\\n# **5.3 Heterogeneities**\\n\\nWe next conduct heterogeneity analyses to examine how differences in municipal and individual characteristics shape families’ exit dynamics in response to “rising tides”.\\n\\n**Positive versus negative shifters** First, we are interested in the question of whether the positive or negative shifters drive our results. Panel A of Table 6 shows the results of splitting the data in those families that live in a municipality that experienced on average positive employment growth and those that experienced on average negative employment growth. Results are driven by positive shifters, causing an increase in the probability of leaving BF. While those results display an interesting heterogeneity, they also increase the credibility of our main question – if employment growth (instead of decrease) causes an increase in the likelihood of leaving the program.\\n\\n**Informality** Second, we aim to disentangle the effect for areas with historically high informality rates and areas with historically low informality rates. We classify municipalities by using the census data from 2010 into municipalities that have informality rates above and below\\n\\n> 32The STATA command _ssaggregate_ by Borusyak _et al._ (2022) is used. The command converts the shift-share IV dataset – in our case, at the municipality level – into a dataset of weighted shock-level aggregates. In other words, the outcome and treatment variables are regressed on exposure weights and a set of control variables to get to the industry-level aggregates.\\n\\n> 33As outlined above we are not trying to isolate a specific factor that affects employment but rather look at general employment trends. This is confirmed by Adao _et al._ (2019), who find that in their application, the first stage differs less when looking at the application of Bartik (1991) since “the Bartik IV absorbs the bulk of the shift-share covariates that affect the change in the employment rate and wages across commuting zones.”\\n\\n21\", \"datasets\": [{\"mention_name\": {\"text\": \"panel data\", \"confidence\": 0.3049046993255615, \"start\": 874, \"end\": 884, \"clean_text\": \"panel data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 874, \"end\": 884, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 874, \"end\": 884, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 874, \"end\": 884}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 874, \"end\": 884, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.5607836842536926, \"start\": 874, \"end\": 884}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3049046993255615, \"start\": 874, \"end\": 884}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4465830624103546, \"start\": 874, \"end\": 884}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4465830624103546, \"start\": 874, \"end\": 884}}, {\"mention_name\": {\"text\": \"census data from 2010\", \"confidence\": 0.806922197341919, \"start\": 2127, \"end\": 2148, \"clean_text\": \"census data from 2010\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2127, \"end\": 2148, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2127, \"end\": 2148, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2127, \"end\": 2148}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2127, \"end\": 2148, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.9972375631332397, \"start\": 507, \"end\": 528}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.806922197341919, \"start\": 507, \"end\": 528}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9774301052093506, \"start\": 507, \"end\": 528}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9774301052093506, \"start\": 507, \"end\": 528}}, {\"mention_name\": {\"text\": \"shift-share IV dataset\", \"confidence\": 0.8981162905693054, \"start\": 2312, \"end\": 2334, \"clean_text\": \"shift-share IV dataset\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2312, \"end\": 2334, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2312, \"end\": 2334, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2312, \"end\": 2334}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2312, \"end\": 2334, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.4774729609489441, \"start\": 692, \"end\": 714}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8981162905693054, \"start\": 692, \"end\": 714}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.6756845116615295, \"start\": 692, \"end\": 714}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6756845116615295, \"start\": 692, \"end\": 714}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"Table 6: Heterogeneity by municipality characteristics\\n\\n|Exit BF<br>IV Approach|\\n|---|\\n|**Panel A: Positive vs. negative employment growth**|\\n|**On average neg. shifter**<br>-0.0177<br>(0.0258|\\n|F-Statistic<br>13.70|\\n|Mean exit probability<br>0.0792|\\n|N<br>36,151,808|\\n|**On average pos. shifter**<br>0.0598***<br>(0.0221)|\\n|F-Statistic<br>87.63|\\n|Mean exit probability<br>0.1026|\\n|N<br>15,220,945|\\n|**Panel B: Informality Level Municipality**|\\n|**Above average informality**<br>0.0821***<br>(0.0254)|\\n|F-Statistic<br>15.43|\\n|Mean exit probability<br>0.0710|\\n|N<br>35,925,648|\\n|**Below average informality**<br>0.0603***<br>(0.0225)|\\n|F-Statistic<br>12.47|\\n|Mean exit probability<br>0.1223|\\n|N<br>15,402,785|\\n|Time-variant controls<br>✓|\\n|Family-fxed efects<br>✓|\\n|Time-fxed efects<br>✓|\\n\\n_∗ ∗∗ ∗∗∗ p <_ 0 _._ 10, _p <_ 0 _._ 05, _p <_ 0 _._ 01 **Notes:** The table presents the results from the IV approach as specified in model (2). In Panel A municipalities are split into municipalities with on average positive and on average negative shifters. In Panel B municipalities are split into below- and above-average informality rates according to the Census of 2010. The regression controls for family- and time-fixed effects. Robust clustered standard errors at the municipality level are shown in parentheses.\\n\\nthe average. As shown in Panel B of Table 6 the effects are positive and significant for both types of municipalities. However, results show that the effects are stronger in municipalities with higher rates of informality indicating that in places, where formal dependent employment is less present, an extension of it causes a greater impact in terms of exits from the program.\\n\\n**Family composition** Third, we split the sample into families that have just one adult living in the family vs. families with various adults. Panel A of Table 7 shows that the overall effect is driven by families with more than 1 adult. Those results seem plausible since single-headed households are commonly facing multiple constraints to benefit from formal employment growth\\n\\n22\", \"datasets\": [{\"mention_name\": {\"text\": \"Census of 2010\", \"confidence\": 0.7302218079566956, \"start\": 1151, \"end\": 1165, \"clean_text\": \"Census of 2010\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1151, \"end\": 1165, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1151, \"end\": 1165, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1151, \"end\": 1165}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1151, \"end\": 1165, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.9938435554504395, \"start\": 501, \"end\": 515}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7302218079566956, \"start\": 501, \"end\": 515}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7457096576690674, \"start\": 501, \"end\": 515}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7457096576690674, \"start\": 501, \"end\": 515}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"- such as childcare duties for example (Blundell _et al._ , 2018). In line with that, we see that families with smaller children – thus also having higher labor supply constraints – respond slightly less to employment growth than families with older children (Panel B of Table 7).\\n\\n**Education** Last, we split our analysis based on distinct educational attainment levels. As shown in Panel C of Table 7, households headed by individuals with moderate education levels exhibit the most pronounced response to enhancements in local employment. Similarly, for families with heads having lower education, we find an increase in the likelihood of program departure when local employment improves. Conversely, households led by individuals with higher educational qualifications also see an increased probability of program exit; however, the effect is smaller and loses statistical significance. One plausible rationale behind this observation is that households headed by highly educated individuals already see an increased probability of departing from the program, as shown by the results obtained from prior survival analyses. Consequently, improvements in local employment might not be a necessary condition for those families to leave. Furthermore, these households might encounter supplementary constraints, as shown by the results in Panel A in Table 7.\\n\\n# **6 Conclusion**\\n\\nCCTs have been a popular tool for poverty alleviation and human capital development. The impact of transfers on education and health outcomes has been the subject of several studies (Kabeer and Waddington, 2015; Mill ́an _et al._ , 2019; Baird _et al._ , 2014). In contrast, graduation patterns of the transfers and the dynamics out of the programs remain vastly understudied. This paper contributes to the literature on cash transfers, CCTs in particular, by providing evidence on employment growth as an opportunity to graduate from social assistance programs. To the best of our knowledge, graduation from these kinds of programs has been highly neglected so far, possibly due to the lack of identified longitudinal data.\\n\\nWe start by analyzing the sociodemographic factors related to the program graduation and show that being male, being better educated, and living in urban areas are associated with a shorter time in BF. To isolate the effect of local employment growth on the probability of leaving BF, we combine several administratively identified datasets and follow families over a period of 84 months. By applying a shift-share IV approach and controlling for family- and time-fixed effects, we find that an increase in local employment has a positive and significant effect on 23\", \"datasets\": [{\"mention_name\": {\"text\": \"administratively identified datasets\", \"confidence\": 0.7119243741035461, \"start\": 2410, \"end\": 2446, \"clean_text\": \"administratively identified datasets\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2410, \"end\": 2446, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2410, \"end\": 2446, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2410, \"end\": 2446}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2410, \"end\": 2446, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.8955310583114624, \"start\": 744, \"end\": 780}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7119243741035461, \"start\": 744, \"end\": 780}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.738320529460907, \"start\": 744, \"end\": 780}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.738320529460907, \"start\": 744, \"end\": 780}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"leaving BF. Those results apply especially to families with less labor supply constraints. Several robustness checks – such as replicating the analysis at the municipality level, and looking at different specifications of the instrument – confirm our results. Additionally, our results show that employment growth increases the likelihood of having a family member in formal dependent employment, highlighting a potential pathway through which employment growth drives exits from the program. However, we acknowledge that we cannot capture all potential mechanisms – such as beneficiaries transitioning into self-employment.\\n\\nOne potential concern revolves around the classification of exits as _desirable_ . Although our analysis reveals that only a small proportion of beneficiaries rejoins the program during the study period, it is crucial to interpret these findings cautiously, considering the presence of a waiting list during the period. The outcomes presented in this paper indicate that beneficiaries of the BF program experience positive outcomes from local employment growth, leading to their exit from the program. It underscores the importance of safety nets being designed with incentives that align with local economic dynamics. However, it is imperative that such safety nets exhibit sensitivity in both directions. While it is desirable for individuals to graduate from the program when their financial situation allows, the safety net should also be responsive in the event of families facing a loss in income. This highlights the need for a balanced and incentivecompatible design that accommodates both positive and negative shifts in beneficiaries’ economic circumstances.\\n\\nOur results indicate that the BF program, in its 2019 and prior design, effectively facilitated the graduation of beneficiaries during periods of economic upturn. However, to enhance the program’s impact, specific public policies, such as employment training and childcare facilities, should be tailored to target the BF program. This targeted approach aims to maximize the opportunities for beneficiaries to capitalize on positive employment trends at the local level. Furthermore, ensuring the responsiveness of the safety net both – in and out of the program – is crucial for its overall effectiveness.\\n\\n24\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099508001182418663/pdf/IDU1527571dc1f7d114c8a1ae5917c1b8225ec9c.pdf\", \"pages\": [26]}}]}"}}},{"rowIdx":82,"cells":{"doc_id":{"kind":"string","value":"34241308"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34241308\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"admreg\": \"Other,Other\", \"admreg_key\": \"912972,912972\", \"authors\": {\"0\": {\"author\": \"Thomas,Alastair Geoffrey Arthur\"}}, \"count\": \"World\", \"count_key\": \"517191\", \"docna\": {\"0\": {\"docna\": \"VAT Rate Structures in Theory and Practice\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"EFI-MTI-Macro/Fiscal-Tax (EMFTX)\", \"projn\": \"1W-Oge - Supporting The Progressivity Of The\\n Tax System And The Re -- P177907\", \"subsc\": \"Mining,Central Government (Central Agencies),Oil and Gas,Other Energy and Extractives\", \"trustfund\": \"TF0B9110-Supporting the Progressivity of\\n the Tax System and the Revenue Mobiliza\", \"trustfund_key\": \"2190600\", \"theme\": \"Inclusive Growth,Energy,Economic Policy,Public Finance Management,Economic Growth and Planning,Fiscal Policy,Energy Policies & Reform,Environment and Natural Resource Management,Domestic Revenue Administration,Public Sector Management,Tax policy\", \"prdln\": \"Advisory Services & Analytics\", \"prdln_key\": \"1416466\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34241308\"}, \"repnb\": \"WPS10677\", \"docdt\": \"2024-01-18T00:00:00Z\", \"datestored\": \"2024-01-18T07:20:20Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"Most countries’ value-added\\n tax (VAT) systems apply reduced VAT rates to a selection of\\n expenditure items in order to achieve distributional goals,\\n and (to a lesser extent) social and cultural objectives.\\n This paper assesses the case for applying reduced VAT rates,\\n with a particular focus on OECD countries where reduced\\n rates feature prominently. It examines both the theoretical\\n and empirical evidence, as well as practical considerations,\\n and concludes that the case for reduced VAT rates is weak.\\n In particular, the optimal indirect tax literature finds no\\n redistributive role for reduced VAT rates when other more\\n direct instruments are available. These theoretical findings\\n are supported by the empirical literature that shows reduced\\n VAT rates to be a poorly targeted means of supporting lower\\n income households, particularly when compared to targeted\\n cash transfer programs. Similarly, reduced VAT rates are\\n unlikely to be a well-targeted way to encourage consumption\\n of merit goods, while they also create significant\\n administrative complexity. These findings have significant\\n implications for tax reform in both developed and developing\\n economies. In particular, where countries have the\\n administrative capacity to implement effectively targeted\\n cash transfer programs, they should use these programs to\\n support poorer households instead of reduced VAT rates.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10677; PROSPERITY\", \"display_title\": \"VAT Rate Structures in Theory and Practice\", \"disclosure_date\": \"2024-01-18T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099920201182414099/text/IDU12d02d19f1336a14e0c1b53714d134b306a9f.txt\", \"listing_relative_url\": \"/research/2024/01/34241308/vat-rate-structures-theory-practice\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099920201182414099/IDU12d02d19f1336a14e0c1b53714d134b306a9f\", \"new_url\": \"2024/01/34241308/IDU12d02d19f1336a14e0c1b53714d134b306a9f\", \"ext_pub_date\": \"2024-01-18T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Taxation m1327615 1121,Tax Policy m1327616 983\", \"docm_id\": \"IDU-2d02d19f-336a-4e0c-b537-4d134b306a9f\", \"chronical_docm_id\": \"IDU-2d02d19f-336a-4e0c-b537-4d134b306a9f\", \"publishtoextweb_dt\": \"2024-01-18T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10677\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"projectid\": \"P177907\", \"guid\": \"099920201182414099\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"26\", \"author_keys\": {\"0\": {\"author_key\": \"2070334\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099920201182414099/IDU12d02d19f1336a14e0c1b53714d134b306a9f\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10677 VAT Rate Structures in Theory and Practice _Alastair Thomas_ Macroeconomics, Trade and Investment Global Practice January 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10677\\n\\n# **Abstract**\\n\\nMost countries’ value-added tax (VAT) systems apply reduced VAT rates to a selection of expenditure items in order to achieve distributional goals, and (to a lesser extent) social and cultural objectives. This paper assesses the case for applying reduced VAT rates, with a particular focus on OECD countries where reduced rates feature prominently. It examines both the theoretical and empirical evidence, as well as practical considerations, and concludes that the case for reduced VAT rates is weak. In particular, the optimal indirect tax literature finds no redistributive role for reduced VAT rates when other more direct instruments are available. These theoretical findings are supported by the empirical literature that shows reduced VAT rates to be a poorly targeted means of supporting lower income households, particularly when compared to targeted cash transfer programs. Similarly, reduced VAT rates are unlikely to be a well-targeted way to encourage consumption of merit goods, while they also create significant administrative complexity. These findings have significant implications for tax reform in both developed and developing economies. In particular, where countries have the administrative capacity to implement effectively targeted cash transfer programs, they should use these programs to support poorer households instead of reduced VAT rates.\\n\\nThis paper is a product of the Macroeconomics, Trade and Investment Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The author may be contacted at athomas@worldbank.org.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# VAT Rate Structures in Theory and Practice\\n\\n## Alastair Thomas\\n\\nJEL codes: H2, H21, H22, H23 Key words : VAT, value-added tax, reduced rates, redistribution\\n\\n> Alastair Thomas is a Senior Economist in the Fiscal Policy and Sustainable Growth Unit of the Macroeconomics, Trade and Investment Global Practice at the World Bank. Helpful comments were gratefully received from Pierre Bachas, Bert Brys, John Creedy, Norman Gemmell, Somik Lall, Emilia Skrok and Kurt Van Dender.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"|Contents|\\n|---|\\n|1<br>Introduction .......................................................................................................................................... 3|\\n|2<br>VAT rate structures in OECD countries ................................................................................................. 4|\\n|3<br>Optimal indirect tax theory ................................................................................................................... 7|\\n|4<br>The case for reduced VAT rates .......................................................................................................... 11|\\n|4.1<br>Redistribution ............................................................................................................................. 11|\\n|4.2<br>Merit goods/externalities ........................................................................................................... 15|\\n|4.3<br>Efficiency ..................................................................................................................................... 16|\\n|4.4<br>Practical implementation issues ................................................................................................. 17|\\n|5<br>Conclusion ........................................................................................................................................... 18|\\n|References .................................................................................................................................................. 20|\\n\\n2\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"# 1 Introduction\\n\\nThe value-added tax (VAT) is a crucial component of most tax systems, generating substantial tax revenue. This is especially the case in developing economies, where reduced administrative capacity and significant informality often restrict the ability to raise significant tax revenue from income and other direct taxes.<sup>1</sup> VAT revenue is therefore crucial to helping countries finance their development needs. However, a key source of revenue loss from VAT systems in almost all countries is through the provision of reduced VAT rates for various goods and services. In particular, most countries apply reduced VAT rates (or exemptions) to a range of products that are typically consumed disproportionately by poorer households in order to achieve distributional goals. Meanwhile, the pursuit of social and cultural objectives, among others, have led to the application of reduced VAT rates on an even broader range of consumption in many countries.\\n\\nThis paper assesses the case for applying reduced VAT rates as compared to adopting a single-rate VAT structure. It examines both the theoretical and empirical evidence on the use of reduced VAT rates, as well as practical considerations. While the paper focuses predominantly on OECD country experience (on which most of the previous literature is based), it also considers recent evidence for developing economies, and draws on this range of evidence to identify policy implications for both developed and developing economies.\\n\\nOverall, the paper finds that the case for applying reduced VAT rates is weak. In particular, the optimal indirect tax literature finds no redistributive role for reduced VAT rates when other more direct instruments are available. These theoretical findings are supported by the empirical literature that shows reduced VAT rates to be a poorly targeted means of supporting lower income households, and the greater – though still imperfect – targeting ability of cash transfer programs. Similarly, reduced VAT rates are unlikely to be a well-targeted way to encourage consumption of merit goods, in part due to their _ad valorem_ nature. Meanwhile, efficiency arguments, with the restricted exception of some substitutes for home production, favor the adoption of a single-rate VAT structure, which would also bring significant administrative benefits.\\n\\nThese findings have significant implications for tax reform in both developed and developing economies. In particular, where countries have the administrative capacity to effectively implement targeted cash transfer programs, they should use these programs to support poorer households instead of using reduced VAT rates. In more limited capacity environments, a simpler universal cash transfer could even be adopted – as this would still better target support to poorer households than reduced VAT rates. Only where there is no administrative capacity to effectively implement a cash transfer program should reduced VAT rates be considered as a tool to target support to poorer households. Even then, the potential impact of informality on the ability to target poorer households should be considered.\\n\\nCountries should also consider removing reduced VAT rates aimed at non-distributional goals where a more effective instrument is available to achieve the particular policy goal. At a minimum, the merits of these reduced VAT rates should be reassessed in light of their negative distributional impact. Finally, countries that currently do not employ reduced rates in their VAT system should be very cautious\\n\\n> 1 VAT revenue comprises almost 33% of total tax revenue, on average, in low-income countries, as compared to around 21%, on average, in OECD countries. (3-year average 2019-2021; Government Finance Statistics Database, 2023).\\n\\n3\", \"datasets\": [{\"mention_name\": {\"text\": \"Government Finance Statistics Database\", \"confidence\": 0.6841052174568176, \"start\": 3756, \"end\": 3794, \"clean_text\": \"Government Finance Statistics Database\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3756, \"end\": 3794, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3756, \"end\": 3794, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3756, \"end\": 3794}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3756, \"end\": 3794, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.7414833903312683, \"start\": 1071, \"end\": 1109}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6841052174568176, \"start\": 1071, \"end\": 1109}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7007675170898438, \"start\": 1071, \"end\": 1109}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7007675170898438, \"start\": 1071, \"end\": 1109}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"regarding any proposals to introduce reduced rates. This is particularly important in light of the potential political economy difficulties of removing them once they are in place.\\n\\nThis paper proceeds as follows: Section 2 summarizes the VAT rate structures currently in place in OECD countries, highlighting the widespread use of reduced VAT rates. Section 3 then reviews the optimal indirect taxation literature on the choice between uniform and multi-rate structures. Drawing on this literature, section 4 then assesses the various arguments used in practice for and against the adoption of reduced VAT rates in OECD countries. Section 5 concludes and highlights policy implications.\\n\\n# 2 VAT rate structures in OECD countries\\n\\nAs of 2022, 174 countries have adopted a VAT (OECD, 2022). This includes 37 of the 38 OECD member countries – the exception being the United States, which operates a range of state-level retail sales taxes instead. Although the general principles underlying each VAT system are the same, there are still significant differences in the systems implemented in different countries, and, in particular, in the rate structures adopted.<sup>2</sup> Table 1 shows the variation in rate structures across OECD countries. Most “older” VAT systems tend to have multi-rate structures, with one or more reduced rates (including zero rates) applying to a significant number of goods and services. This is particularly the case in Europe where countries’ VAT rate structures are guided by the EU VAT Directive, which generally allows for up to two reduced VAT rates in addition to the standard rate.<sup>3</sup> Meanwhile, the VAT systems in a small number of countries – such as Chile and New Zealand – apply a single rate to most, if not all, goods and services.\\n\\nThe types of goods and services subject to these reduced rates also varies. Nevertheless, there are some common trends. For example, as illustrated in Figure 1, the vast majority (31 out of 37 OECD countries that have a VAT) apply a reduced rate to either basic food products or, more often, to a very broad range of food and non-alcoholic beverages purchased for home consumption (“food” hereafter).<sup>4</sup> The general rationale for providing a reduced rate on food is distributional – i.e. to provide support to poorer households. Following a similar rationale, many countries also provide a reduced rate for pharmaceutical products and water supply. Meanwhile, a smaller number of countries apply reduced rates to refuse and sewage services, energy products, and to children’s clothing or shoes. An increasing number of countries now also apply reduced rates to feminine hygiene products. Due to its significant budget share, the reduced rate on food is by far the most significant reduced VAT rate applied in OECD countries.<sup>5</sup>\\n\\n> 2 VAT systems also differ in terms of the extent of exemptions, the registration thresholds for businesses, the use of special taxation methods for specific supplies, and restrictions on the right to deduct VAT on specific inputs. OECD (2022) discusses these in more detail.\\n\\n3 The 2006 EU VAT Directive requires the standard rate to be at least 15%. Reduced rates must be at least 5% and can only be applied to the set of goods and services specified in Annex III of the Directive, though a number of derogations are provided that allow some countries to maintain reduced rates at a rate lower than 5% on some products.\\n\\n4 Unless otherwise specified, expenditure on food and non-alcoholic beverages is separated into two categories in this paper: food and non-alcoholic beverages purchased for home consumption, including takeaway food (“food”); and food and non-alcoholic beverages consumed in restaurants, canteens and cafeterias (“restaurant food”).\\n\\n5 Food made up 18.5% of total household expenditure, on average, in 23 OECD countries examined by Thomas (2020) using household budget survey microdata. In contrast, the next most common expenditure categories – newspapers and books – together made up only around 1% of total household expenditure, on average, in the same 23 countries.\\n\\n4\", \"datasets\": [{\"mention_name\": {\"text\": \"household budget survey microdata\", \"confidence\": 0.6528893113136292, \"start\": 3906, \"end\": 3939, \"clean_text\": \"household budget survey microdata\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3906, \"end\": 3939, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3906, \"end\": 3939, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3906, \"end\": 3939}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3906, \"end\": 3939, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"microdata\", \"confidence\": 0.9371300339698792, \"start\": 769, \"end\": 802}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6528893113136292, \"start\": 769, \"end\": 802}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.970452070236206, \"start\": 769, \"end\": 802}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.970452070236206, \"start\": 769, \"end\": 802}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"**Table 1. VAT rates in OECD countries as at 1 January 2022**\\n\\n||Standard rate|Reduced rates**|\\n|---|---|---|\\n|AUS|10.0|0.0|\\n|AUT*|20.0|10.0/13.0|\\n|BEL<br>|21.0|0.0/6.0/12.0|\\n|CAN*|5.0|0.0|\\n|COL*|19.0|0.0/5.0<br>|\\n|CHE|7.7|0.0/2.5/3.7|\\n|CHL|19.0|-|\\n|CRI|13.0|1.0/2.0/4.0|\\n|CZE|21.0|10.0/15.0|\\n|DEU|19.0|7.0|\\n|DNK|25.0|0.0|\\n|ESP*|21.0|4.0/10.0|\\n|EST|20.0|0.0/9.0|\\n|FIN|24.0|0.0/10.0/14.0|\\n|FRA*|20.0|2.1/5.5/10.0|\\n|GBR|20.0|0.0/5.0|\\n|GRC|24.0|6.0/13.0|\\n|HUN|27.0|5.0/18.0|\\n|IRL|23.0|0.0/4.8/9.0/13.5|\\n|ISL|24.0|0.0/11.0|\\n|ISR*|17.0|0.0|\\n|ITA|22.0|4.0/5.0/10.0|\\n|JPN|10.0|8.0|\\n|KOR|10.0|0.0|\\n|LTU|21.0|5.0/9.0|\\n|LUX|17.0|3.0/8.0/14.0|\\n|LVA|21.0|5.0/12.0|\\n|MEX*|16.0|0.0|\\n|NLD|21.0|9.0|\\n|NZL|15.0|0.0|\\n|NOR|25.0|0.0/12.0/15.0|\\n|POL|23.0|5.0/8.0|\\n|PRT*|23.0|6.0/13.0|\\n|SLV|22.0|5.0/9.5|\\n|SVK|20.0|10.0|\\n|SWE|25.0|0.0/6.0/12.0|\\n|TUR|18.0|1.0/8.0|\\n\\nSource: OECD (2022)\\n\\n# **Table notes**\\n\\n*Austria: a standard rate of 19% applies in Jungholz and Mittelberg. Canada: provincial sales taxes also apply. Colombia: 0% rate applies to hygiene products in the department of Amazonas. France: rates of 0.9%, 2.1%, 10%, 13% and 20% apply in Corsica; rates of 1.05%, 1.75%, 2.1% and 8.5% apply to overseas departments excluding French Guyana and Mayotte. Israel: a rate of 0% applies when an Eilat resident dealer buys goods from Eilat non-residents. Mexico: 8% reduced rate applies in border regions. Portugal: rates of 5%, 10% and 18% apply in the Azores; rates of 5%, 10% and 22% apply in Madeira. Spain: rates of 0%, 2.75%, 3%, 7%, 9.5%, 13.5% and 20% apply in the Canary Islands; rates of 0.5%, 1%, 2%, 4%, 6%, 8%, 9% and 10% apply in Ceuta and Melilla. Temporary COVID-19-related rates are not specified in Table 1 (see OECD, 2022, for details). **Includes zero rates on domestic supplies. Excludes zero rates on exports and other supplies subject to similar treatment such as international transport or supplies to embassies, international organisations and diplomatic missions.\\n\\n5\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"**Figure 1. Common reduced VAT rates in OECD countries as at 1 January 2022**\\n\\n<!-- Start of picture text --> Number of countries<br>35<br>30<br>25<br>20<br>15<br>10<br>5<br>0<br><!-- End of picture text --> Source: Author’s calculations based on OECD (2022) and European Commission (2021).\\n\\nMost countries also use reduced VAT rates to encourage the consumption of certain goods and services with perceived social or cultural benefits. The most common examples are books and newspapers, which are subject to reduced VAT rates in 29 and 28 (out of 37) countries, respectively. As Figure 1 shows, countries often also provide reduced rates for cinema, theatre and concerts, and to a lesser extent for amusement parks, museums, sporting events, sporting facilities and zoos.\\n\\nSome countries also introduce reduced rates on services that are close substitutes with home supply in order to encourage (market) employment, such as repair, domestic cleaning and hairdressing services. Many countries also provide reduced rates with a less clear policy rationale. The most common of these are for hotel (and/or other) accommodation, restaurant food, and passenger transport. Reduced rates for hotels and restaurant food may, for example, be introduced to encourage the employment of low-skilled workers, or for a perceived social benefit. Reduced rates for (domestic) passenger transport may be motivated by environmental concerns, but they may also be justified on employment or distributional grounds (on the basis that they lower the cost of commuting to work).<sup>6</sup> The existence of these reduced rates may also partially result from interest group pressure rather than from a clear policy rationale.\\n\\nThough not the focus of this paper, it is also common for countries to exempt certain expenditures from VAT (i.e. zero rating with no ability to deduct input tax). Again, it is European countries that tend to have the greatest number of exemptions, whereas Chile, Japan and New Zealand tend to have the fewest.<sup>7</sup> Most countries exempt certain sectors considered essential for social reasons – particularly education, healthcare and charities. In some cases, practical reasons have led countries to use exemptions – for example, in every OECD country, most or all financial services are exempted due to the difficulty in\\n\\n> 6 International air travel is universally zero-rated due to the practical difficulty in assigning taxation rights to a service that may occur across multiple countries. Some countries also apply zero rates to other forms of international passenger travel.\\n\\n> 7 Unlike a zero rate, some VAT will still generally be payable on exempted goods and services. This is because the inability to deduct input tax will increase the cost, and hence the price, of the final good or service (unless the cost is fully borne by a party within the supply chain).\\n\\n6\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"determining the appropriate margin on which to apply VAT. Other sectors, such as postal services and gambling have often been exempted for a variety of historical reasons.<sup>8</sup> The application of VAT registration thresholds also reduces the size of the VAT base. These are generally aimed at removing small businesses from the tax net where the associated compliance and administrative costs would be disproportionate relative to the amount of VAT revenue generated. The level of these thresholds varies significantly across OECD countries.<sup>9</sup>\\n\\n# 3 Optimal indirect tax theory\\n\\nBefore assessing the various arguments used for and against the adoption of these reduced VAT rates, this section first reviews the optimal indirect taxation literature on the choice between uniform and differentiated rate structures. The appropriate rate structure for indirect taxation has been the subject of much theoretical work. While early papers by Ramsey (1927) and Corlett and Hague (1953) focused purely on efficiency, the modern optimal indirect taxation literature – beginning with Diamond and Mirrlees (1971) – has attempted to balance equity and efficiency objectives within a single framework.\\n\\nIn a world with a single representative consumer, Ramsey (1927) examines how to raise a given amount of tax revenue from indirect taxation at the lowest distortionary cost. His model excludes the use of any other tax.<sup>10</sup> If cross-price effects are zero, then the famous “inverse elasticity” rule results – where more inelastic goods should be taxed more heavily, and vice versa. If there are cross-price effects then the less prescriptive result is found that taxes should be levied to produce equal proportional reductions in the consumption of each good.<sup>11</sup> Corlett and Hague (1953) also consider a single representative consumer model, but with three goods: leisure and two taxed goods. They examine whether a shift away from uniform taxation would be efficiency improving. They find that efficiency can be improved by taxing more heavily the good that is more complementary with leisure. The rationale here is that, by taxing complements with leisure more heavily than other goods, this will discourage leisure and so reduce the underlying distortion to the labor supply decision. In contrast, if all goods are equally complementary with leisure then uniformity will be optimal.<sup>12</sup> 8 Exemptions are also often applied to cultural services, legal aid, precious metals, public transport and water supply. There is only limited consistency in the types of goods and services that countries apply exemptions to as opposed to applying reduced rates. For example, cultural services, public transport and water supply are exempt in some countries but subject to a reduced rate in others. This can even be the case for the most commonly applied exemptions – for example, Australia zero-rates rather than exempts education and healthcare. In contrast, the Republic of Korea provides no reduced rates and instead uses exemptions to address the distributional, cultural and social goals that many countries attempt to address through reduced VAT rates.\\n\\n9 Full details on VAT design features in each OECD country, including reduced rates, exemptions and registration thresholds, are provided in the OECD’s biennial _Consumption Tax Trends_ publication (see OECD, 2022, for the latest edition).\\n\\n10 Atkinson and Stiglitz (1980) extend the Ramsey model to show that if a lump sum tax is possible and the government only cares about efficiency then there should be no indirect tax at all. This is because the lump sum tax would be a more efficient means of raising the required revenue than the distortionary indirect tax.\\n\\n11 As Heady (1993) notes, this technically only applies for small revenue requirements, and that additional assumptions of no income effects and linear demand curves are necessary for it to hold for large revenue requirements.\\n\\n12 Atkinson and Stiglitz (1972) highlight that Corlett and Hague’s result is driven by the untaxed nature of leisure, not leisure itself. They note the general principle in such a three good model that if there is one untaxed good then we 7\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"Diamond and Mirrlees (1971) and Diamond (1975) extend the Ramsey model to a many person setting in order to take account of distributional considerations.<sup>13</sup> They maximize social welfare functions that apply different weightings to the utility of different individuals. Both studies find that the Ramsey proportional reduction rule must be altered to depend on who consumes the goods. If the weights of the social welfare function are based on income, then the optimal set of tax rates should result in a smaller proportional reduction in demand for goods consumed in greater proportions by the poor.<sup>14</sup> While the above models exclude the possibility of an income tax, the seminal paper by Atkinson and Stiglitz (1976) brings together the optimal income tax and indirect tax literature to examine the simultaneous optimization of both income and consumption taxes. Following Mirrlees (1971), the government’s problem is shaped as one of optimization subject to information constraints. Individuals are assumed to have identical preferences, but to differ in ability level. As ability is not directly observable, income – and potentially consumption – must be taxed instead. The model assumes there is no savings and the only source of income is labor – so the income tax is effectively a wage tax.<sup>15</sup> Within this framework, they show that if an optimal non-linear income tax is available and preferences between consumption and leisure are “weakly separable” (or, in other words, all goods are equally complementary with leisure), then a uniform indirect tax will be optimal.<sup>16</sup> Equivalently, as in their model a uniform indirect tax can be subsumed within an income tax, no indirect taxation is necessary at all.\\n\\nSubsequent papers, such as Christiansen (1984) and Edwards et al. (1994), confirm the uniformity result where preferences are weakly separable. Where weak separability does not hold, they show that higher indirect taxes should be imposed on goods that are more complementary with leisure – consistent with the Corlett and Hague (1953) result.<sup>17</sup> Unlike Diamond and Mirrlees (1971) and Diamond (1975), the usefulness of indirect taxes is purely to improve efficiency as opposed to assist redistribution – which can now be more effectively addressed through the income tax.\\n\\nThe intuition behind this is drawn out by Edwards et al. (1994) using a model where individuals have just two ability levels.<sup>18</sup> They show that taxing complements with leisure makes it less attractive for higher ability individuals to “mimic” lower ability individuals (by working less), and this relaxes the incentive compatibility constraint that otherwise restricts the degree of redistribution that can be achieved through the income tax. This efficiency rationale for differential taxation is in striking contrast to the redistributionbased arguments typically used in practice to justify differential VAT rates. Meanwhile, if all goods are equally complementary with leisure, a mimicking high ability individual will consume the exact same bundle of goods as a lower ability individual with the same income, so consumption will not provide any should tax the greater complement with it more heavily as this is a way of indirectly taxing the untaxed good. The uniformity result in the absence of complements with leisure is also found by Sandmo (1974).\\n\\n> 13 Diamond’s (1975) model also includes the presence of a lump sum tax.\\n\\n14 Diamond and Mirrlees (1971) also develop the famous production efficiency result that, externalities aside, indirect taxes should not be levied on intermediate goods but only on final consumption. The intuition behind this result, as succinctly summarized by Crawford et al. (2010), is that “any distortion of production decisions reduces aggregate output, which cannot be wise so long as there is some useful purpose to which that output could be put.” (p283).\\n\\n15 All references in this section to an income tax should be interpreted as a wage tax.\\n\\n16 Subsequent work by Kaplow (2006) and Laroque (2005) shows that this result still holds even if the income tax is not optimal.\\n\\n17 Atkinson and Stiglitz (1976) actually make the opposite conclusion to Corlett and Hague (1953) when consumption and labor are not weakly separable: that you should tax more heavily goods that are complements with labor – i.e. substitutes with leisure. Kaplow (2010) examines the discrepancy in detail and points to a revised interpretation of the Atkinson and Stiglitz model’s setup where results become consistent with Corlett and Hague (1953).\\n\\n> 18 A similar analysis is undertaken by Nava et al. (1996).\\n\\n8\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"additional information on ability and hence indirect taxes cannot be used to relax the incentive compatibility constraint.\\n\\nThe latter case emphasizes the importance of the assumption of homogeneous tastes. In contrast, if individuals have heterogeneous tastes then a higher and lower ability individual earning the same income would no longer consume the exact same bundle of goods, and so consumption patterns may once again provide information about ability. Mirrlees (1976), Saez (2002) and Gauthier and Henriet (2018) consider models with heterogeneous tastes. They show that if tastes depend on ability, then a higher indirect tax should, in general, be imposed on a good if higher ability individuals have a relatively strong taste for that good.<sup>19</sup> Another strand of literature has considered the impact of home production on optimal tax rates. Kleven et al. (2000) develop a single representative consumer model incorporating home production. This effectively produces a modified Corlett and Hague (1953) rule where the tax distortion favoring (untaxed) home production pushes down the optimal tax rate on consumer services that are substitutes for home production (e.g. house repairs, cleaning services). Where goods and services are equally complementary with leisure (so that uniform taxation would otherwise be optimal), they show that substitutes for home production should unequivocally face a relatively low tax rate. Piggott and Whalley (2001) develop a very similar model incorporating home production. They simulate their model based on Canadian data and find that Canada’s 1990 VAT reform that brought consumer services into the indirect tax net was welfare reducing.<sup>20</sup> Kleven (2004) and Boadway and Gahvari (2006) consider models where both goods and time must be allocated. Kleven (2004) finds that market goods that take more time to be consumed should be taxed more heavily than those that take less time, or those that save time. Boadway and Gahvari (2006) find that goods that are less pleasurable to consume should be taxed more heavily than those that are more pleasurable and that, for less pleasurable goods, those that take longer to be consumed should be taxed more heavily. As Crawford et al. (2010) note, intuitively this is once again acting to reduce tax-induced disincentives to engage in market work, but additionally recognizing that it is preferable to spend any given amount of time in a pleasurable activity as compared to an unpleasurable one. They give the example that DVDs (which are pleasurable but time consuming) should be taxed less heavily than ironing boards (as ironing is dull and time consuming).\\n\\nA key question arising from the above literature is whether preferences between consumption and leisure are in fact weakly separable. Unfortunately, only a small number of empirical studies have examined this question. Browning and Meghir (1991) and Crawford et al. (2010) estimate demand systems for the United Kingdom and both reject weak separability – suggesting a role for differential taxation. Crawford et al. (2010), for example, find that most food products, fuels, tobacco, children’s clothing and public transport are complements with leisure. Meanwhile, alcoholic drinks, food eaten out, motor fuels and leisure goods are complements with work.<sup>21</sup>\\n\\n> 19 Cremer et al. (2001) also show a role for differential taxation if individuals have different initial endowments of goods.\\n\\n> 20 Cremer and Gahvari (2015) also show in an Atkinson and Stiglitz (1976) framework that weak separability in preferences between consumption and leisure is no longer sufficient for uniformity when there is household production. However, if preferences are weakly-separable in market goods versus leisure and household production, then uniformity becomes optimal.\\n\\n> 21 Crawford et al. (2010) suggest the latter result may reflect the use of leisure goods as substitutes for time spent producing relaxation, as per the arguments in Kleven (2004) and Boadway and Gahvari (2006).\\n\\n9\", \"datasets\": [{\"mention_name\": {\"text\": \"Canadian data\", \"confidence\": 0.5128687620162964, \"start\": 1566, \"end\": 1579, \"clean_text\": \"Canadian data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1566, \"end\": 1579, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1566, \"end\": 1579, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1566, \"end\": 1579}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1566, \"end\": 1579, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6091452836990356, \"start\": 736, \"end\": 749}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5128687620162964, \"start\": 736, \"end\": 749}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5356441736221313, \"start\": 736, \"end\": 749}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5356441736221313, \"start\": 736, \"end\": 749}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"Pirttila and Suoniemi (2014), estimate a model using commodity demands and additional administrative data to explain hours worked for Finland. Their results also reject weak separability, although they find statistically significant relationships between consumption and hours worked for a smaller number of consumption categories than Crawford et al. (2010). Expenditure on housing and on books and magazines are found to be negatively related to hours worked, while office meals and car use are positively related to hours worked.<sup>22</sup> They also examine the use of public services and find that the use of childcare is positively related to hours of work.\\n\\nWhile these results suggest an efficiency case for differential taxation, the practical implementation of such an optimal rate structure is challenging. Crawford et al. (2010), for example, emphasize that optimal tax rates will depend not just on the sign of demand responses, but also on their magnitudes as well as on cross-price effects. They conclude that “the limitations and uncertainties of both the theory and empirical work are such that, at least as yet, they provide little firm basis for policy prescription” (p350). They further note that, given the small size of their complementarity estimates, the social gain from implementing differential rates is likely to be small, and these would need to be weighed against the administrative burden (see section 4.4) of implementing a highly differentiated rate structure.\\n\\nThe Mirrlees Review (Mirrlees et al., 2011) to which Crawford et al. (2010) contribute, consequently recommends the adoption of a uniform VAT rate structure for the United Kingdom, with the possible exception of childcare to which a zero rate could be applied.<sup>23</sup> More generally, Sørensen (2007, 2010) argues that a lack of elasticity information, combined with administrative and political economy considerations, suggests that “uniform taxation should be the main guideline for indirect taxation” (Sørensen, 2010; p241). Nevertheless, he also adds that, following the recent home production literature, there is a good case for applying relatively low indirect tax rates to substitutes for home production activities .\\n\\nIn summary, consideration of the above literature provides support for a uniform VAT rate in two ways. First, and most clearly, the optimal indirect tax literature provides the strong message that rate differentiation should not be used to achieve distributional goals when other more direct instruments are available. Second, while the optimal indirect tax literature does provide a potential efficiency rationale for differentiated rates, such an approach is not practically implementable due to the extremely high data requirements. Therefore, from an efficiency perspective, we fall back on a single rate structure to avoid distorting consumption decisions (while also reaping the administrative benefits of a single-rate structure, as discussed subsequently in this paper). The one potential exception on efficiency grounds is regarding substitutes for home production activities – although, as discussed below, this also has practical limitations.\\n\\n22 Additionally, they find capital income to be significantly negatively related to hours of work. Gordon and Kopczuk (2014) make a similar finding based on United States data. Both papers argue therefore that capital income should be taxed. They also argue that current direct tax concessions (e.g. mortgage interest deductibility) for housing should be removed.\\n\\n23 Bastani et al. (2015) argue that Mirrlees et al. (2011) are effectively considering a situation where all goods are separable from leisure with the exception of one good needed for work. They develop a model to show that in such a case there remains an equity argument to apply differential tax rates to the goods that are separable from leisure. The rationale is that, because a family needs to pay for childcare in order to work, a high ability “mimicking” family (that works fewer hours than a non-mimicking family) actually has higher disposable income than a non-mimicking family earning the same gross income because they need to pay for less childcare. Their higher disposable income creates an equity-based rationale for taxing goods consumed proportionately more by higher income households. If childcare is publicly provided, this rationale disappears and uniform taxation is once again optimal.\\n\\n10\", \"datasets\": [{\"mention_name\": {\"text\": \"administrative data\", \"confidence\": 0.8541895151138306, \"start\": 86, \"end\": 105, \"clean_text\": \"administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 86, \"end\": 105, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 86, \"end\": 105, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 86, \"end\": 105}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 86, \"end\": 105, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.5159614682197571, \"start\": 86, \"end\": 105}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8541895151138306, \"start\": 86, \"end\": 105}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6318966746330261, \"start\": 86, \"end\": 105}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6318966746330261, \"start\": 86, \"end\": 105}}, {\"mention_name\": {\"text\": \"United States data\", \"confidence\": 0.3210594654083252, \"start\": 3341, \"end\": 3359, \"clean_text\": \"United States data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3341, \"end\": 3359, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3341, \"end\": 3359, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3341, \"end\": 3359}, \"geography\": {\"text\": \"United States\", \"confidence\": 0.4761580228805542, \"start\": 3341, \"end\": 3354, \"clean_text\": \"United States\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.4630969762802124, \"start\": 887, \"end\": 905}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3210594654083252, \"start\": 887, \"end\": 905}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4485407769680023, \"start\": 887, \"end\": 905}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4485407769680023, \"start\": 887, \"end\": 905}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"# 4 The case for reduced VAT rates\\n\\nWith the above lessons from the optimal indirect tax literature in mind, this section now briefly assesses the main arguments that have been used in practice both for and against the adoption of reduced VAT rates. As already noted, the primary rationale for the adoption of reduced VAT rates has been to achieve distributional goals. That said, merit good and efficiency-based arguments are also used to justify the introduction of reduced rates. Meanwhile, a number of practical arguments favor simplification and the use of a single-rate VAT structure.\\n\\n## 4.1 Redistribution\\n\\nThe discussion above has shown that the theoretical case for using reduced VAT rates to achieve distributional goals is very weak, particularly when other more direct instruments are available. In spite of this, redistribution has been the predominant motivation for reduced VAT rates in OECD countries. On this basis, reduced rates are frequently applied to consumption items that tend to make up a greater proportion of the expenditure of poorer as compared to richer households. As shown in Section 2, these include food, water supply, pharmaceuticals, refuse, sewage, energy and children’s clothing.\\n\\nDistributional arguments are generally premised on the view that the VAT is regressive, and hence measures are necessary to reduce the negative distributional consequences of the VAT. This is problematic for several reasons: first, it is by no means clear that the VAT is indeed regressive; second, it is unlikely that reduced VAT rates will be a well targeted instrument to support the poor; and third, as the theoretical literature emphasizes, other more direct instruments – if available – are likely to be better targeted. In this regard, it is important to remember that it is the progressivity of the overall tax-benefit system (together with expenditure programs) that matters, not one particular component. As such, distributional concerns regarding the VAT do not need to be addressed directly through the VAT system itself.\\n\\nThe distributional impact of a VAT system, including who benefits from reduced VAT rates, will in practice depend on both the design of the VAT system and the consumption patterns of households (including their responses to price changes). As such, whether the VAT is regressive, whether reduced VAT rates are well targeted at poorer households, and whether alternative instruments can better target poorer households are all empirical questions that can be tested. A number of empirical studies have examined these questions, and are summarized briefly below. These studies typically draw on household expenditure survey microdata,<sup>24</sup> as this enables the fine distinctions present in many countries between expenditure categories subject to different VAT rates to be accurately modelled.\\n\\n### 4.1.1 Regressivity of the VAT\\n\\nSeveral studies (e.g., Leahy et al., 2011; Ruiz and Trannoy, 2008; O’Donoghue et al., 2004) examine the overall distributional impact of the VAT by measuring VAT burdens as a percentage of income across the income distribution for a single year. Following this approach, these studies find the VAT to be a highly regressive tax. However, as has been highlighted by various authors (e.g., IFS, 2011a; Creedy, 1998; Metcalf, 1994), a major problem with this approach is that it does not account for savings behavior. More specifically, it ignores the fact that income saved in the current year will incur VAT when it is eventually\\n\\n> 24 Warren (2008) provides a summary of different approaches that can be taken to modelling the distributional effects of consumption taxes more broadly.\\n\\n11\", \"datasets\": [{\"mention_name\": {\"text\": \"household expenditure survey microdata\", \"confidence\": 0.663538932800293, \"start\": 2648, \"end\": 2686, \"clean_text\": \"household expenditure survey microdata\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2648, \"end\": 2686, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2648, \"end\": 2686, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2648, \"end\": 2686}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2648, \"end\": 2686, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"microdata\", \"confidence\": 0.9992128610610962, \"start\": 964, \"end\": 1002}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.663538932800293, \"start\": 964, \"end\": 1002}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9844030141830444, \"start\": 964, \"end\": 1002}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9844030141830444, \"start\": 964, \"end\": 1002}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"consumed (because this VAT cannot be captured by an analysis based on data from a single year).<sup>25</sup> Because savings rates tend to increase with income, this biases income-based VAT burden results downwards at higher income levels – driving the finding that the VAT is regressive.\\n\\nIn contrast, studies that measure VAT burdens as a proportion of expenditure (across either the income or expenditure distribution) tend to find that VAT systems are relatively proportional, or even slightly progressive (see, e.g., Thomas, 2022a; Bird and Smart, 2016; IFS, 2011a; Metcalf, 1994). The largest crosscountry study favoring the expenditure-based approach is Thomas (2022a), who finds the VAT to be either roughly proportional or slightly progressive in 23 of 27 OECD countries examined. However, results for four countries show that broad-based VAT systems with very few reduced VAT rates or exemptions can produce a small degree of regressivity.\\n\\nMost recently, Bachas et al. (2023) have considered the impact of informality in developing countries on the distribution of VAT burdens. They estimate informality Engel curves (relating the informal budget share to log total expenditure) for 32 low and middle-income countries. They find that informality decreases as household expenditure increases, and conclude that this will create a degree of progressivity in the VAT in developing countries. Similarly, Jenkins et al. (2006) find that informality creates progressivity in the VAT in the Dominican Republic.\\n\\n# 4.1.2 The ability of reduced VAT rates to target poorer households\\n\\nA number of empirical studies examine who benefits from reduced rates, finding the VAT to be a poorly targeted tool for supporting poorer households. For example, two large scale studies consider who benefits from reduced VAT rates in nine European (IFS, 2011a) and 20 OECD countries (OECD/KIPF, 2014). Both studies find that reduced VAT rates as a whole have a progressive impact, but that richer households benefit more in aggregate terms than poorer households. IFS (2011a) emphasize a key reason for the overall progressive results is the application of reduced rates to the majority of food in all nine countries. OECD/KIPF (2014) finds significant variation in the distributional impact of reduced rates across expenditure types. Reduced rates typically introduced to achieve distributional goals – such as on food, water supply and energy products – are found to have a small progressive effect, but to be poorly targeted. Reduced rates typically introduced to address social, cultural and other non-distributional goals – such as reduced rates on books, restaurant food and hotel accommodation – are often found to be so poorly targeted that they have a regressive effect.<sup>26</sup> Meanwhile, Warwick et al. (2022) examine the overall distributional impact of reduced VAT rates in six developing countries (Ethiopia, Ghana, Senegal, Sri Lanka, Uzbekistan and Zambia). Additionally, they utilize input-output tables to estimate tax expenditures resulting from VAT exemptions. They find both progressive and regressive results, depending on the country, but that reduced rates and exemptions give significant benefit to richer households and are therefore an expensive way of attempting to support poorer households.<sup>27</sup> The impact of informality in developing countries illustrated by Bachas et al. (2023)\\n\\n> 25 Similarly, current expenditure, and the VAT incurred on it, may have been funded from income earned in a previous year. See Thomas (2022a) for a more detailed discussion.\\n\\n> 26 A range of other single-country microsimulation studies find similar results either for reduced VAT rates as a whole or for selected reduced rates such as on food (e.g., Davis and Kay, 1985, for the United Kingdom; Leahy et al., 2011, for Ireland; Caspersen and Metcalf, 1994, for the United States; Creedy, 2001, for Australia; Ball et al., 2016, for New Zealand; Cseres-Gergely et al., 2017, for Hungary; and Gaarder, 2018, for Norway). See Thomas (2020) for a detailed review of this literature.\\n\\n> 27 Warwick et al. (2022) assume a constant degree of informality across households in their base scenario. In sensitivity analysis for Senegal, they vary the degree of informality across households and products, based on place of purchase data (as used by Bachas et al., 2023). This has only a small impact on their results, and their conclusions\\n\\n12\", \"datasets\": [{\"mention_name\": {\"text\": \"data from a single year\", \"confidence\": 0.4935585558414459, \"start\": 70, \"end\": 93, \"clean_text\": \"data from a single year\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 70, \"end\": 93, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 70, \"end\": 93, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 70, \"end\": 93}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 70, \"end\": 93, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6237924695014954, \"start\": 70, \"end\": 93}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.4935585558414459, \"start\": 70, \"end\": 93}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7504631280899048, \"start\": 70, \"end\": 93}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7504631280899048, \"start\": 70, \"end\": 93}}, {\"mention_name\": {\"text\": \"input-output tables\", \"confidence\": 0.40418994426727295, \"start\": 2993, \"end\": 3012, \"clean_text\": \"input-output tables\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2993, \"end\": 3012, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2993, \"end\": 3012, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2993, \"end\": 3012}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2993, \"end\": 3012, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9986304044723511, \"start\": 568, \"end\": 587}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.40418994426727295, \"start\": 568, \"end\": 587}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5908657312393188, \"start\": 568, \"end\": 587}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5908657312393188, \"start\": 568, \"end\": 587}}, {\"mention_name\": {\"text\": \"place of purchase data\", \"confidence\": 0.7777988910675049, \"start\": 4318, \"end\": 4340, \"clean_text\": \"place of purchase data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4318, \"end\": 4340, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4318, \"end\": 4340, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4318, \"end\": 4340}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4318, \"end\": 4340, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.2922683656215668, \"start\": 375, \"end\": 397}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7777988910675049, \"start\": 375, \"end\": 397}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8995412588119507, \"start\": 375, \"end\": 397}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8995412588119507, \"start\": 375, \"end\": 397}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"may further weaken the ability of reduced VAT rates to target poor households. In particular, Bachas et al. (2023) find (very small) positive slopes on formal food Engel curves in some low-income countries – implying that formal budget shares on food increase with income, and hence that even a reduced rate on food could be slightly regressive in some low-income countries.<sup>28, 29</sup> A standard assumption in the above studies is that the VAT is fully passed on to consumers in prices (and so the savings from a reduced VAT rate are also passed on to the consumer). However, the theoretical and empirical literature suggests that this may not necessarily be the case, casting further doubt on the ability of reduced VAT rates to provide support to poorer households.<sup>30</sup> For example, a detailed review by IHS (2011) finds a wide range of empirical results in the literature, covering full, less than full, and more than full pass-through. They conclude that full pass-through is more likely to be found in more competitive markets and for broader VAT reforms. More recently, Benzarti et al. (2020) find evidence for European countries of stronger pass-through of VAT increases than VAT decreases. Benedek et al. (2019) find roughly full pass-through of standard VAT rate changes, but only around 30 percent pass-through for changes in reduced VAT rates. Unlike Benzarti et al. (2020), they find no significant evidence of asymmetric responses to price changes in European countries. Meanwhile, Gaarder (2018) finds that the introduction of a reduced VAT rate on food in Norway resulted in full pass-through to prices.\\n\\n# 4.1.3 Reduced VAT rates vs cash transfers\\n\\nThe above studies, while showing that reduced VAT rates often have a progressive effect, also show them to be badly targeted at the poor. Nevertheless, a case only exists for removing existing reduced rates if a better alternative policy instrument exists with which to achieve distributional goals. In theory, cash transfer programs (whether implemented through the tax or benefit system) should be able to better target support to poorer households than reduced VAT rates. Not only can they target support based on income, but also on family characteristics, such as number and age of children, to further target support at households most in need. This expectation is supported by several recent studies that show cash transfers to be superior to reduced VAT rates at providing support to poorer households. These include several studies that estimate demand systems to incorporate behavioral responses into the simulation results, as well as simpler non-behavioral studies.\\n\\nFor example, Cseres-Gergely et al. (2017) estimate a Quadratic Almost Ideal Demand System (QUAIDS) for Hungary to simulate the introduction of a range of potential policy measures to help poorer households, finding that an income transfer to the unemployed would be better targeted at the poor than reduced VAT rates on food. Thomas (2022b) uses a QUAIDS model for New Zealand to simulate a revenue neutral reform replacing existing targeted family tax credits with reduced VAT rates on food, beverages, and a remain the same, in part because their data suggests only limited variation in informality across the income distribution in Senegal (88% informality in the bottom decile vs 83% informality in the top decile).\\n\\n> 28 The slopes of formal food Engel curves tend to become negative for upper middle-income countries in their sample. As would be expected, the slopes of total (i.e. formal plus informal) food Engel curves are found to be negative in all countries covered.\\n\\n> 29 Furthermore, if a zero rate was applied to formal food consumption, a situation could arise where low-income informal consumers effectively pay a positive rate of VAT on their food purchases, but high-income formal consumers pay no VAT due to the zero rate. The positive VAT on informal consumption would arise where informal suppliers of food purchase inputs from formal VAT registered businesses. These VAT registered businesses would charge VAT on their output, but this would not be able to be reclaimed by the informal supplier, who may therefore pass this VAT cost on to their (informal) consumer in the price of the food.\\n\\n> 30 This point is emphasized by de la Feria and Walpole (2020).\\n\\n13\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"range of recreational and cultural expenditure. He finds that these reduced VAT rates would be far less effective at supporting poorer households than the existing targeted family tax credits.\\n\\nVan Oordt (2018) uses a QUAIDS model for South Africa to simulate removing reduced VAT rates and using the revenue to extend existing cash transfers. He finds that low- and middle-income households would benefit from the reform, while high-income households would lose. However, he highlights some concerns regarding the ability to effectively implement cash transfers in a developing country context. In this regard, Gcabo et al. (2019), in their non-behavioral analysis, find that better targeting would be achieved with the introduction of new cash transfers rather than the extension of existing cash transfers in South Africa.\\n\\nWarwick et al. (2022) find that the existing cash transfer programs in their six developing countries are better targeted at poor households than reduced VAT rates and exemptions. That said, the targeting of these programs remains imperfect, and would require significant improvements in both generosity and coverage in order to compensate for base-broadening VAT reforms. Strikingly, their non-behavioral simulation results show that even adopting a universal benefit would better target poor households than the reduced VAT rates (and exemptions) in place in these countries.<sup>31</sup> In addition to redistribution (which addresses vertical equity concerns), there are also horizontal equity implications to the adoption of reduced VAT rates. As noted by IFS (2011b), multi-rate VAT systems will effectively reward some households for their preferences and penalize others. This will breach horizontal equity as, for example, two otherwise identical households with different consumption preferences for reduced vs standard-rated goods and services, will face different VAT burdens. In contrast, a single-rate VAT system would result in the same tax burden for both households, irrespective of their consumption preferences.<sup>32</sup> Finally, the application of reduced VAT rates for distributional purposes may conflict with other policy objectives. In particular, applying reduced VAT rates on fossil fuel energy use (e.g. heating fuels) will conflict with environmental goals to efficiently price carbon emissions (and internalize other negative externalities, e.g., local air pollution). A preferable approach is to fully internalize the negative externalities in the price through excise and/or carbon taxes, apply the standard VAT rate on top, and then provide compensation to poorer households through targeted cash transfer programs.<sup>33</sup> More generally, with any good producing negative externalities (e.g. tobacco, alcohol), the standard VAT rate should be imposed on top of the corrective excise tax used to internalize the externality.\\n\\n> 31 Non-behavioral studies for the United Kingdom by Davis and Kay (1985), Crawford et al. (2010) and Mirrlees et al. (2011) find reform packages involving the replacement of reduced VAT rates with changes to direct taxes and transfer payments can improve distributional outcomes. Similarly, Brashares et al. (1988) find that income-tested credits or reimbursements would benefit poor households far more than zero-rating necessities in a VAT in the United States. Again, see Thomas (2020) for a more detailed review of this literature.\\n\\n> 32 IFS (2011b) also discuss the specific egalitarianism argument of Tobin (1970). This argument supposes that society has preferences regarding inequality in the consumption of specific products – such as necessities – as opposed to over total consumption. However, as they note, this argument assumes that households would buy too little of the product if they were provided with the money to buy them (e.g. through a tax credit or benefit). As such, it is more in the nature of a merit good or internality argument, rather than an equity-based argument.\\n\\n33 The same rationale also applies regarding temporary measures introduced in response to the recent surge in food and energy prices, with targeted cash transfers (whether implemented via the tax or benefit system) being preferable to reduced VAT rates or other price-based support measures (see, e.g., Amaglobeli et al., 2022; Van Dender et al., 2022).\\n\\n14\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"# 4.2 Merit goods/externalities\\n\\nAnother common argument for the introduction of reduced VAT rates is to encourage consumption of ‘merit goods’ that generate positive externalities. That is, if consumption of a good or service has benefits to society that the consumer does not take into account in their consumption decision, then there is _prima facie_ a market-failure based case for government intervention to encourage consumption up to the socially optimal level. (Equally, there is a case to discourage consumption of goods and services with negative external effects as noted above).<sup>34</sup> On this basis, as noted in Section 2, many OECD countries have introduced reduced VAT rates on a range of goods and services with perceived cultural or social merit to encourage their consumption. These include: books, newspapers, cinema, theatre, concerts, museums, zoos, amusement parks, and sporting events and facilities. Reduced rates are also applied on environmental externality grounds in some countries to solar panels, insulation, and other environmentally beneficial goods and services. Reduced rates applied to passenger transport may in some cases also be based on environmental externality grounds. Total expenditure on such goods and services typically makes up a far smaller proportion of total household expenditure than on those introduced for distributional reasons, due, in particular, to the large budget share of food.\\n\\nWhile the basic rationale to internalize a positive externality is valid, a key question arises, as it did with redistribution, regarding whether a reduced VAT rate is the most effective way of achieving the desired policy goal.<sup>35</sup> Various authors have suggested this is unlikely to be the case (see, for example, IFS, 2011b; IHS et al., 2015; Abramovsky et al., 2017). IFS (2011b), for example, argue that “[i]f the social problem one wishes to address is affected by business use of a product, or is associated with the consumption of only particular kinds of consumers (e.g. the poor or children), or is unrelated to the price of the product, then applying reduced rates may not be an appropriate policy response” (p554).\\n\\nThe first of these points relates to the ability of businesses to claim input tax credits, meaning that a reduced VAT rate will not lower the price a business pays for a product. As such, where business consumption of a product (e.g. passenger transport, insulation or solar panels) produces positive externalities, then reduced rates – which only incentivize final consumers – will not be well targeted. Second, as reduced VAT rates apply equally to all consumers, they will not be well targeted if underconsumption is specific to a subset of the population – for example, under-consumption of books by young people or poor people.\\n\\nPerhaps most significantly, reduced VAT rates provide a larger subsidy for more expensive purchases. This implies that, for a reduced rate to be well targeted, the positive externality should be correlated with the price of the product subject to the reduced rate. However, in many cases this will not be true. For example, reading an inexpensive paperback book will not provide less social benefit than reading an expensive hardback version of the book. Similarly, taking a taxi is unlikely to provide a greater reduction in environmental\\n\\n> 34 Similar arguments may be made regarding “internalities” where consumption may provide additional benefits to the consumer that they do not fully take into account due, for example, to a lack of knowledge. Negative internality arguments are often used in relation to proposals for health-related excise taxes (see, for example, Marron, 2015). Negative externalities are, of course, a common justification for health-related excise taxes on alcohol and tobacco. 35 Another question that may be asked is whether or not the tax-favored activity really is socially desirable or not. However, as IFS (2011b) note, what is socially desirable is a largely subjective question, and may depend significantly on the culture and history of the particular country. Estimation of the size of the positive externality is also a substantial challenge (as it is for negative externalities). Further difficulty then arises in accurately estimating the behavioral response to a price change, in order to determine the necessary size of the tax concession to fully internalize the externality.\\n\\n15\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"cost than taking a bus for the same journey (assuming an internal combustion engine in both cases). IFS (2011b) suggest better targeted mechanisms are likely to be available in many cases. For example, income-based or age-based subsidies are likely to better target concerns about under-consumption of certain products by young or poor people. Subsidies can also be made available to businesses.\\n\\nApplying a reduced VAT rate for social or cultural purposes may also have broader distributional effects. For example, certain cultural activities may be consumed disproportionately by better-off households, who would then benefit disproportionately from a reduced rate on these activities (OECD/KIPF, 2014).\\n\\nRecently, ‘circular economy’-based arguments have also been put forward suggesting the use of differentiated VAT rates to encourage greater use of recyclable and re-usable goods. (See, for example, European Circular Economy Stakeholder Platform, 2021).<sup>36</sup> While circular economy arguments go beyond the externality-based rationale discussed above, if pursued as a policy goal, the same question remains as to whether the VAT is the most effective way of achieving such a goal. In this regard, the ability of businesses to claim input tax credits, and the provision of a larger subsidy for more expensive purchases (rather than larger resource quantity), again suggest that alternative options may be more effective at achieving those objectives, and with less negative distributional impact.\\n\\n# 4.3 Efficiency\\n\\nAs Mirrlees et al. (2011) note, there is an initial presumption on efficiency grounds in favor of uniformity of indirect taxes to avoid distorting consumption decisions. The optimal tax literature discussed above then points to a theoretical case for differentiated rates on efficiency grounds, but not one in practice due to the lack of reliable information. As such, efficiency arguments point broadly towards a single-rate VAT structure.\\n\\nNevertheless, following Kleven et al. (2000) and subsequent literature, there is a restricted case for applying reduced VAT rates to substitutes for home production activities. This provides support for the reduced rates in place in a number of countries for services such as domestic cleaning, repair services and hairdressing. That said, an assessment by the European Commission (2003) of reduced rates introduced in nine member countries on such labor-intensive services concluded that the reduced rates did not appear to have a positive impact on employment (or on the informal economy). More generally, they concluded that “[c]ompared with other measures, particularly those that directly target labour costs, the budgetary cost of any job creation effects through VAT reductions is always high.” Another argument used to justify applying reduced VAT rates to labor-intensive sectors – such as restaurants and hotels – is to reduce structural unemployment of low-skilled workers (Copenhagen Economics, 2007; IFS, 2011b, IHS et al., 2015). Applying a reduced VAT rate to such sectors that primarily hire low-skilled workers is intended to boost demand for low-skilled workers, and thereby reduce structural unemployment that may have arisen due to factors such as restrictive labor market regulations, high minimum wages, and high non-wage labor costs.<sup>37</sup>\\n\\n> 36 Existing reduced rates in several OECD countries for repair services could be justified on this basis, although they appear to have typically been introduced on efficiency grounds as a substitute for home production (IFS, 2011b). 37 Copenhagen Economics (2007) and IHS et al. (2015) emphasize that this argument also requires disproportionate structural unemployment of low-skilled workers as compared to higher skilled workers. If not, any demand-induced reduction in structural unemployment of low-skilled workers would be countered by a demand-induced increase in structural unemployment of higher skilled workers. They also note that a reduced VAT rate would be unlikely to reduce structural unemployment if it was applied to tradeable goods and services, as part of the impact would be to increase imports.\\n\\n16\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"Little empirical evidence is available on the effectiveness of such measures. However, a study by Copenhagen Economics (2007) suggests that reduced VAT rates are unlikely to be an effective mechanism to increase low-skilled employment. Using a general equilibrium model for the European economy as a whole, they find that reduced VAT rates can lower structural unemployment in sectors such as hotels and restaurants, without significantly reducing employment in other sectors. However, they find that this provides only a very limited boost to overall employment of low-skilled workers, as such sectors only employ a small fraction of all low-skilled workers. Furthermore, they find that applying the standard VAT rate to all sectors currently benefiting from reduced rates would be likely to create a similar boost in demand for low-skilled workers, but without distorting consumption decisions. They note that direct subsidies may be a better-targeted and more transparent approach, with lower compliance costs. IFS (2011b), meanwhile, argue that alternative mechanisms such as active labor market policies, employment regulation reform, and education investment would be likely to be better targeted ways of addressing structural unemployment of low-skilled workers.\\n\\nFinally, applying a reduced VAT rate to increase employment is also likely to have broader distributional effects. In particular, restaurant food and hotel accommodation are likely to be consumed disproportionately by better-off households, who would then benefit disproportionately from a reduced rate on these activities (OECD/KIPF, 2014).\\n\\n# 4.4 Practical implementation issues\\n\\nFinally, the application of a multi-rate VAT system increases the complexity of the system, thereby increasing administrative and compliance costs, creating opportunities for tax avoidance and evasion, as well as increasing susceptibility to lobbying.\\n\\nA large number of studies point to the increased compliance and administration costs associated with multi-rate VAT structures (e.g., Ebrill et al., 2001; Copenhagen Economics, 2007; IFS, 2011b; IHS et al., 2015; Abramovsky et al., 2017). In particular, the use of multiple rates requires accounting, invoicing and tax reporting systems to separately record purchases and sales involving different VAT rates. In contrast, a single VAT rate enables simpler reporting systems with less possibility of error. Ebrill et al. (2001) argue further that simple records, invoices and tax forms will in turn support the effective operation of self-assessment systems as well as more effective taxpayer education and staff training. A single rate will also aid audit activity, as verification is not required of the breakdown of purchases and sales between different rates. A single rate will also reduce scope for disputes over classifications (and limit fraud by reducing possibilities for deliberate misclassification). A single rate will also reduce the number of refunds that tax administrations must process.\\n\\nWhile there is significant empirical evidence on the overall compliance and administrative costs of the VAT, there is unfortunately only limited evidence specifically on the impact of rate structure. An early study by Sandford et al. (1981) found that the average compliance costs of firms in the United Kingdom with output subject to multiple rates were more than twice those of firms with output only subject to a single rate. Copenhagen Economics (2007), meanwhile, point to practical evidence from Sweden showing roughly 20 percent of all VAT disputes to be linked to arguments about whether a particular product should be subjected to a low or high VAT rate. Abramovsky et al. (2017) provide examples of court cases from the United Kingdom involving disputes over classifications. These include cases considering whether Pringles should be classified as potato crisps (standard-rated) or savory snacks (zero-rated), and whether Jaffa Cakes are chocolate-covered biscuits (standard-rated) or chocolate cakes (zero-rated). Meanwhile, for a sample of OECD countries, Agha and Haughton (1996), find that VAT compliance decreases significantly the greater the number of VAT rates – suggesting these problems manifest in lower compliance.\\n\\n17\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"Finally, the application of reduced VAT rates to some products increases vulnerability to lobby group pressure to apply reduced rates to additional products. IFS (2011b) argue this is particularly likely for substitutes for goods already subject to reduced rates, or if the rationale for the existing concessions can be argued as also applicable to other goods. Mirrlees et al. (2011), for example, suggest that lobby group pressure has influenced decisions in the United Kingdom to extend VAT concessions. Drawing on the New Zealand experience, Benge et al. (2013) argue that it is “much easier to deny special treatment in all cases rather than allow certain special cases but not others”, and that “once certain special cases are allowed, decisions on whether or not to allow others are much more likely to be driven by lobbying and political realities than unbiased econometric analysis” (p496).\\n\\n# 5 Conclusion\\n\\nThis paper has investigated the case for a multi-rate VAT structure. The paper has highlighted the widespread use of reduced VAT rates in OECD countries, with these typically being aimed at achieving distributional goals, and (to a lesser extent) social, cultural and other objectives. However, analysis of both the theoretical and empirical evidence, as well as practical considerations, suggests that the case for applying reduced VAT rates is weak. In particular, the optimal indirect tax literature finds no redistributive role for reduced VAT rates when other more direct instruments are available. These theoretical findings are supported by the empirical literature that shows reduced VAT rates to be a poorly targeted means of supporting lower income households, and the greater – though still imperfect – targeting ability of cash transfer programs. Similarly, reduced VAT rates are unlikely to be a well-targeted way to encourage consumption of merit goods. Meanwhile, efficiency arguments, with the restricted exception of some substitutes for home production, favor the adoption of a single-rate VAT structure, which would also bring significant administrative benefits.\\n\\nWhile focusing predominantly on OECD countries, the paper’s findings have significant implications for tax reform in both developed and developing economies. In particular, where countries have the administrative capacity to effectively implement targeted cash transfer programs, they should use these programs to support poorer households instead of using reduced VAT rates. In more limited capacity environments, a simpler universal cash transfer could even be adopted – as this would still better target support to poorer households than reduced VAT rates. Only where there is no administrative capacity to effectively implement a cash transfer program should reduced VAT rates be considered as a tool to target support to poorer households. Even then, the potential impact of informality on the ability to target poorer households should be considered.\\n\\nCountries should also consider removing reduced VAT rates aimed at non-distributional goals where a more effective instrument is available to achieve the particular policy goal. At a minimum, the merits of these reduced VAT rates should be reassessed in light of their negative distributional impact.\\n\\nFrom a political economy perspective, countries that face political difficulty removing reduced VAT rates should, in the short-term, consider removing at least those that have regressive effects. In the long-term, they should still consider removing all reduced rates and adopting more targeted policy solutions with less negative distributional consequences. Not only would a move to a single-rate VAT system improve distributional outcomes, it would also simplify the VAT system – thereby reducing administrative and compliance costs, and reducing opportunities for abuse. The clear “packaging” of the increases in targeted support together with the removal of reduced VAT rates in any public communication is likely to increase public support for the reform.\\n\\n18\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"Finally, countries that currently do not employ reduced rates in their VAT system should be very cautious regarding any proposals to introduce reduced rates. This is particularly important in light of the potential political economy difficulties of removing them once they are in place.\\n\\n19\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099920201182414099/pdf/IDU12d02d19f1336a14e0c1b53714d134b306a9f.pdf\", \"pages\": [20]}}]}"}}},{"rowIdx":83,"cells":{"doc_id":{"kind":"string","value":"34243167"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34243167\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"authors\": {\"0\": {\"author\": \"Canavire Bacarreza,Gustavo Javier\"}, \"1\": {\"author\": \"Gómez Aliaga,Guillermo\"}, \"2\": {\"author\": \"Britton,Chevanne Regina Lygia\"}, \"3\": {\"author\": \"Rios Avila,Fernando\"}, \"4\": {\"author\": \"Jimenez Pozo,Wilson Leonardo\"}, \"5\": {\"author\": \"Granados Ibarra,Silvia Juliana\"}, \"6\": {\"author\": \"Li,Ran\"}}, \"count\": \"Grenada\", \"count_key\": \"82618\", \"docna\": {\"0\": {\"docna\": \"Fiscal Incidence on the Island: Grenada’s\\n Fiscal System and Its Incidence\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"EFI-LCR-POV-Poverty and Equity (ELCPV)\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34243167\"}, \"repnb\": \"WPS10678\", \"docdt\": \"2024-01-22T00:00:00Z\", \"datestored\": \"2024-01-22T01:15:20Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"This paper examines the distributional\\n effects of fiscal policy in Grenada. Using data from the\\n 2017–18 Living Conditions and Household Budgets\\n Survey and following the Commitment to Equity analysis\\n framework, the paper estimates the effects of fiscal policy\\n interventions on inequality and poverty. It analyzes the\\n distributional incidence of direct and indirect taxes,\\n direct transfers provided by social transfers and school\\n feeding programs, and in-kind transfers generated by public\\n services in health and education. The results show that\\n Grenada has a tax system that is neutral on the value-added\\n tax side and progressive on the personal income tax side.\\n Furthermore, direct transfers make a modest contribution to\\n poverty reduction and are almost neutral in their\\n distributive impact. The results contribute to the\\n understanding of who bears the burden of taxation and\\n benefits from transfers and of how Grenada’s\\n fiscal system can improve its redistributive effect.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10678; PROSPERITY\", \"display_title\": \"Fiscal Incidence on the Island:\\n Grenada’s Fiscal System and Its Incidence\", \"disclosure_date\": \"2024-01-22T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099315201222443333/text/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.txt\", \"listing_relative_url\": \"/research/2024/01/34243167/fiscal-incidence-island-grenada’s-fiscal-system-incidence\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099315201222443333/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f\", \"new_url\": \"2024/01/34243167/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f\", \"ext_pub_date\": \"2024-01-22T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Public Finance m1326315 713,Fiscal Incidence Analysis m1326940 711,Conditional Cash Transfers(CCT) m1327202 761,Growth, Poverty and Inequality m1327633 1091,Labor Taxes m1504100 2786\", \"docm_id\": \"IDU-a5a6417d-a77d-4ba5-b095-8a83e3fc689f\", \"chronical_docm_id\": \"IDU-a5a6417d-a77d-4ba5-b095-8a83e3fc689f\", \"publishtoextweb_dt\": \"2024-01-22T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10678\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"guid\": \"099315201222443333\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"32\", \"author_keys\": {\"0\": {\"author_key\": \"715971\"}, \"1\": {\"author_key\": \"NA\"}, \"2\": {\"author_key\": \"2136902\"}, \"3\": {\"author_key\": \"1974997\"}, \"4\": {\"author_key\": \"1987580\"}, \"5\": {\"author_key\": \"2128001\"}, \"6\": {\"author_key\": \"1678637\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099315201222443333/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10678 Fiscal Incidence on the Island Grenada’s Fiscal System and Its Incidence _Gustavo Canavire-Bacarreza Guillermo Gómez Aliaga Chevanne Britton Fernando Rios-Avila Wilson Jimenez Pozo Silvia Granados Ibarra Ran Li_ Poverty and Equity Global Practice & Macroeconomics, Trade and Investment Global Practice January 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10678\\n\\n# **Abstract**\\n\\nThis paper examines the distributional effects of fiscal policy in Grenada. Using data from the 2017–18 Living Conditions and Household Budgets Survey and following the Commitment to Equity analysis framework, the paper estimates the effects of fiscal policy interventions on inequality and poverty. It analyzes the distributional incidence of direct and indirect taxes, direct transfers provided by social transfers and school feeding programs, and in-kind transfers generated by public services in health and education. The results show that Grenada has a tax system that is neutral on the value-added tax side and progressive on the personal income tax side. Furthermore, direct transfers make a modest contribution to poverty reduction and are almost neutral in their distributive impact. The results contribute to the understanding of who bears the burden of taxation and benefits from transfers and of how Grenada’s fiscal system can improve its redistributive effect.\\n\\nThis paper is a product of the Poverty and Equity Global Practice and the Macroeconomics, Trade and Investment Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at gcanavire@worldbank.org; ranli@worldbank.org; Chevanne.britton@gmail.com; friosa@gmail.com; wjimenezpozo@gmail.com; sgranadosibarra@worldbank.org; and ggomezaliaga@worldbank.org.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [{\"mention_name\": {\"text\": \"Living Conditions and Household Budgets Survey\", \"confidence\": 0.8718152642250061, \"start\": 157, \"end\": 203, \"clean_text\": \"Living Conditions and Household Budgets Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 157, \"end\": 203, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 157, \"end\": 203, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"2017–18\", \"confidence\": 0.5897348523139954, \"start\": 149, \"end\": 156}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 157, \"end\": 203, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.7252625823020935, \"start\": 157, \"end\": 203}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8718152642250061, \"start\": 157, \"end\": 203}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8801406025886536, \"start\": 157, \"end\": 203}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8801406025886536, \"start\": 157, \"end\": 203}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# **Fiscal Incidence on the Island: Grenada’s Fiscal System and Its Incidence**<sup>*****</sup>\\n\\nGustavo Canavire-Bacarreza<sup>†</sup> Guillermo Gómez Aliaga<sup>‡</sup> Chevanne Britton<sup>§</sup> Fernando Rios-Avila<sup>**</sup> Wilson Jimenez Pozo<sup>††</sup> Silvia Granados Ibarra<sup>‡‡</sup> Ran Li<sup>§§</sup>\\n\\n**Keywords:** Fiscal incidence, Poverty, Inequality, Taxes, Social Transfers, Public Expenditure, Public Revenue **JEL CODES:** D31 **,** H11 **,** H22, H5, I14, I24, I3, O54\\n\\n> * The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the World Bank, the Executive Directors of the World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The authors would like to thank Carlos Rodríguez-Castelan, Gabriela Inchauste, Cesar Cancho and Gabriel Lara for helpful comments and suggestions.\\n\\n> † The World Bank, Poverty and Equity Global Practice; gcanavire@worldbank.org.\\n\\n> ‡ The World Bank, Poverty and Equity Global Practice; <u>ggomezaliaga@worldbank.org.</u>\\n\\n> § Grenada’s Ministry of Finance; <u>chevanne.britton@gmail.com.</u>\\n\\n> ** Levy Institute at Bard College; <u>f.rios.a@gmail.com.</u>\\n\\n> †† Fundacion Aru; wjimenezpozo@gmail.com.\\n\\n> ‡‡ The World Bank, Poverty and Equity Global Practice; sgranadosibarra@worldbank.org.\\n\\n> §§ The World Bank, Macroeconomic and Trade Global Practice; ranli@worldbank.org.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **1. Introduction**\\n\\nBecause of their idiosyncratic economic characteristics, small island developing states (SIDS) face greater challenges that elevate the costs associated with delivering public services, including education, health, and social programs. The direct impacts of these challenges have repercussions on public finances, pushing governments to assess and prioritize sectors in determining where to focus their investments (OECD 2018). This problem is aggravated by the vulnerability of SIDS to shocks that limit their domestic revenue generation, largely due to their dependence on tourism activities (OECD 2018). In this context, a fiscal incidence analysis becomes of prime importance, as Lustig (2020) points out.\\n\\nThis paper focuses on Grenada, an SIDS with a population of around 113,000, that is highly dependent on tourism. As a result, the country is vulnerable to climatic shocks and economic downturns in developed countries. Grenada is at high risk for climate-related disasters, ranking third globally in terms of GDP vulnerability and seventh in terms of fatalities per 100,000 inhabitants due to weather-related losses. Approximately one-fifth of the country’s population lives below the national poverty line (OECD 2018; World Bank 2021; Grenada, Ministry of Foreign Affairs 2022). The COVID-19 pandemic severely impacted Grenada’s economy, particularly its tourism sector, leading to a significant economic contraction and increased unemployment.<sup>1</sup> The government’s response measures contained the initial spread of the virus, but subsequent waves caused further economic challenges. Despite some recovery in 2021, Grenada’s GDP remains below pre-pandemic levels, and the government had to suspend the fiscal rule for three consecutive years to address the pandemic’s extended impacts.<sup>2</sup> The extended impacts of the pandemic have made additional fiscal stimulus necessary (see Appendix A) to support vulnerable groups and maintain economic activities.\\n\\nThe paper examines the redistributive impact of taxes and public spending, following the Commitment to Equity (CEQ) framework (Lustig 2018). The framework aims to provide a comprehensive picture of the redistributive effects of fiscal policies on household income and consumption. Here the CEQ framework is used to assess Grenada’s fiscal policies, focusing on pre and post fiscal incomes in order to determine the net beneficiaries and payers of these policies (Lustig 2020). For the Grenada context, a fiscal incidence analysis through the CEQ framework is relevant, because it can provide evidence concerning the extent to which the government’s fiscal policies are impacting the welfare of the country’s households and inform future policy decisions.\\n\\nThis paper makes multiple contributions to the existing literature. First, we assess the fiscal system of an SIDS, Grenada, which faces resource constraints, natural hazards, and a high poverty rate. These characteristics make it important to determine whether the country’s fiscal system has a redistributive impact. Furthermore, Grenada is located in a region where fiscal incidence is understudied. Previous work on Caribbean islands using the CEQ methodology has focused on bigger and more developed islands such as Jamaica (Katayama et al. 2021),\\n\\n> 1 Before the COVID-19 pandemic, Grenada was making a strong commitment to economic reform and resilience building with the support of international organizations. It achieved an average annual growth of 4.5 percent between 2014 and 2019, surpassing the regional average. Fiscal reforms, including the Fiscal Responsible Act (FRA) of 2015, led to a significant reduction in public debt and the accumulation of a fiscal surplus. Nevertheless, Grenada’s economy remains less diversified and highly susceptible to climate change and natural disasters.\\n\\n> 2 MFMOD Database, the World Bank’s World Development Indicators (WDI), and GEM databases, IMF. Most sources available in https://data.worldbank.org/.\\n\\n2\", \"datasets\": [{\"mention_name\": {\"text\": \"MFMOD Database\", \"confidence\": 0.8515267372131348, \"start\": 3869, \"end\": 3883, \"clean_text\": \"MFMOD Database\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3869, \"end\": 3883, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3869, \"end\": 3883, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3869, \"end\": 3883}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3869, \"end\": 3883, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9577265381813049, \"start\": 502, \"end\": 516}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8515267372131348, \"start\": 502, \"end\": 516}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7392987012863159, \"start\": 502, \"end\": 516}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7392987012863159, \"start\": 502, \"end\": 516}}, {\"mention_name\": {\"text\": \"World Development Indicators\", \"confidence\": 0.5233849287033081, \"start\": 3902, \"end\": 3930, \"clean_text\": \"World Development Indicators\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3902, \"end\": 3930, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3902, \"end\": 3930, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3902, \"end\": 3930}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3902, \"end\": 3930, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9577265381813049, \"start\": 535, \"end\": 563}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5233849287033081, \"start\": 535, \"end\": 563}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7392987012863159, \"start\": 535, \"end\": 563}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7392987012863159, \"start\": 535, \"end\": 563}}, {\"mention_name\": {\"text\": \"GEM databases\", \"confidence\": 0.7454903721809387, \"start\": 3942, \"end\": 3955, \"clean_text\": \"GEM databases\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3942, \"end\": 3955, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3942, \"end\": 3955, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3942, \"end\": 3955}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3942, \"end\": 3955, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9577265381813049, \"start\": 575, \"end\": 588}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7454903721809387, \"start\": 575, \"end\": 588}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7392987012863159, \"start\": 575, \"end\": 588}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7392987012863159, \"start\": 575, \"end\": 588}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"Barbados (García-Peña Bersh 2019), and the Dominican Republic (Aristy-Escuder et al. 2016). Second, we provide a comparison with similar countries in the Caribbean in order to understand the differences between a small developing island and larger counterparts. Finally, we incorporate into the analysis fiscal interventions that have not been previously analyzed, such as the personal income tax (PIT), which was readjusted in 2019; the most important direct transfer to poor and vulnerable households, the Support for Education, Empowerment and Development Programme (SEED); and school transfers.\\n\\nThe results indicate that Grenada’s fiscal system has the potential to decrease poverty rates and inequality in the country, but it shows neutral incidence. Grenada’s fiscal policy design is indeed oriented toward the low-income population. Focusing only on inequality and accounting for all fiscal interventions does decrease inequality, as observed in countries like Jamaica and the Dominican Republic. These results point to the redistributive potential of fiscal policies pursued by the government of Grenada. As expected, higher segments of the income distribution pay more taxes than the low-income population, and furthermore, the middleincome population pays more than they receive, as does the population making up the high part of the income distribution. However, overall indirect taxes tend to be neutral, meaning they affect household income proportionally across the income distribution; these taxes do lower inequality, but also increase poverty rates. While direct taxes (PIT) do not have an effect on poverty, they help to reduce inequality in Grenada, indicating a well-targeted structure. In sum, indirect taxes are neutral, and direct taxes are almost progressive.\\n\\nIn the case of direct transfers, their effect on reducing poverty is higher than their impact on inequality. However, SEED reduces headcount poverty by almost three percent, signaling that the policy is well targeted, but it has a low redistributive effect, even though low-income households receive a higher amount, indicating this program is relatively regressive. In-kind transfers show mixed results: government funding of education benefits the poor, with primary and secondary education benefiting low- and middle-income households, while government funding of tertiary education benefits the higher echelons of the income distribution. Health is clearly regressive, providing a greater benefit to poorer households. However, their impact on monetary poverty is challenging to describe due to the need to monetize these services to estimate their effect.\\n\\nFollowing this introduction, section 2 provides the context of Grenada’s economic performance and poverty and inequality situation, as well as an overview of fiscal revenues and expenditures. The fiscal incidence methodology is described in section 3, while section 4 presents the main results. Section 5 concludes the paper.\\n\\n# **2. Grenada’s Context: Growth, Poverty, and Fiscal Performance**\\n\\nGrenada’s economic growth is heavily affected by external factors, though structural reforms contributed to the country’s sustained positive growth between 2014 and 2019 (figure 13, Appendix A1). Thanks to growth-enhancing reforms that expanded tourism and a benign global economic environment, real output growth averaged 4.5 percent annually between 2014 and 2019. The high growth rate led to a dramatic increase in Grenada’s per capita GDP. By 2019, Grenada’s GDP per capita was only surpassed by those of St. Kitts and Nevis and Antigua and Barbuda (figure 14, Appendix A1). However, the pandemic, compounded by the war in Ukraine, ended the trend of solid growth and led to a sharp economic contraction of 13.8 percent in 2020, followed by a slow recovery in 2021–22. The economic volatility driven by external factors reflects the intrinsic vulnerabilities of Grenada as an SIDS.\\n\\n3\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"In line with the macroeconomic behavior, moderate poverty decreased from 2008 to 2018. However, in the same period extreme poverty increased slightly. As of 2018, Grenada’s poverty rate had decreased by almost 13 percentage points compared to 2008 (the previous year for which data were available). In contrast, extreme poverty in Grenada increased by 1.2 percentage points (figure 1, panel a). Regarding inequality, we observe that during this 10-year period, the Gini coefficient, which measures the inequality of the income distribution, increased from 0.37 to 0.40 (figure 1, panel b). This increase in inequality from 2008 to 2018 may imply some degree of deterioration of conditions for the lower part of the consumption distribution, which is also shown in the increase of the extreme poverty rate during this period.\\n\\n**Figure 1 Poverty and Inequality Measures**\\n\\n<!-- Start of picture text --> (a) Poverty and Indigence (%) (b) Gini Index<br><!-- End of picture text --> <!-- Start of picture text --> 37.7<br>32.1 0.45<br>0.40<br>24.8<br>0.37<br>12.9<br>2.4 3.6<br>1998 2008 2018 1998 2008 2018<br>Extreme poverty Total poverty Gini<br><!-- End of picture text --> _Sources_ : Government of Grenada, Central Statistical Office, Survey of Living Conditions and Household Budget Survey (SLCHBS) 1998, 2007–08, and 2018–19.\\n\\nGrenada’s fiscal position has varied over the years, partly due to natural and external shocks. Following a modest performance in 2009–13, revenues and grant collection shifted to an upward trajectory in 2014, growing by an average of 11.2 percent annually until 2019. Taxes accounted for 83.2 percent of total government revenue between 2009 and 2021 (figure 2). Tax revenue comprises direct and indirect taxes levied on individuals and businesses. The direct taxes are taxes on income and profits and on property, while taxes on domestic goods and services and on international trade and transactions comprise the indirect taxes. For the purposes of this paper, we focus on the PIT under direct taxes and the value-added tax (VAT) and excise tax under indirect taxes, because these have the most significant impact on individuals and households while also being crucial income earners for the government, as shown in figure 3. PIT in 2021 accounted for 53.4 percent of Taxes on Income and Profits and 10.2 percent of tax revenue. The VAT accounted for 83.8 percent of Taxes on Domestic Goods and Services and 35.0 percent of total tax revenue in 2021. Because the VAT is charged to consumers at the point of purchase, it has implications for individual and household consumption. The excise tax is charged at a fixed rate per quantity of specific goods (imported and locally manufactured goods) referred to as “excisable goods” and accounted for 2.8 percent of tax revenue in 2021.\\n\\n4\", \"datasets\": [{\"mention_name\": {\"text\": \"Survey of Living Conditions and Household Budget Survey\", \"confidence\": 0.7949100732803345, \"start\": 1238, \"end\": 1293, \"clean_text\": \"Survey of Living Conditions and Household Budget Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.3106438219547272, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"Central Statistical Office\", \"confidence\": 0.324101984500885, \"start\": 1210, \"end\": 1236, \"clean_text\": \"Central Statistical Office\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1238, \"end\": 1293}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1238, \"end\": 1293, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9946606755256653, \"start\": 470, \"end\": 525}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7949100732803345, \"start\": 470, \"end\": 525}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6612604260444641, \"start\": 470, \"end\": 525}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6612604260444641, \"start\": 470, \"end\": 525}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"**Figure 2 Revenue and Grants Collection**\\n\\n_Source_ : Eastern Caribbean Central Bank Statistics.<sup>3</sup>\\n\\n**Figure 3 Tax Revenue by Tax Type**\\n\\n_Source_ : Eastern Caribbean Central Bank Statistics.<sup>4</sup> Turning to government expenditures, on average, from 2017 to 2021, 32 percent was allocated annually to social programs, including education, health, and public assistance transfers (figure 15, Appendix A2).<sup>5</sup> The most important direct transfer to poor and vulnerable households is done through SEED, which is administered through the Ministry of Social Development. Also important is the School Feeding Programme, which focuses on providing meals to needy students.\\n\\nExpenditures in education increased over the period 2017–21 and accounted for roughly 10 percent of annual government expenditures (figure 16, Appendix A3). Government recurrent expenditure in public education generally is provided in four stages: pre-primary, primary, secondary, and tertiary (including vocational training). Several institutions also provide special education to serve students with disabilities and special needs. The government largely subsidizes education up to the secondary level (including special education). The level of subsidization decreases at the tertiary level, with higher fees required to access training leading\\n\\n> 3 Available in https://www.eccb-centralbank.org/statistics.\\n\\n> 4 Available in https://www.eccb-centralbank.org/statistics.\\n\\n> 5 This is based on data retrieved from the Annual National Budget and excludes debt repayments (both principal and interest). In 2022, the Grenada government instituted a cap on gasoline prices, which constituted an energy subsidy. This measure will be analyzed in a future iteration of the tool.\\n\\n5\", \"datasets\": [{\"mention_name\": {\"text\": \"Eastern Caribbean Central Bank Statistics\", \"confidence\": 0.876897931098938, \"start\": 55, \"end\": 96, \"clean_text\": \"Eastern Caribbean Central Bank Statistics\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 55, \"end\": 96, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 55, \"end\": 96, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 55, \"end\": 96}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 55, \"end\": 96, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.9945932030677795, \"start\": 55, \"end\": 96}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.876897931098938, \"start\": 55, \"end\": 96}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8900628089904785, \"start\": 55, \"end\": 96}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8900628089904785, \"start\": 55, \"end\": 96}}, {\"mention_name\": {\"text\": \"data retrieved from the Annual National Budget\", \"confidence\": 0.5883022546768188, \"start\": 1489, \"end\": 1535, \"clean_text\": \"data retrieved from the Annual National Budget\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1489, \"end\": 1535, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1489, \"end\": 1535, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1489, \"end\": 1535}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1489, \"end\": 1535, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.5041912198066711, \"start\": 740, \"end\": 786}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.5883022546768188, \"start\": 740, \"end\": 786}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5166198015213013, \"start\": 740, \"end\": 786}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5166198015213013, \"start\": 740, \"end\": 786}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"to associate degrees and certifications from the institutions providing this level of schooling. Government expenditure in public education is considered an in-kind transfer because the service can be accessed by all school-age children regardless of socioeconomic background.\\n\\nExpenditures on health also increased in the period 2017–21, accounting for 7 percent of annual government expenditure as of 2018 (figure 16, Appendix A3). In this paper, funds spent by the government in the health sector amount to in-kind transfers to the public because these services are available to all. Citizens, residents, and visitors can freely access the services in the public health system with no differentiation in fees<sup>6</sup> according to income level or socioeconomic class.\\n\\nIn addition to the in-kind transfers in education and health, the government implemented a social protection system within the SEED program. The program targeted school-age children, people with disabilities, the elderly, the chronically ill, pregnant women, lactating mothers, and the adult poor. Beneficiaries are selected via proxy means testing, which uses household characteristics such as age, household size, marital status, assets, disability, and housing conditions as indicators or “proxies” to determine household expenditure or consumption. Cash is transferred to households: in 2017, it was EC$300 per month or EC$3,600 per annum. In 2018, the amount given to elderly households was increased by EC$100 per month, and in 2022, the government raised the amount paid to all beneficiaries by EC$150 per month. Government annual expenditure on SEED increased from EC$14 million in 2017 to approximately EC$18 million in 2021 and is expected to have increased further in 2022. From 2017 to 2021, the government spent an average of EC$2.9 million per year on the school feeding program.\\n\\n# **3. Data Sources**\\n\\nThe primary data source is the Survey of Living Conditions and Household Budgets (SLCHBS) conducted by the Central Statistical Office (CSO), the country’s national statistics office, between April and May of 2018–19. This survey encompassed a sample of approximately 1,689 households. The primary objective of this questionnaire was to collect information related to household socioeconomic and living conditions as well as the household budget in order to construct the consumption variable (as the survey did not collect income information), data that are crucial for measuring poverty in Grenada. It should be noted that this survey did not gather information regarding pensions received or any social security contributions, neither by individuals nor the government. The survey comprised two distinct parts, one focusing on household conditions and another for obtaining individual characteristics of each household member. The person interviewed was one adult member (older than 18 years old), though the survey collected information on every single household member.<sup>7</sup> The use of administrative data on taxable wages for the PIT complements the survey information. The source of this data set was the Inland Revenue Division (IRD), a division within Grenada's Ministry of Finance (MOF). This data set covers wages for different workers and is disaggregated into two parts: the estimation of the share of the wage allocated for pension funds, and the other being the gross income, which is the taxable part of the wage received.\\n\\n> 6 Fees are waived for some services provided to children, the elderly, and the indigent. All fees and co-payments were not included in the analysis given the low amounts reported in the SLCHBS of 2017–18.\\n\\n> 7 To evaluate the accuracy of the data for fiscal interventions, a macroeconomic validation is presented in Appendix B3.\\n\\n6\", \"datasets\": [{\"mention_name\": {\"text\": \"Survey of Living Conditions and Household Budgets\", \"confidence\": 0.832214891910553, \"start\": 1924, \"end\": 1973, \"clean_text\": \"Survey of Living Conditions and Household Budgets\"}, \"acronym\": {\"text\": \"SLCHBS\", \"confidence\": 0.3075333833694458, \"start\": 1975, \"end\": 1981, \"clean_text\": \"SLCHBS\"}, \"producer\": {\"text\": \"Central Statistical Office\", \"confidence\": 0.6656690835952759, \"start\": 2000, \"end\": 2026, \"clean_text\": \"Central Statistical Office\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1924, \"end\": 1973}, \"geography\": {\"text\": \"Grenada\", \"confidence\": 0.6329283118247986, \"start\": 2484, \"end\": 2491, \"clean_text\": \"Grenada\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.989252507686615, \"start\": 273, \"end\": 322}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.832214891910553, \"start\": 273, \"end\": 322}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.999984860420227, \"start\": 273, \"end\": 322}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.999984860420227, \"start\": 273, \"end\": 322}}, {\"mention_name\": {\"text\": \"administrative data on taxable wages\", \"confidence\": 0.6602208018302917, \"start\": 2990, \"end\": 3026, \"clean_text\": \"administrative data on taxable wages\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2990, \"end\": 3026, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2990, \"end\": 3026, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2990, \"end\": 3026}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2990, \"end\": 3026, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9440786242485046, \"start\": 556, \"end\": 592}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6602208018302917, \"start\": 556, \"end\": 592}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.9998421669006348, \"start\": 556, \"end\": 592}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9998421669006348, \"start\": 556, \"end\": 592}}, {\"mention_name\": {\"text\": \"SLCHBS\", \"confidence\": 0.84920334815979, \"start\": 3627, \"end\": 3633, \"clean_text\": \"SLCHBS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.36259955167770386, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3627, \"end\": 3633, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3627, \"end\": 3633}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3627, \"end\": 3633, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.2722231149673462, \"start\": 374, \"end\": 380}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.84920334815979, \"start\": 374, \"end\": 380}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6606495380401611, \"start\": 374, \"end\": 380}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6606495380401611, \"start\": 374, \"end\": 380}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"We utilize this administrative database for two years, 2017 with 7,442 observations and for 2018 with 8,146.\\n\\nBecause income information is not available in the survey, we compare the administrative data, specifically the taxable portion, with the consumption estimation. This allows us to obtain a more comprehensive understanding of the income distribution. In some cases, we complement missing survey information with administrative data. Because this data set is disaggregated by gender and occupational category, both with values for the taxable income, the administrative data prove useful. Additionally, the administrative data enable us to compare the income distribution to ensure data consistency and improve the exercise's macroeconomic validation. The result is an accurate set of information that makes constructing the taxable income distribution possible.\\n\\n# **4. Methodology**\\n\\nThis paper uses the Commitment to Equity (CEQ) methodology to assess the distributional impact of the fiscal system in Grenada. According to Lustig (2018), “The CEQ Assessment is a diagnostic tool that uses fiscal incidence analysis to determine the extent to which fiscal policy reduces inequality and poverty in a particular country” (62). Fiscal redistribution effects refer to the process by which a state collects revenue from individuals and households (through taxation) and allocates this revenue to finance direct transfers, subsidies, and in-kind benefits individuals and households enjoy (Lustig and Higgins 2018). Because this study focuses on the distributional effects on the population, state interventions that affect firms or other privatesector institutions are excluded from the methodology. The CEQ framework aims to answer four main questions: (1) How much income redistribution and poverty reduction is being accomplished through fiscal policy? (2) How equalizing and pro-poor are specific taxes and government spending? (3) How effective are taxes and government spending in reducing inequality and poverty? and (4) What is the impact of fiscal reforms that change a particular tax or benefit's size and/or progressivity? (Lustig 2018).\\n\\nTo calculate the distributional effects of fiscal policy, the CEQ framework constructs different income concepts in sequential stages, starting from pre-fiscal income that includes only private income sources to final income that incorporates the (almost) full set of taxes and government benefits. In summary, the CEQ (Lustig and Higgins 2018) identifies four main concepts of household income (when the household surveys include variables of different sources of income).<sup>8</sup>\\n\\n> 8 _Market income_ includes factor income (all wages and salaries received by workers and capital income), plus private transfers (remittances, private pensions, etc.) plus imputed housing rent, all income before taxes and social security contributions, government transfers and old-age pensions, minus the contribution to social insurance oldage pensions (only in cases where pensions are considered deferred. The contribution to old-age pensions is not subtracted when contributory pensions are financed by the state). _Disposable income_ is market income plus direct cash and near-cash transfers, minus personal income taxes and employee social security contributions (only when contributory pensions are treated as transfers). _Consumable income_ includes indirect subsidies (energy, food, and other general targeted price subsidies) and excludes indirect taxes (VAT, excise taxes, and other indirect taxes). _Final income_ (post-fiscal) is consumable income plus the value of in-kind transfers, such as free or subsidized government services in education and health with copayments and user fees subtracted; it reflects post-fiscal income.\\n\\n7\", \"datasets\": [{\"mention_name\": {\"text\": \"administrative database\", \"confidence\": 0.8073523640632629, \"start\": 16, \"end\": 39, \"clean_text\": \"administrative database\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 16, \"end\": 39, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 16, \"end\": 39, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 16, \"end\": 39}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 16, \"end\": 39, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.8652870059013367, \"start\": 16, \"end\": 39}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8073523640632629, \"start\": 16, \"end\": 39}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7794397473335266, \"start\": 16, \"end\": 39}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7794397473335266, \"start\": 16, \"end\": 39}}, {\"mention_name\": {\"text\": \"administrative data\", \"confidence\": 0.8501173853874207, \"start\": 563, \"end\": 582, \"clean_text\": \"administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 563, \"end\": 582, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 563, \"end\": 582, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 563, \"end\": 582}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 563, \"end\": 582, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9969760179519653, \"start\": 563, \"end\": 582}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8501173853874207, \"start\": 563, \"end\": 582}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.9263186454772949, \"start\": 563, \"end\": 582}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9263186454772949, \"start\": 563, \"end\": 582}}, {\"mention_name\": {\"text\": \"household surveys\", \"confidence\": 0.873302161693573, \"start\": 2560, \"end\": 2577, \"clean_text\": \"household surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2560, \"end\": 2577, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2560, \"end\": 2577, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2560, \"end\": 2577}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2560, \"end\": 2577, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.7849442362785339, \"start\": 800, \"end\": 817}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.873302161693573, \"start\": 800, \"end\": 817}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6349677443504333, \"start\": 800, \"end\": 817}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6349677443504333, \"start\": 800, \"end\": 817}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"For each income concept, the inequality and poverty indexes are calculated and both magnitudes are compared before and after fiscal interventions. In this regard, the CEQ framework considers variants according to the characteristics of each country, harmonizing the aggregates but not the specific items.\\n\\n# **4.1 Income Aggregates Estimation**\\n\\nGrenada’s information system has limitations when it comes to applying the detailed CEQ framework. Certain data characteristics are needed to model fiscal incidence within this framework, such as an adequate household income aggregate. While Grenada has conducted the SLCHBS recently,<sup>9</sup> as well as maintaining administrative records of social programs and the characteristics of the tax system, the module of incomes within the SLCHBS is limited. The survey captures income data in brackets, records few details on nonlabor incomes, and does not have information on pension affiliation or amount received by individuals, so it is not possible to construct net market income plus pensions for this exercise in the Granada context. Thus, the NSO opted for building welfare aggregates from household consumption<sup>10</sup> instead of household income. This poses a challenge for the implementation in this paper because we need household incomes to estimate the fiscal incidence and run simulations. Following (Katayama & al.,2021) for the case of Jamaica, in this paper we assume that current consumption equals disposable income.<sup>11</sup> This assumption is the most important for applying the CEQ methodology to Grenada, in that it simplifies the reconstruction of income aggregates by tracing back to market income definition as our measure of base income (before taxes or fiscal interventions). As is known in the estimation of welfare measures in developing countries, the distribution of household consumption tends to be more concentrated than income. Of course, this characteristic could introduce some bias and underestimate the extremes of the distribution; however, there is no other option in the absence of household income data.\\n\\nOnce household consumption is defined and thus consumable income is available, the rest of the income aggregates established in the CEQ methodology are calculated according to the tax structure, social protection, and social delivery in Grenada. Our point of departure is disposable income, which is reported by the survey and is assumed to equal household consumption. Then, to identify the distribution of gross income and, after that recreate the distribution of the market income, it is necessary to work backward, following the steps shown in Figure 4: _Net market income (NMI):_ This concept (a pretax income) is obtained by adding the amount of personal income tax ( _PIT_ ) to disposable income ( _DI_ ): N NN= DDNN+ PPNNPP\\n\\n> 9 Through the NSO, Grenada implemented SLCHBSs in 2007–08 and 2017–18. The latter (used in this paper) collected information on household characteristics (housing, demographics, education, health status, social protection, and shocks and coping strategies). In addition, this survey focused on household consumption expenditures on food and nonfood items and other transactions such as financial payments, investment acquisitions, fees, and voluntary contributions.\\n\\n> 10 This consumption aggregate is constructed from the following components: (1) expenses and received gifts of food and beverage items, (2) expenses of nonfood articles, (3) value of consumption of durable goods (such as vehicles), and (4) imputed rent of own housing. Household consumption was used as the basis for the official measurement of poverty and inequality in Grenada (World Bank 2021), following the guidelines for measuring welfare aggregates (Deaton and Zaidi 2002).\\n\\n> 11 Other studies that use consumption as primary income are Lara Ibarra et al. (2019) (Arab Republic of Egypt), World Bank (2018), (Kenya), and Majia-Mantilla et al. (2019) (Uganda).\\n\\n8\", \"datasets\": [{\"mention_name\": {\"text\": \"SLCHBS\", \"confidence\": 0.7592219710350037, \"start\": 614, \"end\": 620, \"clean_text\": \"SLCHBS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.34380194544792175, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 614, \"end\": 620, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 614, \"end\": 620}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 614, \"end\": 620, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9693247079849243, \"start\": 614, \"end\": 620}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7592219710350037, \"start\": 614, \"end\": 620}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5433735251426697, \"start\": 614, \"end\": 620}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5433735251426697, \"start\": 614, \"end\": 620}}, {\"mention_name\": {\"text\": \"household income data\", \"confidence\": 0.9029008746147156, \"start\": 2080, \"end\": 2101, \"clean_text\": \"household income data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2080, \"end\": 2101, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2080, \"end\": 2101, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2080, \"end\": 2101}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2080, \"end\": 2101, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.22227002680301666, \"start\": 419, \"end\": 440}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9029008746147156, \"start\": 419, \"end\": 440}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.479783296585083, \"start\": 419, \"end\": 440}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.479783296585083, \"start\": 419, \"end\": 440}}, {\"mention_name\": {\"text\": \"SLCHBSs\", \"confidence\": 0.43487927317619324, \"start\": 2878, \"end\": 2885, \"clean_text\": \"SLCHBSs\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2878, \"end\": 2885, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2878, \"end\": 2885, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2878, \"end\": 2885}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2878, \"end\": 2885, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.2510642111301422, \"start\": 342, \"end\": 349}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.43487927317619324, \"start\": 342, \"end\": 349}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6489195227622986, \"start\": 342, \"end\": 349}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6489195227622986, \"start\": 342, \"end\": 349}}, {\"mention_name\": {\"text\": \"consumption aggregate\", \"confidence\": 0.334253191947937, \"start\": 3316, \"end\": 3337, \"clean_text\": \"consumption aggregate\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3316, \"end\": 3337, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3316, \"end\": 3337, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3316, \"end\": 3337}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3316, \"end\": 3337, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.751491904258728, \"start\": 5, \"end\": 26}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.334253191947937, \"start\": 5, \"end\": 26}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4777575135231018, \"start\": 5, \"end\": 26}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4777575135231018, \"start\": 5, \"end\": 26}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"_Market income (MI)_ : This aggregate is constructed by subtracting direct transfers ( _DTr_ , for example, via SEED and the School Feeding Programme) from _MI_ : N = N Consumable and final income, on the other hand, are constructed forward from NN−DDPPDD _DI_ .\\n\\n_Consumable income_ : This is equal to _DI_ minus indirect taxes, including VAT and Excise taxes plus indirect subsidies ( _IS_ ): 19F<sup>12</sup> _Final income (FI)_ : This aggregate is obtained after including in-kind transfers for inpatient CCNN = DDNN − (VAT + Excise) + NNII.\\n\\n_Final income (FI)_ and outpatient care provided by the health system (hospitals and community health centers) ( _ikTrH_ ), as well as care provided by the educational system at the preschool, primary, secondary, university, and special levels ( _ikTrE_ ): . **Figure 4 CEQ Income Concepts** FFNN= CCNN+ i PPDDii+ i PPDDii <!-- Start of picture text --> PRE- FISCAL<br><!-- End of picture text --> <!-- Start of picture text --> TRANSFERS<br><!-- End of picture text --> <!-- Start of picture text --> MARKET INCOME<br> Wages, salaries, incomes from<br><!-- End of picture text --> <!-- Start of picture text --> TAXES<br><!-- End of picture text --> <!-- Start of picture text --> PLUS Direct<br><!-- End of picture text --> <!-- Start of picture text --> SEED (child, school<br>support and disability<br><!-- End of picture text --> <!-- Start of picture text --> PLUS Indirect<br>Transfers<br><!-- End of picture text --> <!-- Start of picture text --> PLUS Monetized<br>values of main in-<br><!-- End of picture text --> <!-- Start of picture text --> Outpatient and inpatient<br>care, Community health<br><!-- End of picture text --> <!-- Start of picture text --> NET MARKET INCOME<br><!-- End of picture text --> <!-- Start of picture text --> DISPOSABLE INCOME<br><!-- End of picture text --> <!-- Start of picture text --> CONSUMABLE INCOME<br><!-- End of picture text --> <!-- Start of picture text --> FINAL INCOME<br><!-- End of picture text --> <!-- Start of picture text --> MINUS Direct Taxes<br><!-- End of picture text --> <!-- Start of picture text --> Personal income<br><!-- End of picture text --> <!-- Start of picture text --> MINUS Indirect Taxes<br><!-- End of picture text --> <!-- Start of picture text --> VAT, Excise Tax<br><!-- End of picture text --> <!-- Start of picture text --> MINUS Copayments<br>and<br>user fees<br><!-- End of picture text --> _Source_ : Author’s adaptation based on CEQ Framework (Lustig 2018).\\n\\n> 12 This component was considered equal to 0, because the data did not have information to estimate indirect subsidies.\\n\\n9\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"# **4.2 Assumptions and Implementation of the Fiscal Incidence Analysis**\\n\\n# **_PIT paid by households_**\\n\\nWe assume that PIT is only paid on the labor income of the employed population in the formal sector. Because we do not have a direct measure of labor income, we use an approximation based on our definition of market income. The following steps were taken to calculate the PIT payment for formal workers using the SLCHBS from 2018: (1) we identify the disposable income; (2) we obtain the net market income, subtracting direct transfers; (3) we divide the resulting value by the number of employed adults; (4) we calculate the taxes paid based on the PIT rates corresponding to the brackets; and (5) we aggregate again at the household level.\\n\\n# **_VAT paid by households_**\\n\\nThe VAT is levied on all forms of consumer spending—goods and services. The tax is computed on the value of imports and the value added (or markup) that one business charges another or the final consumer when a good or service is provided. The VAT rate is 15 percent on the selling price; however, hotels and dive operations are charged a rate of 10 percent. Some goods and services, for example flour, rice, sugar, and powdered and evaporated milk are not taxed, nor are postage stamps and supply of textbooks as prescribed by the regulations.\\n\\nExpenses associated with cell phone services have a tax rate of 20 percent and water and electricity services are partially subsidized, based on consumption bounds. In the case of electricity, the quantities are obtained from the costs paid by the households and the exogenous estimates of the prices per kwH are obtained from the Grenada authorities. Households who consume less than 99 kwH per month are exempt from paying the VAT, while households who exceed this consumption pay the VAT on the excess over the limit. Similarly, the water consumption limit is 2,900 gallons per month, with households paying the VAT on the excess consumption over the stated limit.\\n\\nBecause the VAT is already incorporated in the price, a pretax rate is first calculated for each E <u>E</u> good using the equation P = and then the rate is applied to the pretax value. v jj 1+rr vv Although the VAT regulations require companies to have a special registration document showing that they offer taxable products for EC$12,000 per year or operate for at least 12 months, this information is not available in the data source consulted. Therefore, it is difficult to know whether households buy in stores exempt from the VAT registration requirement or sell informally. Likewise, the application of the incidence of indirect taxes does not consider cascading effects, given the difficulty of imputing household deductions to reduce the net payment of these taxes. Consequently, there is an overestimation of the effect of the VAT and excise taxes.\\n\\n# **_Transfers received by households_**\\n\\nDirect transfers in the model are mainly associated with the SEED program and the School Feeding Programme. The SEED transfer is identified in the SLCHBS through a direct question to the households. Based on administrative records, we identify the amounts of program transfers to children, youth, and elderly individuals. Finally, we aggregate the amounts at the household level. The SEED amounts are reduced for households to ensure that base/market 10\", \"datasets\": [{\"mention_name\": {\"text\": \"SLCHBS\", \"confidence\": 0.9226505756378174, \"start\": 420, \"end\": 426, \"clean_text\": \"SLCHBS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.338127464056015, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 420, \"end\": 426, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 420, \"end\": 426}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 420, \"end\": 426, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.3237978518009186, \"start\": 420, \"end\": 426}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8663672804832458, \"start\": 420, \"end\": 426}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8132964372634888, \"start\": 420, \"end\": 426}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8132964372634888, \"start\": 420, \"end\": 426}}, {\"mention_name\": {\"text\": \"administrative records\", \"confidence\": 0.6924851536750793, \"start\": 3108, \"end\": 3130, \"clean_text\": \"administrative records\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3108, \"end\": 3130, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3108, \"end\": 3130, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3108, \"end\": 3130}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3108, \"end\": 3130, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9792612195014954, \"start\": 875, \"end\": 897}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6924851536750793, \"start\": 875, \"end\": 897}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7162983417510986, \"start\": 875, \"end\": 897}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7162983417510986, \"start\": 875, \"end\": 897}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"income is never negative so that if the income is less than the transfer, the latter is adjusted to avoid the definition of income being negative.\\n\\n# **_Estimated value of health care and education services received by households_**\\n\\nUsing data from the SLCHBS we identify individuals who used public education and health services and imputed values for the use of these services. Based on the utilization of health services by the individuals who responded to the survey, a distinction is made between those who received inpatient care and those who received outpatient care in hospitals and community health centers. We then found data on average unit costs by type of care and imputed it to the individuals who used the service to proxy for the state’s current expenditure on health services. As for education, we assume that the public education system serves the entire preschool, primary, secondary, and postsecondary school population. The current expenditure on education was derived from the costs per level of education specified by the Grenada government. Once the annual cost of each level of education was obtained, it was imputed to each student in a household according to the level at which they were enrolled.\\n\\nDue to the information limitations on household income structure and national accounts in Grenada, we could only perform a basic application with direct effects, omitting considerations on agents’ behavior. As more information and parameters on fiscal interventions become available, they could be added to the estimates of fiscal incidence.\\n\\n# **5. Main Results**\\n\\nA set of poverty, income distribution, and indicators of progressivity measures were constructed to answer the questions posed above. Next, we present the main results of the simulations.\\n\\n# **5.1. How Does the Fiscal System Affect Poverty?**\\n\\nThe fiscal system has a high potential for poverty reduction in Grenada (table 1), with the highest impact being associated with in-kind transfers. After adjusting the disposable income, the poverty rate in Grenada for 2018 reached 22.77 percent under the definition of consumption per equivalent adult (equivalent to the definition of disposable income in this document), using the national poverty lines as defined by the CSO (EC$6,782 per year per person, 2017–18). Departing from this measure, the model builds all the other income aggregates as described in the previous section. In the absence of fiscal interventions, such as PIT and cash transfers, 25 percent of the population would live in poverty, with an estimated poverty gap of 8.24 percent. Moreover, after accounting for all taxes and direct transfers, the poverty headcount would be 26.35 percent. In sum, four relevant aspects can be noted. First, the direct transfers (SEED and School Feeding Programme) reduce the incidence of poverty by 2.2 percentage points; these are also reflected in other indicators sensitive to changes in income distribution among the poor, such as poverty severity and the Sen and Watts indexes. Second, the PIT does not change the poverty measures, because it is paid by the population with incomes above the national poverty lines. In fact, based on our estimates, only around 10 percent of the employed population has earnings high enough to be subject to the PIT. Third, indirect taxes such as the VAT (and excise taxes) increase poverty by about 3.6 percent; this is explained by their being structured to be neutral taxes.\\n\\n11\", \"datasets\": [{\"mention_name\": {\"text\": \"SLCHBS\", \"confidence\": 0.7862102389335632, \"start\": 254, \"end\": 260, \"clean_text\": \"SLCHBS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.4127049744129181, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 254, \"end\": 260, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 254, \"end\": 260}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 254, \"end\": 260, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.5650547742843628, \"start\": 254, \"end\": 260}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7862102389335632, \"start\": 254, \"end\": 260}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9500045776367188, \"start\": 254, \"end\": 260}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9500045776367188, \"start\": 254, \"end\": 260}}, {\"mention_name\": {\"text\": \"Watts indexes\", \"confidence\": 0.48040249943733215, \"start\": 3015, \"end\": 3028, \"clean_text\": \"Watts indexes\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3015, \"end\": 3028, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3015, \"end\": 3028, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3015, \"end\": 3028}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3015, \"end\": 3028, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.8886744976043701, \"start\": 567, \"end\": 580}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.48040249943733215, \"start\": 567, \"end\": 580}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5472117066383362, \"start\": 567, \"end\": 580}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5472117066383362, \"start\": 567, \"end\": 580}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"**Table 1 Poverty Measures According to CEQ-Income Definitions** **<u>(In percentage)</u>**\\n\\n|**Poverty measures**|**Market**<br>**income**|**Net market**<br>**income**|**Disposable**<br>**income**|**Consumable**<br>**income**|\\n|---|---|---|---|---|\\n|Headcount poverty|25.00|22.77|22.77|26.35|\\n|Poverty gap * 100|8.24|6.73|6.73|7.80|\\n|Poverty severity * 100|4.10|2.97|2.97|3.45|\\n|Watts Index * 100|11.94|9.23|9.23|10.71|\\n|Sen Index * 100|11.56|9.06|9.06|10.54|\\n\\n_Source_ : Authors’ estimation based on SLCHBS 2018.\\n\\nAlthough the net balance of transfers and taxes (direct and indirect) translates into a higher incidence of poverty, the same is not true for the poverty gap (it decreases the poverty gap by about one percentage point) and other poverty measures. All of these show a reduction after the fiscal interventions.\\n\\n# **5.2. How Does the Fiscal System Affect Inequality?**\\n\\nThe fiscal system in Grenada has heterogeneous redistributive characteristics with a significant potential to reduce inequality. The effects of fiscal interventions on income distribution are heterogeneous, as seen in Table 2. First, the direct transfers (SEED and School Feeding Programme) have low redistributive potential; between the market income and net market income, the Gini index change is below one percentage point. The potential of these transfers is better understood when examining the lower section of the income distribution through the ratio P90/P10, which decreases from 7.4 to 6.8. Second, because the PIT is concentrated in the middle- and upper-income distribution, the effect is noticeable only in the P90/P10 ratio (from 6.8 to 6.6) and on the Gini coefficient, which declines from 0.406 to 0.397. Third, though the VAT and excise taxes seem to be very neutral, they, in fact, cause all inequality measures to decline, although not significantly.\\n\\n**Table 2 Inequality Measurements According to CEQ-Income Definitions**\\n\\n|**Inequality**<br>**measures**|**Market**<br>**income**|**Net market**<br>**income**|**Disposable**<br>**income**|**Consumable**<br>**income**|\\n|---|---|---|---|---|\\n|P90/P10 ratio|7.461|6.856|6.639|6.775|\\n|P90/P50 ratio|2.822|2.781|2.701|2.733|\\n|P10/P50 ratio|0.378|0.406|0.407|0.403|\\n|Gini index|0.416|0.406|0.397|0.395|\\n|GE(-1) index|0.300|0.362|0.346|0.340|\\n|GE(0) index|0.304|0.279|0.267|0.264|\\n|GE(1) index|0.302|0.288|0.272|0.270|\\n|GE(2)index|0.411|0.393|0.364|0.360|\\n|Atkinson (0.5) index|0.140|0.132|0.126|0.125|\\n|Atkinson (1) index|0.262|0.244|0.234|0.232|\\n|Atkinson(2)index|1.000|0.420|0.409|0.405|\\n\\n_Source_ : Own calculation based on SLCHBS 2018.\\n\\n12\", \"datasets\": [{\"mention_name\": {\"text\": \"SLCHBS 2018\", \"confidence\": 0.7515371441841125, \"start\": 502, \"end\": 513, \"clean_text\": \"SLCHBS 2018\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 502, \"end\": 513, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 502, \"end\": 513, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 502, \"end\": 513}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 502, \"end\": 513, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.944446861743927, \"start\": 132, \"end\": 143}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7515371441841125, \"start\": 132, \"end\": 143}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7160964012145996, \"start\": 132, \"end\": 143}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7160964012145996, \"start\": 132, \"end\": 143}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"<!-- Start of picture text --> 7000<br>6000<br>‘ 5000<br>3<br>§ 4000<br>a<br>v<br>3000<br>g<br><4 2000<br>1000> - me O £ | U i |<br>1 2 3 4 5 6 7 8 9 10<br>Decile<br>@ All Indirect Taxes Personal Income Tax<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> 800<br>700<br>o 600<br>8 500<br>a<br>co 400<br>2<br>3 300<br>“ 200 | |<br>; | ie<br>1 2 3 4 5 6 7 8 9 10<br>Decil<br>mSEED School feed program<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"<!-- Start of picture text --> 1,000 25%<br>0 = P| 20%<br>-1,000 15%<br>-2,000 10%<br>-3,000 5%<br>-4,000 0%<br>-5,000 -5% | [<br>-6,000 -10%<br>-7,000 -15%<br>1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10<br>All Direct Transfers Mmmm All Taxes 9 == Net Payers All Direct Transfers mmm All Taxes 9 Net Payers<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"<!-- Start of picture text --> 1000 18.00<br>900800 | 14.0016.00<br>3 12.00<br>&8 600 | 2g 10.00<br>S500 Fi<br>8 400 a<br>a 300 6.00 r] = a<br>200 | 4.00 Py =<br>-0 [ | .0.00 i | i l i| =<br>1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 re<br>Decil Decil<br>m Benef. In-patient attendance Benef. Out-patient attendance = Benef. Communitary health m Benet. In-patient attendance Benef. Out-patient attendance = Benef. Communitary health<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> 900 30.00<br>800<br>;00 25.00<br>500¥8 600 22 20.00<br><48.g&& 400300 a§ég 10.0015.00 |<br>5.00<br>0 | | | | | 0.00 a a = |lf FSF<br>1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10<br>Decil Decil<br>m Pre-school attendance Primary school attendance _mTM Secondary school attendance m Pre-school attendance Primary school attendance _TM Secondary school attendance<br>w Tertiary attendance m Special education attendance w Tectiary attendance m Special education attendance<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> 1600 50.00<br>1400 45.00<br>40.00<br>1200<br>Pa 35.00<br>3 1000 5<br>aQ8&SoC g%S95§ 30.0025.00<br>2 oo & 20.00<br>4 15.00<br>400<br>10.00<br>1 2 3 + 5 6 8 9 10 1 2 3 4 5 6 7 8 9 10<br>Decil Decil<br>WAM health attendance = All education attendance WAM health attendance = All education attendance<br><!-- End of picture text -->\", \"datasets\": [{\"mention_name\": {\"text\": \"Pre-school attendance\", \"confidence\": 0.37954679131507874, \"start\": 691, \"end\": 712, \"clean_text\": \"Pre-school attendance\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 691, \"end\": 712, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 691, \"end\": 712, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 691, \"end\": 712}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 691, \"end\": 712, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.21658553183078766, \"start\": 285, \"end\": 306}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.37954679131507874, \"start\": 285, \"end\": 306}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6544097065925598, \"start\": 285, \"end\": 306}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6544097065925598, \"start\": 285, \"end\": 306}}, {\"mention_name\": {\"text\": \"Primary school attendance\", \"confidence\": 0.35284703969955444, \"start\": 796, \"end\": 821, \"clean_text\": \"Primary school attendance\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 796, \"end\": 821, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 796, \"end\": 821, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 796, \"end\": 821}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 796, \"end\": 821, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.21658553183078766, \"start\": 390, \"end\": 415}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.35284703969955444, \"start\": 390, \"end\": 415}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6544097065925598, \"start\": 390, \"end\": 415}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6544097065925598, \"start\": 390, \"end\": 415}}, {\"mention_name\": {\"text\": \"All education attendance\", \"confidence\": 0.40267887711524963, \"start\": 1250, \"end\": 1274, \"clean_text\": \"All education attendance\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1250, \"end\": 1274, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1250, \"end\": 1274, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1250, \"end\": 1274}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1250, \"end\": 1274, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.237912118434906, \"start\": 2, \"end\": 26}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.40267887711524963, \"start\": 2, \"end\": 26}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7950109243392944, \"start\": 2, \"end\": 26}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7950109243392944, \"start\": 2, \"end\": 26}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"<!-- Start of picture text --> Concentration Index - All Interventions<br>Special Education = -0.7)<br>Tertiary Education Ms (0.24<br>Secondary School -0.16<br>Primary School -0.23<br>Pre-School -0.24<br>Education in-kind Transfers -0.15<br>Community Center -0.07<br>Inpatient Care Mmm 0.12<br>Outpatient Care Mm 0.10<br>Health in-kind Transfers mm 0.07<br>In-Kind Transfer -0.04 &<br>VAT 0.42<br>Indirect Taxes 0.42<br>Personal Income Tax (PIT) es (0.87<br>Direct Taxes es 0.87<br>All Taxes 0.52<br>School Feeding Program -0.31<br>SEED -040<br>Transfers -0.39 DE<br>-1.00 -080 -060 -040 -0.20 0.00 020 040 060 0.80 1.00<br>m Concentration Index Baseline<br><!-- End of picture text -->\", \"datasets\": [{\"mention_name\": {\"text\": \"m Concentration Index Baseline\", \"confidence\": 0.3675740659236908, \"start\": 624, \"end\": 654, \"clean_text\": \"m Concentration Index Baseline\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 624, \"end\": 654, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 624, \"end\": 654, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 624, \"end\": 654}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 624, \"end\": 654, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.8307936191558838, \"start\": 169, \"end\": 199}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3675740659236908, \"start\": 169, \"end\": 199}, \"usage_context\": {\"text\": \"background\", \"confidence\": 0.5778519511222839, \"start\": 169, \"end\": 199}, \"is_used\": {\"text\": \"False\", \"confidence\": 0.5778519511222839, \"start\": 169, \"end\": 199}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"# **5.5. How Progressive or Regressive Are Interventions and Overall Fiscal Policy?**\\n\\nGrenada has managed to establish a tax system with progressive direct taxes, neutral indirect taxes, and globally progressive transfers that is redistributive but that has substantial space to increase efficiency in the reduction of inequalities. Calculation of the concentration indexes of fiscal interventions in Grenada and obtaining the revenue concepts yielded the following results, which are presented in table 3. All taxes have a concentration index of 0.522, higher than the Gini index of market income, and a Kakwani index that is positive but close to 0, which defines this intervention as globally progressive. This result is stronger in the PIT case, while the VAT behaves more like a proportional tax. Similarly, when the transfers are analyzed, they show a negative concentration index and a positive Kakwani measure, so that they can be considered globally progressive in absolute terms. Both transfers have this characteristic in such a way that the benefits (per equivalent adult) decrease when (prefiscal) market income increases.\\n\\n**Table** **<u>3 Concentration Index and Progressivity of Fiscal Intervention</u>**\\n\\n||Size<br>(wrt Market<br>income)|Gini index|Concentration<br>index|Kakwani<br>index|Inequality<br>reduction<br>effect|\\n|---|---|---|---|---|---|\\n|**MARKET INCOME**|**1**|**0.416**||||\\n|**DISPOSABLE INCOME**|**0.998**|**0.397**||||\\n|- SEED|0.016||-0.396|0.812|0.013|\\n|- School Feeding Programme|0.001||-0.312|0.729|0.001|\\n|**All direct transfers**|0.017||-0.387|0.803|0.014|\\n|- PIT|0.019||0.865|0.449|0.009|\\n|**CONSUMABLE INCOME**|**0.927**|**0.395**||||\\n|- VAT|0.070||0.422|0.006|0.000|\\n|- Excise Tax|||||0.000|\\n|**All indirect Taxes**|0.070||0.422|0.006|0.000|\\n|**All taxes**|0.089||0.522|0.106|0.010|\\n|**FINAL INCOME**|**1.007**|**0.367**||||\\n|- Benef. Inpatient attendance|0.024||0.121|-0.295|-0.007|\\n|- Benef. Outpatient attendance|0.007||0.104|-0.312|-0.002|\\n|- Benef. Community health|0.010||-0.066|-0.482|-0.005|\\n|**All health attendance**|0.040||0.072|-0.344|-0.014|\\n|- Preschool attendance|0.006||-0.241|-0.657|-0.004|\\n|- Primary school attendance|0.013||-0.230|-0.647|-0.009|\\n|- Secondary school attendance|0.015||-0.157|-0.573|-0.009|\\n|- Tertiary attendance|0.005||0.240|-0.177|-0.001|\\n|- Special education attendance|0.000||-0.788|-1.204|0.000|\\n|**All education attendance**|0.039||-0.147|-0.563|-0.023|\\n|**All in-kind transfers**|0.079||-0.039|-0.455|-0.039|\\n\\n_Source_ : Own calculation based on SLCHBS 2018.\\n\\n# **5.6. What Is the Marginal Contribution to the Redistributive Effect of Fiscal Interventions?**\\n\\nOther than in the area of tertiary education, Grenada’s fiscal policies contribute to the reduction of inequality. In this part of the paper, we use the definitions of the marginal contribution to the redistributive effect of fiscal interventions, mainly those contributions with respect to the relative size of specific items as well as vertical equity, included in the concepts presented in the CEQ framework. Both help with showing which interventions (or groupings of them) are effective in reducing inequality through fiscal policy. This exercise shows that, with the exception of tertiary education, the rest contribute to the reduction of inequalities. When 17\", \"datasets\": [{\"mention_name\": {\"text\": \"SLCHBS 2018\", \"confidence\": 0.8044499158859253, \"start\": 2532, \"end\": 2543, \"clean_text\": \"SLCHBS 2018\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2532, \"end\": 2543, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2532, \"end\": 2543, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2532, \"end\": 2543}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2532, \"end\": 2543, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.43844932317733765, \"start\": 123, \"end\": 134}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8044499158859253, \"start\": 123, \"end\": 134}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7024673223495483, \"start\": 123, \"end\": 134}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7024673223495483, \"start\": 123, \"end\": 134}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"<!-- Start of picture text --> All net in-kind transfers All net in-kind transfers<br>Net education transfers Net education transfers<br>Educational Tranfer Educational Tranfer<br>Edu. Transfer Tertiary [i Edu. Transfer Tertiary as<br>Edu. Transfer Secundary Es Edu. Transfer Secundary |<br>Edu. Transfer Primary Edu. Transfer Primary i]<br>Edu. Transfer Pre-School Edu. Transfer Pre-School a<br>Net health transfers Net health transfers a<br>Health Tranfer Health Tranfer a<br>DiecttTaxss DE Direct Taxes |<br>School Feeding Progra School Feeding Progrma |<br>SEED SEED |<br>0.00 020 040 060 080 100 120 140 160 180 -0.60 -040 -0.20 0.00 020 040 060 080 1.00<br>m Derivative of Marginal Contribution to Redistributive Effect with respect to the relative m Derivative of Marginal Contribution to Vertical Equity with respect to the<br>size of a specific item relative size of a specific item<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"Analyzing the Gini index, this changes by 4.9 percentage points between net market income and postfiscal income, which includes all the interventions analyzed. However, the changes attributed to direct transfers and direct and indirect taxes account for about 2 percentage points, while the remainder is attributed to in-kind transfers associated with education and health services. This demonstrates that the fiscal policy elaborated by the government does have a redistributive effect that decreases inequality.\\n\\nTo deepen our analysis, we calculated the concentration indexes of transfers and taxes in order to determine the progressivity of the interventions. The results are summarized as follows: (1) both the SEED program and the School Feeding Programme mainly target low-income people, which is observed through the negative sign of the concentration coefficient, indicating a negative relationship between income and transfer; (2) the PIT concentration index is very high and positive, which shows that this tax impacts high incomes; (3) as noted above, indirect taxes are associated with a positive concentration index, although significantly lower than in the previous case; and (4) health transfers are almost neutral, even though the middle- and highincome population uses the services. However, health services in community centers show a negative concentration index, indicating that these are preferentially used by the population (on both an inpatient and outpatient basis) in the lower part of the income distribution, as opposed to hospital care.\\n\\nRegarding education services, it should be noted that, except for tertiary education, all levels serve mainly the low- and middle-income population. Tertiary education is available mainly to the high-income population, though education as a whole is progressive.\\n\\nRegarding changes in inequality attributed to fiscal interventions, Grenada presents results in line with what is expected in comparable countries (mainly Central American countries and Jamaica; see table 8 in Appendix E). In Grenada, the Gini index is more sensitive than in Jamaica but less sensitive than in Costa Rica. These results point to the redistributive potential of the fiscal policies pursued by the Government of Grenada.\\n\\nThis analysis has aimed to obtain as complete a picture as possible of the redistributive effect of taxes and expenditures in Grenada. With the support of the World Bank Group, the International Monetary Fund, and other international organizations, Grenada is developing economic reforms to promote growth and support economic activities, especially in the wake of the COVID-19 pandemic, which impacted the economy and the health situation of the population quite negatively.\\n\\n19\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099315201222443333/pdf/IDU1a5a6417d1a77d14ba51b09518a83e3fc689f.pdf\", \"pages\": [20]}}]}"}}},{"rowIdx":84,"cells":{"doc_id":{"kind":"string","value":"34243399"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34243399\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"admreg\": \"Other,Other\", \"admreg_key\": \"912972,912972\", \"authors\": {\"0\": {\"author\": \"Atamanov,Aziz\"}, \"1\": {\"author\": \"Hoogeveen,Johannes G.\"}, \"2\": {\"author\": \"Reese,Benjamin Christopher\"}}, \"count\": \"Uganda,World\", \"count_key\": \"82583,517191\", \"docna\": {\"0\": {\"docna\": \"The Costs Come before the Benefits: Why\\n Donors Should Invest More in Refugee Autonomy in Uganda\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"EFI-Poverty and Equity-GE (EPVGE),EFI-AFR1-POV-Poverty and Equity (EAEPV)\", \"projn\": \"1W-Developing A Methodology For Measuring\\n The Impact Of Hosting, P -- P179836\", \"subsc\": \"Social Protection,Other Education\", \"trustfund\": \"TF0B9573-Developing a methodology for\\n measuring the impact of hosting, protectin\", \"trustfund_key\": \"2216237\", \"prdln\": \"Advisory Services & Analytics\", \"prdln_key\": \"1416466\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34243399\"}, \"repnb\": \"WPS10679\", \"docdt\": \"2024-01-22T00:00:00Z\", \"datestored\": \"2024-01-22T06:25:47Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"When host countries allow refugees to\\n earn income, two main groups benefit: refugees, who become\\n financially autonomous, and international institutions that\\n can reduce the humanitarian aid that would otherwise be\\n needed to support refugees. Uganda is one of the more\\n progressive countries when it comes to promoting the\\n financial independence of refugees and shifting from\\n humanitarian aid to development ways of working. This note\\n considers how successful refugees in Uganda have been in\\n becoming financially independent and estimates how\\n assistance has been saved due to these efforts at economic\\n inclusion. Using the international poverty line of US$2.15\\n in 2017 purchasing power parities to proxy the costs of\\n basic needs, the results suggest that the amount of total\\n aid needed was reduced by almost 45 percent. They also show\\n that many refugees live in poverty, implying that the\\n present combination of aid and work is inadequate to assure\\n a decent standard of living. While more assistance is needed\\n in the short run, reductions in development assistance are\\n feasible but require upfront investments in refugee earning\\n capacity to realize them.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10679; PEOPLE\", \"display_title\": \"The Costs Come before the Benefits:\\n Why Donors Should Invest More in Refugee Autonomy in Uganda\", \"disclosure_date\": \"2024-01-22T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099825401222436103/pdf/IDU16a019d351e7a014fca1a36d11e801ee4a2d0.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099825401222436103/text/IDU16a019d351e7a014fca1a36d11e801ee4a2d0.txt\", \"listing_relative_url\": \"/research/2024/01/34243399/costs-come-before-benefits-donors-invest-more-refugee-autonomy-uganda\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099825401222436103/IDU16a019d351e7a014fca1a36d11e801ee4a2d0\", \"new_url\": \"2024/01/34243399/IDU16a019d351e7a014fca1a36d11e801ee4a2d0\", \"ext_pub_date\": \"2024-01-22T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Voice and Agency m1326001 924,Jobs and Poverty m1327656 701,Humanitarian Aid m1503719 3039,Refugee Workers m1504034 2731\", \"docm_id\": \"IDU-6a019d35-e7a0-4fca-a36d-1e801ee4a2d0\", \"chronical_docm_id\": \"IDU-6a019d35-e7a0-4fca-a36d-1e801ee4a2d0\", \"publishtoextweb_dt\": \"2024-01-22T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10679\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"projectid\": \"P179836\", \"guid\": \"099825401222436103\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"14\", \"author_keys\": {\"0\": {\"author_key\": \"1182029\"}, \"1\": {\"author_key\": \"435354\"}, \"2\": {\"author_key\": \"1865436\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099825401222436103/IDU16a019d351e7a014fca1a36d11e801ee4a2d0\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10679 The Costs Come before the Benefits Why Donors Should Invest More in Refugee Autonomy in Uganda _Aziz Atamanov Johannes Hoogeveen Benjamin Reese_ Poverty and Equity Global Practice January 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099825401222436103/pdf/IDU16a019d351e7a014fca1a36d11e801ee4a2d0.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10679\\n\\n# **Abstract**\\n\\nWhen host countries allow refugees to earn income, two main groups benefit: refugees, who become financially autonomous, and international institutions that can reduce the humanitarian aid that would otherwise be needed to support refugees. Uganda is one of the more progressive countries when it comes to promoting the financial independence of refugees and shifting from humanitarian aid to development ways of working. This note considers how successful refugees in Uganda have been in becoming financially independent and estimates how assistance has been saved due to these efforts at economic inclusion. Using the international poverty line of US$2.15 in 2017 purchasing power parities to proxy the costs of basic needs, the results suggest that the amount of total aid needed was reduced by almost 45 percent. They also show that many refugees live in poverty, implying that the present combination of aid and work is inadequate to assure a decent standard of living. While more assistance is needed in the short run, reductions in development assistance are feasible but require upfront investments in refugee earning capacity to realize them.\\n\\nThis paper is a product of the Poverty and Equity Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at aatamanov@worldbank.org.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099825401222436103/pdf/IDU16a019d351e7a014fca1a36d11e801ee4a2d0.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# The Costs Come before the Benefits<sup>1</sup>\\n\\nWhy Donors Should Invest More in Refugee Autonomy in Uganda Aziz Atamanov, Johannes Hoogeveen and Benjamin Reese* JEL classification: I30, I38, O15 Keywords: Uganda, refugees, aid, self-reliance, inclusion\\n\\n* All authors are with the World Bank. Corresponding author is Aziz Atamanov\\n\\n(aatamanov@worldbank.org).\\n\\n**Acknowledgments: The authors are thankful to Xavier Devictor, R. Mukami Kariuki and Pierella Paci for overall support and guidance.\\n\\n> 1 Dutch expression (de cost gaet voor de baet uyt) displayed on the front of an 18th century former trade information office in Amsterdam conveying the idea that one must make investments first, whether in terms of time, effort, or finances, before one can reap the rewards.\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099825401222436103/pdf/IDU16a019d351e7a014fca1a36d11e801ee4a2d0.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"<!-- Start of picture text --> 2,000 K x<br>3 Lobule =<br>a Oruchinga<br>1,500 K Kiryandongo =<br>wv v Invepi =<br>1,000 K S° 3 RwamwanjaPalabek [—<br>= = Kampala<br>500 K w 8 ” so Kyaka |i 2<br>OK Fr° =5v |ax |Ng * KyangWaliPalorinydRh 2<br>x Ss. & RNG » x Ss ; & Ss < NakivaleBiCibicli meee<br>ss Ss es eS »Re + & S we AGU<br>S OK 50K 100K 150K 200K 250K<br>m Refugees m Asylum-seekers<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099825401222436103/pdf/IDU16a019d351e7a014fca1a36d11e801ee4a2d0.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"<!-- Start of picture text --> 2,000,000<br>1,500,000<br>1,000,000<br>500,000<br>-<br>8s 8s 8 8 SS SB EBSA<br>oN oN SoNSN oN oN °oNN°o °oNN°o fo}N lo}NANfo}<br>lm Refugees under UNHCR's mandate m@ Asylum-seekers<br><!-- End of picture text --> <!-- Start of picture text --> 2,000,000<br>1,500,000<br>1,000,000<br>500,000<br>7<br>o N st oO foe} o N t+ oO foe} Oo N<br>f=} So So So So a a nl a a N N<br>SoN oN SoN oN SoN SoN fo}N SoN io}N lo}N SoN lo}N<br>m Dem. Rep. of the Congo m South Sudan<br>= Somali. m Eritrea.<br>TM Sudan TM Other<br>@ Burundi<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099825401222436103/pdf/IDU16a019d351e7a014fca1a36d11e801ee4a2d0.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"# Data and descriptive statistics\\n\\nPrimary data used in this brief comes from the Uganda Refugee and Host Communities Household Survey 2018 (RHCS). The RHCS is one of the few surveys in the region providing a representative picture of the socio-economic status of the refugee and host community population. The survey is representative of the refugee and host population in the West Nile and Southwest regions, and the city of Kampala. The host population is defined as the native population in districts where refugee settlements are situated. One important advantage of the RHCS is that it uses an identical survey instrument as the Uganda National Household Survey (UNHS). The UNHS is the key survey conducted by the Uganda Bureau of Statistics to measure poverty. Using the same instrument allows measuring poverty in the RHCS in a way comparable to how the UNHS measures official national poverty rates.\\n\\nAccording to the World Bank (2019), despite all economic inclusion efforts and basic service integration, refugees in Uganda were much more likely to be poor than Ugandans in 2018, 46 versus 18 percent using the national poverty line. This implies that almost half of refugees did not have enough resources to satisfy the minimum daily caloric requirements and basic non-food needs – an issue that has become more acute recently as WFP’s general food assistance has had to be prioritized and reduced for many refugees due to financing shortfalls. Using a different daily poverty line, the international poverty line of US$2.15 expressed in 2017 purchasing power parities (PPP), one finds even higher poverty rates: 70 versus 38 percent accordingly. The highest poverty rate among refugees is observed in the West Nile region reaching 58 percent using the official national poverty line and almost 80 percent using the international poverty line. Having a better educated household head is associated with lower poverty among refugees. There is little difference in poverty incidence among refugees who depend on income from economic activities or aid. However, the poverty rate for refugees who depend on remittances is lower.<sup>3</sup> Higher poverty rates among refugees are associated with higher food insecurity and lower asset ownership compared to hosts. About seven out of ten refugee households in Uganda experience severe food insecurity, while for hosts the proportion is five out of ten (World Bank 2019). Food insecurity among refugees in Kampala (57 percent) is lower than in the Southwest (89 percent of households) and West Nile (85 percent). Regarding assets, refugees are less likely to have access to productive assets such as agricultural land and livestock compared to Ugandans. Land ownership and size of land are also significantly lower respectively smaller among refugees than the hosts but with some variation among refugees across settlements and their time of arrival.\\n\\nAccess to basic services among refugees is either on par with hosts or sometimes better according to World Bank (2019). Access to improved water and sanitation is found to be higher among refugee households compared to host households (94 and 39 percent versus 66 and 26 percent accordingly). There are no substantial differences in school enrollment among refugees and hosts with primary enrollment being high, but secondary enrollment being remarkably low for both groups. Elevated costs and language of instruction (which changes from local vernacular to English) are often mentioned as important constraints to staying in school among refugees. There are no major differences in illness prevalence between refugees and hosts, but refugees have slightly better access to health care.\\n\\nRefugees are less likely to be employed than hosts, with agriculture being the largest sector of employment. Only 28 percent of refugees are employed compared to 64 percent among Ugandans. Unemployed refugees are more likely to be young, with low levels of education, who arrived in Uganda less than two years ago. One in five refugee households own a non-agricultural enterprise which is slightly less than among Ugandans. Among the employed, seven out of ten hosts and half of refugees work in the agriculture industry. As mentioned earlier, refugees are less likely to own land than hosts and mostly have user rights instead of owning the land (World Bank 2019).\\n\\nThe COVID-19 shock had a substantial negative impact on refugees who demonstrated a much slower recovery than Ugandans. According to results from phone surveys conducted since onset of the COVID-19 pandemic (Atamanov et\\n\\n> 3 These are mostly Somalian refugees living in Kampala and who are not getting aid.\\n\\n4\", \"datasets\": [{\"mention_name\": {\"text\": \"Uganda Refugee and Host Communities Household Survey 2018\", \"confidence\": 0.8181377649307251, \"start\": 82, \"end\": 139, \"clean_text\": \"Uganda Refugee and Host Communities Household Survey 2018\"}, \"acronym\": {\"text\": \"RHCS\", \"confidence\": 0.5820210576057434, \"start\": 141, \"end\": 145, \"clean_text\": \"RHCS\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 82, \"end\": 139, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"2018\", \"confidence\": 0.4078899621963501, \"start\": 1085, \"end\": 1089}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 82, \"end\": 139, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9962739944458008, \"start\": 82, \"end\": 139}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8181377649307251, \"start\": 82, \"end\": 139}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9999936819076538, \"start\": 82, \"end\": 139}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9999936819076538, \"start\": 82, \"end\": 139}}, {\"mention_name\": {\"text\": \"Uganda National Household Survey\", \"confidence\": 0.8376170992851257, \"start\": 635, \"end\": 667, \"clean_text\": \"Uganda National Household Survey\"}, \"acronym\": {\"text\": \"UNHS\", \"confidence\": 0.7016875743865967, \"start\": 669, \"end\": 673, \"clean_text\": \"UNHS\"}, \"producer\": {\"text\": \"Uganda Bureau of Statistics\", \"confidence\": 0.6737700700759888, \"start\": 720, \"end\": 747, \"clean_text\": \"Uganda Bureau of Statistics\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 635, \"end\": 667}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 635, \"end\": 667, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9972503781318665, \"start\": 635, \"end\": 667}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8376170992851257, \"start\": 635, \"end\": 667}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9681218266487122, \"start\": 635, \"end\": 667}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9681218266487122, \"start\": 635, \"end\": 667}}, {\"mention_name\": {\"text\": \"phone surveys\", \"confidence\": 0.6140731573104858, \"start\": 4510, \"end\": 4523, \"clean_text\": \"phone surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4510, \"end\": 4523, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4510, \"end\": 4523, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4510, \"end\": 4523}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4510, \"end\": 4523, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9995753169059753, \"start\": 256, \"end\": 269}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6140731573104858, \"start\": 256, \"end\": 269}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8007429242134094, \"start\": 256, \"end\": 269}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8007429242134094, \"start\": 256, \"end\": 269}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099825401222436103/pdf/IDU16a019d351e7a014fca1a36d11e801ee4a2d0.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"<!-- Start of picture text --> 100%<br>Ranking ofper capita<br>consumption from<br>lowest to highest<br>Consumption<br>en ee Poverty Line<br>XN X<br>Basic needs taking into,<br>0% \\\\<br>2 5<br>8 § &<br>Seu9 Oo 3=_ Full Cost of Basic Needs<br>ogoa<o @ labor/other .income @ aid. =—é Poorest Househokis Richest<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099825401222436103/pdf/IDU16a019d351e7a014fca1a36d11e801ee4a2d0.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"<!-- Start of picture text --> 61%<br>55%<br>x 47%<br>c 44%<br>2<br>rom<br>—<br>aa<br>5oO 29%<br>&<br>Qa<br>2<br>mo]<br>mPoorest #2 3 =4_— Richest<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099825401222436103/pdf/IDU16a019d351e7a014fca1a36d11e801ee4a2d0.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"<!-- Start of picture text --> vovo 343 we eee nenee<br>Z<br>5 Inclusion<br>N —— dividend<br>°<br>N<br>s one eee _oe<br>iyavo 193<br>Qa<br>£<br>rom<br>oO.S)<br>© 39<br>a<br>3na) ||<br>Scenario 1: no Scenario 2: current Scenario 3: full<br>economic scenario inclusion<br>opportunities<br>250 costs to cover under current<br>scenario<br>=<br>24v 200 =<<br>a<br>Q 150<br>fo}<br>N<br>&<br>2 100<br>5<br>50<br>ie)<br>own use noland <0.05ha>0.05ha work doesnotoneyear two three Ugandan no<br>rights per per work orless years yearsor friend Ugandan<br>capita capita longer friend<br>Land ownership Land size Head of Years living in Uganda _ Social integration<br>household<br>working status<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099825401222436103/pdf/IDU16a019d351e7a014fca1a36d11e801ee4a2d0.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"characteristics discussed earlier enter the regression as explanatory right-hand side variables. We also added a variable measuring the number of mature cattle refugee households own and a dummy capturing if households borrowed or paid back money or borrowed goods. Finally, we also added the dependency ratio to proxy a household’s capability to work. These variables are expected to affect the capacity of refugees to earn income. Finally, we controlled for the region of residence, household size, age, sex, and literacy of the head of the household. Variables with a positive sign are associated with reducing the gap between consumption and the poverty line (lower costs of basic needs), while variables with a negative sign are associated with widening the gap between consumption and the poverty line.\\n\\n|Table 1. Results from OLS regression explaining difference in log ratio of pre-assistance consumption to int<br>line byrefugee household characteristics(positive coefficient means reduction ofpoverty gap)|ernational poverty|\\n|---|---|\\n|**Variables**|**Coefficient**|\\n|Head of household working (base=not working)|0.265***|\\n|**_Household owns at least one plot of land_**||\\n|Household does not have land or have use rights on land|-0.168*|\\n|**_Household does not have land_**||\\n|Household owns less than 0.05 hectare per capita|-0.116|\\n|Household owns more than 0.05 hectare per capita|0.263**|\\n|**_Household lives in Uganda one year or less_**||\\n|Household lives in Uganda 2 years|-0.0762|\\n|Household lives in Uganda longer than 3 years|0.410***|\\n|Household child/children have Ugandan friend|0.233***|\\n|**_One member household_**||\\n|Two household members|-0.533**|\\n|Three household members|-0.537**|\\n|Four and above household members|-0.760***|\\n|Dependency ratio (children and elderly over total household size)|-0.712***|\\n|Head of household is literate|0.05|\\n|Number of cattle owned|0.0195***|\\n|Any member borrowed money or goods or paying back|0.289***|\\n|**_West Nile_**||\\n|Kampala|1.806***|\\n|Southwest|0.259***|\\n|Head of household age|-0.00533*|\\n|Male head of household|-0.00418|\\n|Constant|-0.129|\\n|Observations|691|\\n|R-squared|0.578|\\n|Source: RHCS 2018, WB staff calculations.||\\n|Note: *** significant at 1 percent, ** at 5 percent, * at 1 percent. Regression is run for refugee sample only using<br>population weight.||\\n\\nConsistent with the earlier descriptive analysis, there are several household characteristics which are significantly correlated with the consumption to poverty line ratio. Thus, having an employed household head helps to reduce the gap between household consumption and the poverty line. Having at least 0.05 hectare of land per household member and owning at least one plot helps to reduce the gap. Social integration proxied by refugee children having Ugandan friends also reduces the ratio between pre-assistance consumption to international poverty. Given that integration takes time, staying in Uganda three or more years reduces the poverty gap compared to staying in Uganda only one to two years. Ownership of large cattle and the ability to borrow are also found to be associated with a lower poverty gap. The gender of the head of household does not affect the size of the poverty gap controlling for other characteristics of refugees and their location. Households with larger shares of children or with more elderly tend to have a higher gap, presumably because of a lower capacity to work.\\n\\n8\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099825401222436103/pdf/IDU16a019d351e7a014fca1a36d11e801ee4a2d0.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"<!-- Start of picture text --> 100<br>TM Non-refugee @ Refugee<br>80<br>3v<br>2 60<br>oPaJ<br>xo°<br>% 40<br>x<br>20; o oon 608 hl l |<br>head is Head is % of At least one Use rights on No land Land size Land size 0.05 Any member<br>working literate households _ land plot is land below 0.05 ha ha per capita took or have<br>with owned per capita or above outstanding<br>dependency loan<br>ratio>50%<br>Household characteristics Access and size of land Credit<br>constraint<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099825401222436103/pdf/IDU16a019d351e7a014fca1a36d11e801ee4a2d0.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"food insecurity, children dropping out of school, and negative coping mechanisms like transactional sex and even suicide.\\n\\nIt is straightforward to estimate by how much aid falls short. The post assistance poverty gap of refugees is 28 percent, implying an average annual shortfall in per capita consumption of US$97, or US$145 million per year when applied to all refugees. Rather than offering their full contribution to global burden sharing by providing US$290 million in aid, refugees receive only half, US$145 million.<sup>10</sup> Given the growing global demand and increasing costs of humanitarian refugee assistance, it would be more equitable and financially smarter not only to meet the obligations for burden sharing but also to invest additionally in the productivity of refugees. Development partners needs to find ways to capitalize on Uganda’s progressive refugee policy context to better facilitate ways that refugees can capitalize on economic opportunities and access to jobs. This way, considerable savings can be realized in the future for hosting refugees.\\n\\n# Conclusion\\n\\nUganda has been at the forefront of refugee integration and social inclusion. The country has created a legal framework that allows refugees to be economically active and offers access to land and housing. This approach has been partially successful. Despite having the right to work, and obtaining land, many refugees remain poor and fail to gain economic independence. Too many refugees remain dependent on humanitarian aid.\\n\\nThis note estimates that at present US$290 million is needed to ensure consumption for all refugees in Uganda is brought up to the international poverty line. Refugees do not receive this much, and the note estimates a shortfall of US$145 million in assistance. This shortfall in burden sharing, due to underfunding is borne by refugees who end up in poverty with levels of consumption that fall below the international poverty line with follow on consequences for food security, nutritional status etc.<sup>11</sup> US$290 million is, however, a lot less than the US$515 million which would have been needed, had refugees not had the right to work. The question is whether refugee income could be boosted further and the need for aid reduced?\\n\\nA comparison between refugees and host communities suggests there is potential for refugees to do better. Poverty incidence and the poverty gap for hosts are much lower than for refugees. It is beyond this paper to establish how to increase the productive capacity of refugees. Yet lower ownership of land and its small size, along with limited access to credit are obvious obstacles to productivity which have been found to increase the poverty gap of refugees. Others (Bohnet and Schmitz-Pranghe 2019) have pointed to limited access to agricultural inputs (seed, fertilizer, irrigation) and limited access to markets as obstacles for refugees to reach their productive potential.\\n\\nThere is much to be said to invest in increasing the income earning capacity of refugees through more development focused investments by the international community. When poor refugees earn income, there are two key beneficiaries. Refugee households, who gain dignity and financial autonomy. And the international community, which can provide less humanitarian aid. This note has provided an estimate of the money currently saved by refugees earning their own income: US$225 million. With an eye on the future, it would be worthwhile to invest some of the current savings to increase the productive capacity of refugees.\\n\\n> 10 It is tempting to compare the estimates of aid needed and aid received by refugees with the numbers available in the account of humanitarian agencies. Unfortunately, without additional analysis it is impossible to make direct comparisons. Not all resources provided by humanitarian agencies will be received by eligible beneficiaries. All assistance programs experience ’leakage’, so some of the spending on basic needs assistance reported by humanitarian agencies is not picked up as having been received by poor households. Second, the assessment is complicated because reporting by humanitarian agencies mixes direct benefits for households (like food assistance), with cost-of-service provision for services that are then provided free of charge (e.g., for health), with investments in e.g. shelter, which are then made available free of charge to refugees.\\n\\n> 11 See UNHCR 2023a.\\n\\n10\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099825401222436103/pdf/IDU16a019d351e7a014fca1a36d11e801ee4a2d0.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"By allowing refugees to participate economically, the Government of Uganda has already created the legal environment which makes it possible to reduce refugee dependence on humanitarian assistance. Thus, Uganda provides an ideal environment to also demonstrate how development investments in refugee earning capacity, can bring about savings in future assistance and provide more long-term and sustainable management models for refugees in the country.\\n\\nNot only does Uganda provide an ideal environment to further promote economic inclusion, as savings over and above the present US$225 million are made in basic needs assistance, the international community might want to consider sharing this ‘inclusion dividend’ with the host country. After all, Uganda’s policies make economic inclusion feasible, so why should the international community reap all its monetary benefits? In a situation where host communities are at times worse off than refugees (e.g. refugees have better access to certain services) it is critical that some of the inclusion dividend is invested in host communities. Only when all parties benefit - refugees, donors and host communities - is continued economic inclusion of refugees sustainable.\\n\\n11\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099825401222436103/pdf/IDU16a019d351e7a014fca1a36d11e801ee4a2d0.pdf\", \"pages\": [12]}}]}"}}},{"rowIdx":85,"cells":{"doc_id":{"kind":"string","value":"34243842"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34243842\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"authors\": {\"0\": {\"author\": \"Ambler,Kate\"}, \"1\": {\"author\": \"Herskowitz,Sylvan Rene\"}, \"2\": {\"author\": \"Maredia,Mywish Karimbhai\"}}, \"count\": \"Malawi\", \"count_key\": \"82524\", \"docna\": {\"0\": {\"docna\": \"Rural Labor and Long Recall Loss\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"DIME Infra & Climate Change (DIME4)\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34243842\"}, \"repnb\": \"WPS10680\", \"docdt\": \"2024-01-23T00:00:00Z\", \"datestored\": \"2024-01-23T01:14:32Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"Surveys frequently rely on annual recall\\n to capture individuals’ labor activities over the\\n preceding year. This paper uses a panel of rural households\\n in Malawi for a survey experiment to test the effect of a\\n long, annual recall window on reported labor supply relative\\n to a set of quarterly interviews. The paper documents large\\n losses in reported labor participation using the long recall\\n window with reductions of over 20 percent of reported\\n activities and months worked and a 2.5 times greater\\n incidence of reported unemployment relative to the shorter\\n window. These losses are greater for activities further in\\n the past and especially for individuals whose labor supply\\n is reported by other family members, reaching up to 50\\n percent for some outcomes. The profile of\\n households’ primary respondents, predominantly\\n male and older, and differential effects by age further\\n suggest that long recall may cause meaningful biases in the\\n resulting data for women and younger household members.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10680; RRR; PEOPLE\", \"display_title\": \"Rural Labor and Long Recall Loss\", \"disclosure_date\": \"2024-01-23T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099314301232420189/text/IDU1f0a4a5fa17a29146e919386190683a48ead2.txt\", \"listing_relative_url\": \"/research/2024/01/34243842/rural-labor-long-recall-loss\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099314301232420189/IDU1f0a4a5fa17a29146e919386190683a48ead2\", \"new_url\": \"2024/01/34243842/IDU1f0a4a5fa17a29146e919386190683a48ead2\", \"ext_pub_date\": \"2024-01-23T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Poverty Measurement and Analysis m1326941 707,Household Surveys m1504071 2761,Labor Diagnostics m1583605 4430\", \"docm_id\": \"IDU-f0a4a5fa-7a29-46e9-9386-90683a48ead2\", \"chronical_docm_id\": \"IDU-f0a4a5fa-7a29-46e9-9386-90683a48ead2\", \"publishtoextweb_dt\": \"2024-01-23T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10680\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"guid\": \"099314301232420189\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"34\", \"author_keys\": {\"0\": {\"author_key\": \"NA\"}, \"1\": {\"author_key\": \"1111373\"}, \"2\": {\"author_key\": \"434944\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099314301232420189/IDU1f0a4a5fa17a29146e919386190683a48ead2\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10680\\n\\n# Rural Labor and Long Recall Loss\\n\\n_Kate Ambler_ _Sylvan René Herskowitz Mywish K. Maredia_ Development Economics Development Impact Group January 2024 A verified reproducibility package for this paper is available at http://reproducibility.worldbank.org, click **here** for direct access.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10680\\n\\n# **Abstract**\\n\\nSurveys frequently rely on annual recall to capture individuals’ labor activities over the preceding year. This paper uses a panel of rural households in Malawi for a survey experiment to test the effect of a long, annual recall window on reported labor supply relative to a set of quarterly interviews. The paper documents large losses in reported labor participation using the long recall window with reductions of over 20 percent of reported activities and months worked and a 2.5 times greater incidence of reported unemployment relative to the shorter window. These losses are greater for activities further in the past and especially for individuals whose labor supply is reported by other family members, reaching up to 50 percent for some outcomes. The profile of households’ primary respondents, predominantly male and older, and differential effects by age further suggest that long recall may cause meaningful biases in the resulting data for women and younger household members.\\n\\nThis paper is a product of the Development Impact Group, Development Economics. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at sherskowitz@worldbank.org. A verified reproducibility package for this paper is available at http:// reproducibility.worldbank.org, click **here** for direct access.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [{\"mention_name\": {\"text\": \"panel of rural households in Malawi\", \"confidence\": 0.5131652355194092, \"start\": 178, \"end\": 213, \"clean_text\": \"panel of rural households in Malawi\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 178, \"end\": 213, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 178, \"end\": 213, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 178, \"end\": 213}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 178, \"end\": 213, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9912132620811462, \"start\": 178, \"end\": 213}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.5131652355194092, \"start\": 178, \"end\": 213}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8324426412582397, \"start\": 178, \"end\": 213}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8324426412582397, \"start\": 178, \"end\": 213}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# **_Rural Labor and Long Recall Loss_**\\n\\nKate Ambler Sylvan René Herskowitz Mywish K. Maredia IFPRI World Bank Michigan State University\\n\\n**Keywords:** Labor, Survey Methodology, Measurement Bias, Recall Windows, Proxy Response **JEL Codes** : O1, J2, C8, Q1\\n\\nThis work was supported by the CGIAR Research Program on Policies, Institutions, and Markets and was approved by the IFPRI Institutional Review Board and the Malawi Committee on Social Science and Humanities, National Commission for Science and Technology. We thank the MwAPTA Institute for collecting the baseline data, GeoPoll for data collection, and Laura Leavens for research support. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **1. Introduction and Background**\\n\\nUnderstanding peoples’ productive activities over the course of the year is important both to document how people earn their livelihoods and also to ensure that poverty programs are well designed and well-targeted. Measuring employment is especially challenging in settings where productive activities are informal, with irregular intensity of participation, and seasonal, where much of the effort and earnings are concentrated in specific periods of the year (Feuerbacher et al. 2020). These characteristics are particularly relevant in rural labor markets in low-income countries that rely heavily on employment in the agriculture sector (Fink, Jack, and Masiye 2020; World Bank, 2017).\\n\\nThe prevailing survey methodology literature considers shorter recall periods to be more accurate, often relying on a standard seven-day recall window (Durazo et al. 2021). However, relying on reported labor activities over the most recent seven days is likely to miss individuals’ full range of labor contributions and work activities over a full year, especially in highly seasonal and informal labor markets. To address this gap, standard data collection practices typically also include questions about productive activities of each household member over the past year.<sup>1</sup> While existing evidence is mixed, use of such a long window may lead to losses in data and data quality.<sup>2</sup> In this paper, we use a survey experiment to test the effect of recall windows on reported annual work behaviors, contrasting reported labor participation using a long (annual) recall method to those resulting from a set of quarterly interviews that follow a similar structure and set of questions, but over shorter, 90-day, recall windows.\\n\\nA commonly used format for annual recall labor questions in multitopic household surveys asks the household’s primary respondent to report on their labor activities over the preceding year, including type of activity, periods worked, and intensity of participation. These questions are asked about their main activity over the year as well as any secondary activity. While there is often a stated preference for having each household member respond for themselves, frequently, either when that respondent is unavailable or when survey budget resources are insufficient to allow for\\n\\n> 1 The most common format for large multi-topic surveys such as the World Bank’s Living Standards Measurement Study is to ask about activities in the last seven days in detail, with follow-up questions regarding the last 12 months (see Durazo et al. 2021).\\n\\n> 2 See de Weerdt, Gibson, and Beegle (2020) for an excellent review of the survey methods literature that includes an overview of both recall windows and proxy reporting.\", \"datasets\": [{\"mention_name\": {\"text\": \"reported labor activities\", \"confidence\": 0.5149959921836853, \"start\": 921, \"end\": 946, \"clean_text\": \"reported labor activities\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 921, \"end\": 946, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 921, \"end\": 946, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 921, \"end\": 946}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 921, \"end\": 946, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"report\", \"confidence\": 0.9493194222450256, \"start\": 31, \"end\": 56}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5149959921836853, \"start\": 31, \"end\": 56}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7881458401679993, \"start\": 31, \"end\": 56}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7881458401679993, \"start\": 31, \"end\": 56}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"self-reporting by all household members, the household’s primary respondent will then report on behalf of other members as well, “proxying” their responses.<sup>3</sup> Recently, the global penetration of cell phone access has introduced phone surveys as a method of data collection that could allow for more frequent interview intervals along with shorter recall periods at a lower cost. While weekly interviews have been shown to be less vulnerable to recall loss than longer windows (Arthi et al. 2018), weekly interviews across an entire year or season would be prohibitively expensive for most research budgets and highly fatiguing for respondents. In this paper we use a survey experiment to test an intermediary option, comparing quarterly (90day) with an annual recall question using a series of phone surveys.\\n\\nLabor outcomes measured using the long recall responses result in much lower recorded levels of any work involvement (22%), number of unique activities reported (24%), and number of months worked (20%) compared to those based on the quarterly measures, although we do not find significant losses in total hours worked. These gaps increase with the amount of time that has passed. Proportional losses from long recall (relative to short recall averages) are, depending on the outcome variable, between two and ten times larger for proxied individuals than for households’ primary respondents who self-report. Splitting the sample by self and proxy reported individuals, we explore heterogeneity by gender and age of the respondent as well as of the proxied household member. We do not see clear patterns of heterogeneity for self-reports or by the primary respondent’s characteristics when reporting for others. We do, however, see that younger household members have greater relative losses from long recall when their responses are proxied than older proxied household members. The profile of primary respondents (generally household heads) further biases annual recall data against women and younger household members, as they will be exposed to larger losses from long recall as a consequence of having their labor contributions more frequently reported by proxy.\\n\\nThis paper contributes to a growing body of work on the effects of recall periods in the survey methodology literature. A considerable body of evidence suggests that longer recall periods tend to undercount household consumption, provision of agricultural inputs, and negative health events\\n\\n> 3 The Living Standards Measurement Study (LSMS) surveys follow a similar structure and have a similarly stated preference for self-reporting whenever possible. In a review of six LSMS surveys, Desiere and Costa (2019) find that the use of proxy reporting is still widespread, ranging between 24% in Nigeria up to 85% in Mali.\", \"datasets\": [{\"mention_name\": {\"text\": \"annual recall data\", \"confidence\": 0.7873232364654541, \"start\": 1977, \"end\": 1995, \"clean_text\": \"annual recall data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1977, \"end\": 1995, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1977, \"end\": 1995, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1977, \"end\": 1995}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1977, \"end\": 1995, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8292874097824097, \"start\": 351, \"end\": 369}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7873232364654541, \"start\": 351, \"end\": 369}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9291054606437683, \"start\": 351, \"end\": 369}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9291054606437683, \"start\": 351, \"end\": 369}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"relative to shorter recall windows.<sup>4</sup> The evidence on recall windows as it applies to labor activities, however, is more mixed. In a study based in Tanzania, Arthi et al. (2018) compare reported contributions of farm labor in an agricultural plot-based module using weekly recall with those from an end of season recall window. Longer recall windows exaggerate the number of hours worked per person by a factor of four, although the reported number of active plots and individual level participation are undercounted. In our rural sample in nearby Malawi, we find reductions in reported labor contributions using the longer recall periods relative to a shorter recall window, but no overall difference in hours worked, the intensive margin on which Arthi et al. (2018) find positive effects. In urban labor markets, Heath et al. (2021) examine very short recall windows and find modest effects of switching from weekly to daily recall with higher reported self- employment spells but no impact on wage employment. However, Garlick et al. (2020) find that weekly versus monthly surveys did not influence data quality or reported microenterprise activities.\\n\\nThis paper also links to a second area of the methods literature centered on the use of proxy responses for data collection. Of note, Bardasi et al. (2011) find that males are more affected by proxy losses in reported agricultural labor activities than women, though this is reduced when the primary respondent is his wife or well-educated. By contrast, Serneels, Beegle, and Dillon (2017) find that proxying does not affect estimates of returns to education. Another related paper in this literature is by Kilic et al. (2022), that also intersects with the recall literature on labor measurement. They use two nationally representative surveys in Malawi conducted in the same year but following different research protocols. They find underreporting of certain labor activities in the survey using “business as usual,” that allowed for proxy reporting and presence of other household members, relative to the other survey that required own response and privacy during the interview. Contrasting their setup, with multiple differences between the “business as usual” and the more stringent protocols, our study uses experimental variation designed to isolate a single mechanism, recall windows. Our heterogeneous results, showing greater long recall loss for\\n\\n> 4 Das and Sanchez-Paramo (2012) find that one-third of acute illness is unreported when comparing one month to one week recall periods. Other notable examples showing high sensitivity of consumption to recall windows include Beegle et al. (2012b), De Weerdt et al. (2016), Backiny-Yetna et al. (2017), and Di Maio and Fiala (2019). Beegle et al. (2012a) find little evidence of distortions from longer recall lags in recording agricultural inputs and production.\", \"datasets\": [{\"mention_name\": {\"text\": \"nationally representative surveys in Malawi\", \"confidence\": 0.6325014233589172, \"start\": 1778, \"end\": 1821, \"clean_text\": \"nationally representative surveys in Malawi\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1778, \"end\": 1821, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1778, \"end\": 1821, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1778, \"end\": 1821}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1778, \"end\": 1821, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9980219602584839, \"start\": 985, \"end\": 1028}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6325014233589172, \"start\": 985, \"end\": 1028}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7412692904472351, \"start\": 985, \"end\": 1028}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7412692904472351, \"start\": 985, \"end\": 1028}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"proxied individuals, are consistent with Kilic et al. as they compare the two survey methods with both annual and weekly recall.\\n\\nFinally, the findings in our paper have implications for a range of labor linked papers, especially those in rural settings that rely on similarly constructed data.<sup>5</sup> Large, primarily descriptive literatures currently exist on rural labor, gender and age differences in labor contributions, and rural income diversification. Our results suggest that all these estimates could be meaningfully impacted by reliance on long recall-based data and further distorted by the use of proxy reporting.\\n\\nWe discuss our survey experiment and data in Section 2 and empirical strategy in Section 3. Results are presented in Section 4, followed by a brief discussion in Section 5, and conclusion in Section 6.\\n\\n# **2. Data**\\n\\nOur survey experiment is built on a sample of households included in a separate study, the MwAPATA Instituteʼs Malawi Rural Agricultural Livelihood Survey (MRALS), conducted inperson in the fourth quarter of 2019. Their original sample is representative of farm households in rural areas of eight selected districts and used a multitopic household survey to collect information on demographics, health, socio-economic status, time use, and agricultural production from a sample of 3,259 households (Muyanga et al. 2020).<sup>6</sup>\\n\\n# _2.1 Sample and survey structure_\\n\\nOur analysis uses a sample of 701 households for whom a complete set of four quarterly phoneinterviews were successfully completed. Starting with the MRALS sample of 3,259 households only 2,435 households had contact phone numbers and were therefore eligible for inclusion in our phone-based survey. Stratifying by region, 1,505 eligible households were randomly selected to be included in our study. 1,020 of these households were successfully contacted and included in Round 1 and ultimately, 701 households were successfully reached for the full set of interviews.\\n\\n> 5 See Dzanku (2020), Abay et al. (2019), Asfaw et al. (2019), Yeboah and Jayne (2018), Imai et al. (2015), Himanshu et al. (2013), Djurfeldt (2013), Haggblade et al. (2010), Ellis (1998), and Ellis and Freeman (2004) for some recent examples.\\n\\n> 6 The selected districts include two in the Northern Region (Rumphi and Mzimba), four in the Central Region (Lilongwe Rural, Dowa, Kasungu, and Mchinji), and two in the Southern Region (Neno and Blantyre Rural).\", \"datasets\": [{\"mention_name\": {\"text\": \"Malawi Rural Agricultural Livelihood Survey\", \"confidence\": 0.6443154215812683, \"start\": 962, \"end\": 1005, \"clean_text\": \"Malawi Rural Agricultural Livelihood Survey\"}, \"acronym\": {\"text\": \"MRALS\", \"confidence\": 0.5337125658988953, \"start\": 1007, \"end\": 1012, \"clean_text\": \"MRALS\"}, \"producer\": {\"text\": \"MwAPATA Instituteʼs\", \"confidence\": 0.6369479298591614, \"start\": 942, \"end\": 961, \"clean_text\": \"MwAPATA Instituteʼs\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 962, \"end\": 1005}, \"geography\": {\"text\": \"Malawi\", \"confidence\": 0.7053338885307312, \"start\": 962, \"end\": 968, \"clean_text\": \"Malawi\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9960289001464844, \"start\": 962, \"end\": 1005}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6443154215812683, \"start\": 962, \"end\": 1005}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4994092881679535, \"start\": 962, \"end\": 1005}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4994092881679535, \"start\": 962, \"end\": 1005}}, {\"mention_name\": {\"text\": \"multitopic household survey\", \"confidence\": 0.5153130888938904, \"start\": 1178, \"end\": 1205, \"clean_text\": \"multitopic household survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1178, \"end\": 1205, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1178, \"end\": 1205, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1178, \"end\": 1205}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1178, \"end\": 1205, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9885643124580383, \"start\": 345, \"end\": 372}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.5153130888938904, \"start\": 345, \"end\": 372}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8889870643615723, \"start\": 345, \"end\": 372}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8889870643615723, \"start\": 345, \"end\": 372}}, {\"mention_name\": {\"text\": \"phone-based survey\", \"confidence\": 0.6249783039093018, \"start\": 1702, \"end\": 1720, \"clean_text\": \"phone-based survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1702, \"end\": 1720, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1702, \"end\": 1720, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1702, \"end\": 1720}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1702, \"end\": 1720, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9957931041717529, \"start\": 90, \"end\": 108}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6249783039093018, \"start\": 90, \"end\": 108}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6445021629333496, \"start\": 90, \"end\": 108}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6445021629333496, \"start\": 90, \"end\": 108}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"Interviews were conducted by phone and the primary respondent was the respondent from the initial MRALS study, typically the household head. The labor module asked this respondent about their own labor participation and additionally to report the labor activities of up to two other adult household members ages 18 to 65. These adults were pre-selected randomly in households with more than three adults. While survey protocols allowed interviews to be done with other household members if the respondent was unavailable, in practice this happened very infrequently.<sup>7</sup> Column 1 of Table 1 shows descriptive statistics of study participants who were successfully reached for the full set of quarterly interviews and who constitute the sample used in the empirical analysis. Household level characteristics, those of the designated respondent, and those of proxied respondents are shown in panels A, B, and C, respectively. Panel A shows that our final sample contains 701 households, with an average of just over five members split evenly between children and adults. In Panel B we see that 75 percent of primary respondents are men and their average age is 44. Sixty-nine percent are in a monogamous marriage while 10% are in polygamous marriages with 6.6 years of education on average. Panel C shows the characteristics of the other household members whose data was collected in the survey. These household members are younger, 30years-old on average, and 76% are female with similar levels of education.<sup>8</sup> Appendix Table 1 shows these summary statics from the analysis sample with those targeted from the sampling of the initial MRALS sample. They are nearly indistinguishable.\\n\\nHouseholds were contacted approximately once per quarter over the course of a year. While the surveys were intended to be spaced evenly, three months between surveys, in practice logistical issues affected this timing and led to some variation in the actual interval between interviews. Dates and timing of each survey wave are illustrated in Figure 1. These quarterly surveys were designed to be brief, averaging under 20 minutes, focused on labor activities of adult household members. The labor module followed a similar structure to that used in many household surveys where respondents are asked about their primary work activity, with a recall period of the preceding 90 days. Follow-up questions capture further details about that activity in particular, which months they spent doing it, and how many days and hours they spent on it in a typical month that they were doing it (in that recall window). After reporting that activity, they were asked if\\n\\n> 7 This occurred in 2.3%, 2.0%, 0.9%, and 0.7% instances across rounds 1 to 4, respectively.\\n\\n> 8 There is slightly higher attrition among this group resulting from a requirement that individuals were living in the household in all four rounds in order to be included in the analysis sample.\", \"datasets\": [{\"mention_name\": {\"text\": \"MRALS study\", \"confidence\": 0.8033003807067871, \"start\": 98, \"end\": 109, \"clean_text\": \"MRALS study\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 98, \"end\": 109, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 98, \"end\": 109, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 98, \"end\": 109}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 98, \"end\": 109, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.7626455426216125, \"start\": 98, \"end\": 109}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8033003807067871, \"start\": 98, \"end\": 109}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9999028444290161, \"start\": 98, \"end\": 109}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9999028444290161, \"start\": 98, \"end\": 109}}, {\"mention_name\": {\"text\": \"MRALS\", \"confidence\": 0.4106280207633972, \"start\": 1651, \"end\": 1656, \"clean_text\": \"MRALS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.36961501836776733, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1651, \"end\": 1656, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1651, \"end\": 1656}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1651, \"end\": 1656, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"microdata\", \"confidence\": 0.35177305340766907, \"start\": 756, \"end\": 761}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4106280207633972, \"start\": 756, \"end\": 761}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4046855866909027, \"start\": 756, \"end\": 761}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4046855866909027, \"start\": 756, \"end\": 761}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"there was a secondary activity that they were engaged in over that same 90-day window and, if so, to provide similar follow-up information.<sup>9</sup> After completing this, the respondent was then asked the same sequence of questions about up to two other adult household members (if available). The interviewers asked to speak to the same primary household respondent in each follow-up round of the survey and asked them about the labor activities of the same family members as those identified and included in the first interview.\\n\\nIn the final interview, respondents were additionally asked to report primary and secondary activities over the past 12 months. Aside from the reference window, the follow-up questions on the activity details were held the same, with the list of months spanning the whole year instead of the last three months. If the sequence of these questions affects respondents’ responses, possibly through either framing effects or fatigue, as has been shown in other studies (e.g., Ambler et al. 2021; Jeong et al. 2023), this could bias the resulting data when using both measures for all individuals in the same order. To address this concern, we therefore randomized the order of the annual recall questions to be either just before or just after the 90-day recall for all family members. In the analysis, we use the reported labor participation of individuals using the recall window which they were asked first, to avoid the possibility of order-induced bias. The analysis then leverages this randomization to test the effect of long versus short recall windows.\\n\\nColumns 3 and 4 of Table 1 show summary statistics and baseline balance between individuals in households who were randomly assigned to have the short or the long recall questions asked first, respectively. The final column shows the p-value of a test for difference between the two groups. We do not find any statistically significant differences between our two randomly assigned groups.\\n\\n# _2.2 Outcome variables_\\n\\nOur analysis focuses on four primary labor supply measures. We examine two extensive margin measures: whether or not the person had any work activities reported about them at all and the number of unique activities they worked; and two intensive margin measures: the number of months worked and the number of hours worked.\\n\\n> 9 Respondents were then asked about their primary and secondary activities over the preceding week and, if distinct from the 90-day activities, details about those activities. Across all rounds, new jobs were reported in the 7-day category fewer than ten times.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"We use the 90-day recall periods to construct quarterly estimates of each of these outcomes. For the intensive margin outcomes, we use the survey date and the variables indicating which months the indicated person worked to calculate the percentage of the month that is applicable for each individual. This is then used to calculate the months and hours worked during that 90-day period.<sup>10</sup> Using the 12-month recall data we create similar quarterly measures to match the dates of the shorter recall.<sup>11</sup> Figure 1 provides a timeline of the four rounds of the surveys, showing the survey dates, the months covered in each recall period, along with the agricultural calendar for maize, the main staple crop in Malawi. We note that there was a delay between the third and fourth (final) survey. This has the effect that some of the time covered in the 90-day recall asked in the Round 1 interview is not well covered by the 12-month recall questions asked at the endline. As such, we observe very low levels of work in Quarter 1 as measured in the 12-month recall data. In particular, we were concerned that including all four rounds of short recall from the four surveys would result in artificially low estimates of work in Quarter 1 for the long recall group. Additionally, these delays resulted in capturing labor data for May and June—the harvest months for the primary agricultural season, in both Round 1 and Round 4.\\n\\nTo address both of these issues, we exclude Quarter 1 from our main analysis and use quarters 2- 4 to improve the comparability of the short and long recall periods. As seen in Figure 1, these three quarters cover the main components of one maize growing season, although there is a small gap between the final two interviews and some seasonal activities just after harvest time may be undercounted and missed in our estimates. This would likely lead to an understatement of long recall losses. We show robustness (and strengthening) of our main results using all four interviews in Appendix Table 1.\\n\\nTo identify the number of jobs done by individuals over the full study period, we linked reported activities across survey rounds, so that if a respondent reported working the same job in all four\\n\\n> 10 This is necessary because for a 90-day recall period, there are 4 calendar months in which the respondent may have worked. To estimate the months worked or hours worked, we must calculate the relevant fraction of the first and fourth month options.\\n\\n> 11 We calculate these quarterly measures by using the question in the 12-month recall module that asks respondents to indicate which months they worked in each job, and then use the survey dates from the quarterly interviews to reflect the same time periods covered by the short recall questions. We use the day of survey for each household, so the period covered varies slightly by household based on the date of interview.\", \"datasets\": [{\"mention_name\": {\"text\": \"12-month recall data\", \"confidence\": 0.6686238646507263, \"start\": 411, \"end\": 431, \"clean_text\": \"12-month recall data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 411, \"end\": 431, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 411, \"end\": 431, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 411, \"end\": 431}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 411, \"end\": 431, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.30705687403678894, \"start\": 411, \"end\": 431}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6686238646507263, \"start\": 411, \"end\": 431}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5348876714706421, \"start\": 411, \"end\": 431}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5348876714706421, \"start\": 411, \"end\": 431}}, {\"mention_name\": {\"text\": \"labor data\", \"confidence\": 0.36782076954841614, \"start\": 1329, \"end\": 1339, \"clean_text\": \"labor data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1329, \"end\": 1339, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1329, \"end\": 1339, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1329, \"end\": 1339}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1329, \"end\": 1339, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.3852542042732239, \"start\": 562, \"end\": 572}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.36782076954841614, \"start\": 562, \"end\": 572}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9874952435493469, \"start\": 562, \"end\": 572}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9874952435493469, \"start\": 562, \"end\": 572}}, {\"mention_name\": {\"text\": \"survey dates from the quarterly interviews\", \"confidence\": 0.32867223024368286, \"start\": 2682, \"end\": 2724, \"clean_text\": \"survey dates from the quarterly interviews\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2682, \"end\": 2724, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2682, \"end\": 2724, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2682, \"end\": 2724}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2682, \"end\": 2724, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9035269618034363, \"start\": 328, \"end\": 370}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.32867223024368286, \"start\": 328, \"end\": 370}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7116365432739258, \"start\": 328, \"end\": 370}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7116365432739258, \"start\": 328, \"end\": 370}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"rounds it is only counted as one unique job.<sup>12</sup> Finally, we sum the quarterly estimates to create aggregate measures comparable to a 9-month recall. The main analysis compares the aggregated estimates collected by quarterly or long recall.\\n\\nIn comparing labor aggregates using the short and long recall windows, two features of the data should be acknowledged. In the short recall estimates, respondents were able to list different jobs in each round so that, if they reported a different primary and secondary activity in each round, they could have up to eight unique activities.<sup>13</sup> Consistent with common survey practices, the long recall questions only allow for a maximum of two work activities. The opportunity to report a greater diversity of work activities may be an advantage of the short recall estimates. In practice, only 2% of individuals reported to have participated in more than two unique activities based on the short recall data and only 25% of individuals reported to have done two unique tasks in the long recall data (the value at which their responses would be censored). While this mechanical difference in the range of possible reported values may influence the analyses on a handful of individuals, the low incidence of cases in which they would be binding makes it unlikely to be driving the results. We return to this later in our discussion of results.\\n\\nA second difference is that when calculating hours worked, we rely on a series of questions that ask respondents which months were worked in that time frame, the number of days in a typical month, and then the number of hours in a typical day. In the short recall responses, the number of days and hours apply to those months worked in that quarter, for long recall those values are applied to all months worked throughout the year. The short recall responses therefore allow for greater variability in recorded labor supply within periods worked for the same activity. We also return to this later in the discussion but generally consider the greater flexibility in reported labor intensity across the year using the short recall window to be an important benefit of higher frequency interviews where differences in intensity are missed when relying on annual recall.\\n\\nThe main 9-month aggregate outcome variables described here are summarized in Table 2, separately by short and long recall group (and anticipating the first, main result). Ninety-one\\n\\n> 12 This was done by preloading the activities reported in earlier survey rounds and asking respondents to indicate when they were referencing the same activity as earlier.\\n\\n> 13 Unique activities are distinct both by type of employment (e.g., self, wage, etc.) and the employer. For example, if a person reported wage employment with entity A and entity B, these would be considered as two different activities.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"percent of those in the short recall group report working at all, in an average of 1.19 activities in 3.6 months for 360 total hours. With the exception of total hours worked these measures are all higher for short recall compared to long recall. Ninety percent of subjects in the short recall group report working on the household farm or in other household agricultural businesses. Much smaller numbers report working in a non-agricultural family business (9.4%) or as wage employees (13.7%). Like the overall measures, these numbers are all lower in the long recall group. Overall, this sample is heavily involved in household agricultural activities, with limited participation in other activities. Non-farm activities, when reported, are more likely to be secondary activities.\\n\\n# **3. Empirical Approach**\\n\\nIn the final interview, we randomized the order of the questions in the labor module so that some households used the long (annual) recall window first while others used the short (90-day) recall first. For those who received the short recall questions first, we use labor measures based exclusively on their quarterly survey responses, and for those who received the long recall first, we use measures based exclusively on the annual recall responses.<sup>14</sup> Our main analysis leverages this randomization to test for differences between these alternative recall windows using the following main empirical specification: L L L L L L 0 1L L L L L ii ii ii ii The coefficient of interest is 1, the difference in average reported labor measures between = ββ + ββ + XX + εε individuals using long and short recall. is a set of gender by age group by reporting status (proxy ββ ii or self) fixed effects. Age groups are defined as individuals under 25, 25-34, 35-49, and 50 and XX above, approximately quartiles of the study sample. Robust standard errors are clustered at the household level, the level of the treatment. In additional analyses we test for heterogeneity of these differences by proxy status, gender, and age. In these analyses we use fully saturated regression models to test each group’s differences against a null hypothesis of no effect and report p-values for differences across groups.\\n\\n# **4. Results**\\n\\n> 14 As discussed in Section 2, we have taken this approach of only using the data from the module that was administered first in the final round to avoid concerns over the potential influence of asking 90-day recall questions just before annual recall questions (or vice versa). Appendix Table 2 shows that this ordering indeed has a substantial impact, particularly on the long recall reported values.\", \"datasets\": [{\"mention_name\": {\"text\": \"quarterly survey responses\", \"confidence\": 0.32598480582237244, \"start\": 1122, \"end\": 1148, \"clean_text\": \"quarterly survey responses\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1122, \"end\": 1148, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1122, \"end\": 1148, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1122, \"end\": 1148}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1122, \"end\": 1148, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9947182536125183, \"start\": 336, \"end\": 362}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.32598480582237244, \"start\": 336, \"end\": 362}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9895283579826355, \"start\": 336, \"end\": 362}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9895283579826355, \"start\": 336, \"end\": 362}}, {\"mention_name\": {\"text\": \"data from the module\", \"confidence\": 0.39154738187789917, \"start\": 2320, \"end\": 2340, \"clean_text\": \"data from the module\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2320, \"end\": 2340, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2320, \"end\": 2340, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2320, \"end\": 2340}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2320, \"end\": 2340, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.28008830547332764, \"start\": 27, \"end\": 47}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.39154738187789917, \"start\": 27, \"end\": 47}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9858470559120178, \"start\": 27, \"end\": 47}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9858470559120178, \"start\": 27, \"end\": 47}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"# _4.1 Main outcomes_\\n\\nTable 3, Panel A, presents the main results. In columns 1 and 2 we examine whether the person worked at all and the total number of unique jobs reported. We find that long recall led to large reductions in both of these measures. Long recall reduces any reported work participation by 20 percentage points relative to short recall. Just 8% of individuals report no work whatsoever in the short recall, so an increase by 20 percentage points constitutes an enormous difference in reported labor force participation. Long recall also reduces the number of unique jobs reported by 0.3, which is 24% of the short recall mean. On the intensive margin of labor supply, column 3 shows a similar pattern for the number of months worked. Long recall reduces the reported number of months worked by 0.7 relative to short recall, 20% of the short recall mean. However, when considering hours worked, in column 4, there is only a small negative coefficient that is not statistically different from zero.\\n\\nIn columns 5, 6, and 7 of Table 3, Panel A, we consider the types of work reported for each individual. We report the impact of long recall on having worked on the household farm or in an agricultural-related home business, participation in a non-agricultural home business, or having done wage labor of any kind. As expected in a rural sample, agricultural work is the most common with 89% of individuals reporting participation in the short recall group. Long recall reduces this rate by 22%. These relative losses are comparable to those for having worked at all. Looking at non-agricultural household business and wage work, long recall reduces reported participation by six and seven percentage points, respectively, from short recall bases of 9 and 14 percent. These translate into very large proportional effects of 65% and 51%, suggesting disproportionate losses in reported non-agricultural labor contributions when using long recall that may reflect important sources of income diversification in this rural sample.\\n\\nGiven such strong effects on the other measures, the lack of detectable effect on the number of hours worked is surprising. One possibility is that respondents report hours worked in high intensity seasons instead of values that correspond more closely to a true mean across all months in a given range. This would mean that while reported hours worked vary between relatively high and low seasons in short recall, the maximal levels of labor intensity end up getting applied to all months worked across the full year in the long recall, generating an inflated estimate of hours even when months and activities are held constant. This pattern of reporting is similar to behaviors\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"shown in Arthi et al. (2018). It is also consistent with the reporting in our data. For those participating, the average hours per month worked under short recall (averaged across rounds) is 96.4 for primary jobs and 83.4 for secondary jobs. The hours per month for the long recall are notably higher, 111.9 for primary jobs and 115.7 for secondary jobs.\\n\\n# _4.2 Heterogeneity by self and proxy reporting_\\n\\nNext, we examine heterogeneous impacts of long recall by proxy/self-reporting, gender, and age. Because a respondent is thought to have more complete knowledge of their own labor contributions than those of other household members, use of long recall when reporting on behalf of others could be more taxing and more susceptible to omission. We present these results in Table 3, Panel B, estimating effects by proxy status using a fully saturated regression model, such that the reported interaction is the long recall effect for each group compared to no effect. Beneath each column, we also report the means in the short-recall group separately for self and proxy reports, the scaled differences of the treatment effects, and the p-value for the difference between the two groups. In considering these patterns we emphasize that a household’s primary respondent, and therefore proxy or self-reporting status, was not randomly assigned and therefore we cannot make causal claims of the effects of proxying itself on reported measures.<sup>15</sup> However, we can show how absolute and relative gaps between long and short recall differ by proxy status in our sample.\\n\\nAcross the first four outcomes, losses from long recall are larger for proxied individuals, in both absolute and relative (to their short recall mean) terms, although we lack sufficient precision to say whether these differences are statistically significant for the two intensive margin measures. Long recall loss for working at all among those who self-report is 5 percentage points, or 5% considering a short recall mean of almost 100%. The effect among those proxied is large: 42 percentage points or 54% of the short-recall mean for proxied individuals. These large differences between self and proxy reports may be due to increased recall bias in proxy reports or to the fact that primary respondents are engaged in different types of labor activities that may be less vulnerable to omission. When considering total activities and months worked (columns 2 and 3), the proportional effect is larger than for working at all (15% compared to 5%) suggesting that recall bias may differentially affect the reporting of a second job or the specific months worked for those\\n\\n> 15 The survey respondents are almost exclusively household heads and, as reported in Table 1, are overwhelmingly male and significantly older than the proxied respondents.\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"who are self-reporting. For these measures, the effect of long recall continues to be larger among the proxied respondents, both in absolute magnitude and percent effect. However, the difference is not statistically significant for months worked. As in the overall sample, there is no impact on hours worked, even while the point estimates are consistent with proxy reported labor being more vulnerable to long recall loss.<sup>16</sup>\\n\\n# _4.3 Heterogeneity by gender and age_\\n\\nNext, we examine differences in long recall loss by gender and age within respondent type (self or proxy). We begin with heterogeneity by gender, focusing on the impacts of long recall on total activities and number of months worked in Table 4, with worked at all and hours worked reported in Appendix Table 3. For each outcome we estimate three specifications testing for different types of heterogeneity: (1) respondents (self-reports) only, considering heterogeneity by their own gender, (2) proxied individuals only, considering the gender of the household’s respondent, and (3) proxied individuals only, considering heterogeneity by their own gender. We do not observe any meaningful differences in long recall loss by respondent gender among self-reported measures in columns 1 and 4 or proxied measures in columns 2 and 5. Losses among proxied women appear to be larger in absolute and relative terms, especially for number of months worked in column 6, but the estimates lack precision to distinguish these differences from statistical noise.\\n\\nFinally, we perform a similar analysis on effects by age group in Table 5 and Appendix Table 4, replicating the specifications and table structure from the gender analysis. While tests for differences between all four age groups do not always show statistical significance, it appears that older respondents have higher levels of long recall loss when reporting their own labor activities. Although column 2 does not show clear patterns, column 5 suggests that this pattern may be reversed when reporting about other household members, with older respondents exhibiting less long recall loss than younger respondents. Finally, younger household members appear to be more affected by long recall loss in both absolute and relative magnitude with significant differences among those under 25 or 25-34 when tested against people 50 and above.\\n\\n> 16 Regarding types of activities in columns 5-7 the patterns for farm work are similar to working at all. For nonagricultural businesses and wage work we observe larger absolute effects for those who are self-reporting; the proportional effect for non-agricultural businesses is larger for respondents and the effect on wage work is larger for proxies. In the latter case however, we cannot reject that the two effects are equal.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"# _4.4 Time pattern of results_\\n\\nTo better understand the dynamics and mechanics of these results, we show results for the four main outcomes by quarter, again focusing on quarters 2, 3, and 4 in Figure 2. Consistent with our discussion in section 2, we omit quarter 1 as the long recall window did not sufficiently overlap with the first short-run recall to afford comparable reference periods. Each panel shows the regression adjusted means by quarter, with a 95% confidence interval on the difference between the two recall groups.\\n\\nAcross all four outcomes, we see that long recall labor measures are furthest below the short run measures in the second quarter, the quarter furthest from the time of the final interview, and the difference is highly significant, showing proportionate reductions of approximately 40%. Reported differences in labor participation for any work, number of activities, and number of months worked grow smaller as the amount of time since the endline is reduced in quarters 3 and 4. However, the intensive measure of hours worked shows a sudden re-widening in the final quarter. And for both the number of months and hours worked, the long recall responses in quarter 4 are significantly greater than those based on the short recall-based responses.\\n\\nThe patterns shown in these results are puzzling given that major activities in the long recall should not be missed by the major activities that take place in 90-day segments within that range. One behavior that could lead to increasing labor measures over time in the long recall is if respondents have a tendency to “pull forward” the attributed months of a given activity from the months they report having started it up towards the present day at the time of the endline (in quarter 4). If, as discussed in section 4.1 respondents then apply their highest levels of participation for these activities, this could result in especially dramatic exaggerations in quarter 4. Despite this caveat, we generally observe meaningfully lower reported labor measurements when using long recall, with effects that increase along with time farther from the endline.\\n\\n# **5. Discussion and Interpretation**\\n\\nWhile we have discussed these gaps between long and short recall measurements as “long recall losses”, a few additional points deserve consideration in interpreting these differences as losses. First, the majority of current research suggests that reported data is less accurate with longer recall windows, however even the shorter, three-month recall window that we use as our benchmark may\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"be missing meaningful activities that could be captured using even shorter interview intervals and recall periods. To the extent that our short recall responses are themselves suffering from recall loss, our estimated long recall losses will underestimate the full extent of missed labor contributions.\\n\\nSecond, telescoping, the crowding in of actions that actually occurred just outside of the intended recall window, could lead to exaggeration of reported behaviors (Abate et al. 2022). While this phenomenon can affect responses using either short or long recall, short recall estimates could be especially impacted when aggregating across multiple survey rounds to characterize longer windows of time. However, the use of quarterly interviews may have also served as natural (if not explicit) time markers, bounding the time frame of their responses, which has been shown in other settings to limit the influence of telescoping (Abate et al. 2022). Further, asking individuals to specify specific months of participation in activities may further limit risks of these effects.\\n\\nA final consideration is that the act of responding to additional interviews, as used to collect the short recall data, might itself impact peoples’ survey responses or affect attrition (Arthi et al. 2018; Zwane et al. 2011). We note that because everyone in our analysis sample was interviewed the same number of times, this does not threaten the internal validity of our results, though it could influence whether we think we would get the same long-recall responses in the absence of the earlier interviews. The setup of our study prevents us from directly testing this although it could be an important question for future research.\\n\\nDespite these caveats, this work suggests that there are advantages to the three-month windows, although future research could evaluate their performance against even shorter recall periods to further inform these tradeoffs along with their vulnerability to telescoping or distortions induced by repeated interviews. Ultimately, researchers will need to carefully consider their context and the objectives of their research. They must weigh the value of more precise shorter recall measurements and the improved ability to capture seasonal, secondary, and informal labor contributions against survey costs and the potential for respondent fatigue.\\n\\n# **6. Conclusion**\\n\\nThe results presented in this paper suggest that using an annual recall window to measure seasonal variability in employment can lead to considerable losses relative to the use of a shorter recall\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"window in a set of quarterly interviews. Secondary, non-agricultural sources of income are disproportionately impacted. The extent of the losses that we observe in our analysis are heavily influenced by whether an individual’s labor contribution is being self-reported or reported by the household’s primary respondent. Labor contributions of youth are more likely to be omitted when using longer recall windows and relying on proxy reports. Because women and youth are less likely to be household heads and therefore more likely to be proxied in household surveys, the use of long recall periods may especially undercount their labor contributions. The time pattern of results is suggestive of more underreporting the further is the time from the survey time frame, coupled with possible overestimation of hours worked.\\n\\nGiven these losses, use of long recall windows can fundamentally affect our understanding of time patterns in peoples’ labor contribution, even when respondents are asked to disaggregate their work over the course of that period. This is not a principal goal of many multi-topic surveys that are primarily focused on the number of months worked and with even less emphasis on which ones. However, as policy makers increasingly demand data driven insights, continuing to refine and improve our methods of data collection is becoming more important. Understanding the seasonality, intensity, and diversity of labor is central to rural development planning. As this interest in the intensity, diversity, and seasonality of work continues to grow, an accompanying understanding of how to accurately measure labor contributions is becoming increasingly important. This objective cannot be separated from a need to understand how survey design and measurement choices, such as recall windows and proxy reporting, impact the resulting data, and finding ways to reduce their influence and potential biases.\", \"datasets\": [{\"mention_name\": {\"text\": \"household surveys\", \"confidence\": 0.353989839553833, \"start\": 547, \"end\": 564, \"clean_text\": \"household surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 547, \"end\": 564, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 547, \"end\": 564, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 547, \"end\": 564}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 547, \"end\": 564, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8888990879058838, \"start\": 547, \"end\": 564}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.353989839553833, \"start\": 547, \"end\": 564}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7764018177986145, \"start\": 547, \"end\": 564}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7764018177986145, \"start\": 547, \"end\": 564}}, {\"mention_name\": {\"text\": \"multi-topic surveys\", \"confidence\": 0.3389784097671509, \"start\": 1089, \"end\": 1108, \"clean_text\": \"multi-topic surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1089, \"end\": 1108, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1089, \"end\": 1108, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1089, \"end\": 1108}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1089, \"end\": 1108, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9909862875938416, \"start\": 1089, \"end\": 1108}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3389784097671509, \"start\": 1089, \"end\": 1108}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8404938578605652, \"start\": 1089, \"end\": 1108}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8404938578605652, \"start\": 1089, \"end\": 1108}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099314301232420189/pdf/IDU1f0a4a5fa17a29146e919386190683a48ead2.pdf\", \"pages\": [17]}}]}"}}},{"rowIdx":86,"cells":{"doc_id":{"kind":"string","value":"34243850"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34243850\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"admreg\": \"East Asia and Pacific,East Asia and Pacific\", \"admreg_key\": \"119225,119225\", \"authors\": {\"0\": {\"author\": \"Cameron,Lisa\"}, \"1\": {\"author\": \"Contreras Suarez,Diana Stella\"}, \"2\": {\"author\": \"Setyonaluri,Diahhadi\"}}, \"count\": \"Indonesia\", \"count_key\": \"82599\", \"docna\": {\"0\": {\"docna\": \"Leveraging Women’s Views to Influence Gender\\n Norms around Women Working : Evidence from an Online\\n Intervention in Indonesia\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"Social EAP 2 (SEAS2)\", \"projn\": \"ID-Indonesia Gender Equality For Growth\\n Program -- P172182\", \"subsc\": \"Social Protection\", \"trustfund\": \"TF0B1202-Indonesia and Timor Leste\\n Gender PASA,TF0B3379-Gender Equality for Growth,TF0B4399-High-level Dialogue on Gender\\n for Growth,TF0B4400-Reducing Gender Gaps in\\n Indonesia Labor Force,TF0B9888-Indonesia: Support for the Care Economy\", \"trustfund_key\": \"1824899,1931113,1976123,1976124,2230452\", \"theme\": \"Human Development and Gender,Gender\", \"prdln\": \"Advisory Services & Analytics\", \"prdln_key\": \"1416466\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34243850\"}, \"repnb\": \"WPS10681\", \"docdt\": \"2024-01-23T00:00:00Z\", \"datestored\": \"2024-01-23T02:24:04Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"How to influence social norms that drive\\n behavior in relation to women’s participation in\\n employment is not well understood. Providing randomly\\n selected participants with information on the extent of (i)\\n women’s support for women with children working;\\n (ii) husband’s support for sharing day-to-day\\n childcare with wives; and (iii) mothers’ and\\n mother-in-law’s support for working women,\\n increased the probability of choosing an online career\\n mentoring course for women over a shopping voucher of equal\\n value by 25 percent. Information beyond women’s\\n support for working women further increased support for\\n women working for some groups, although not strongly so.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10681; Impact Evaluation series; PEOPLE\", \"display_title\": \"Leveraging Women’s Views to\\n Influence Gender Norms around Women Working : Evidence from\\n an Online Intervention in Indonesia\", \"disclosure_date\": \"2024-01-23T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099424001232435449/text/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.txt\", \"listing_relative_url\": \"/research/2024/01/34243850/leveraging-women’s-views-influence-gender-norms-around-women-working-evidence-online-intervention-indonesia\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099424001232435449/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c\", \"new_url\": \"2024/01/34243850/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c\", \"ext_pub_date\": \"2024-01-23T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Gender and Labor Market m1326905 923,Gender m1327621 615,Social Norms and Poverty m1503846 3751\", \"docm_id\": \"IDU-edba8d7d-0b72-4eb1-bfe3-bb7ad3731c0c\", \"chronical_docm_id\": \"IDU-edba8d7d-0b72-4eb1-bfe3-bb7ad3731c0c\", \"publishtoextweb_dt\": \"2024-01-23T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10681\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"projectid\": \"P172182\", \"guid\": \"099424001232435449\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"43\", \"author_keys\": {\"0\": {\"author_key\": \"NA\"}, \"1\": {\"author_key\": \"1312517\"}, \"2\": {\"author_key\": \"860599\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099424001232435449/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10681 Leveraging Women’s Views to Influence Gender Norms around Women Working Evidence from an Online Intervention in Indonesia _Lisa Cameron Diana Contreras Suarez Diahhadi Setyonaluri_ Social Sustainability and Inclusion Global Practice & Poverty and Equity Global Practice January 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10681\\n\\n# **Abstract**\\n\\nHow to influence social norms that drive behavior in relation to women’s participation in employment is not well understood. Providing randomly selected participants with information on the extent of (i) women’s support for women with children working; (ii) husband’s support for sharing day-to-day childcare with wives; and (iii) mothers’ and mother-in-law’s support for working women, increased the probability of choosing an online career mentoring course for women over a shopping voucher of equal value by 25 percent. Information beyond women’s support for working women further increased support for women working for some groups, although not strongly so.\\n\\nThis paper is a product of the Social Sustainability and Inclusion Global Practice and the Poverty and Equity Global Practice.. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www. worldbank.org/prwp. The authors may be contacted at lisa.cameron@unimelb.edu.au, diana.contreras@unimelb.edu.au, and diahhadi.s@ui.ac.id.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# Leveraging Women’s Views to Influence Gender Norms around Women Working: Evidence from an Online Intervention in Indonesia<sup>_∗_</sup>\\n\\nLisa Cameron<sup>_†_</sup> Diana Contreras Suarez<sup>_‡_</sup> Diahhadi Setyonaluri<sup>_§_</sup>\\n\\n**Keywords** : gender norms, women’s employment, social desirability bias, Indonesia. **JEL Classification Code** : O12, O15, J16.\\n\\n> _∗_ We thank seminar participants at the Australian Gender Economics Workshop, Asian Development Bank, George Washington University, International Food Policy Research Institute, the International Bank for Reconstruction and Development, World Bank Jakarta Office, NBER Summer Workshop, University of Indonesia, University of Melbourne and Western Economic Association International Meetings for helpful comments that have improved the paper. We thank Saskia Rizqina Maulida for excellent research assistance and gratefully acknowledge funding from JPAL-SEA’s Indonesia Research Fund and from the Australian Government’s Department of Foreign Affairs and Trade through the Australia-World Bank Indonesia Partnership (ABIP). This paper was commissioned by the Social Sustainability and Inclusion (SSI) and Poverty and Equity (POV) Global Practices under the Indonesia Gender Equality and Growth research program (project code P172182) managed by Emcet O. Tas and Ririn S. Purnamasari.\\n\\n> _†_ Melbourne Institute of Applied Economic and Social Research, University of Melbourne; Email: lisa.cameron@unimelb.edu.au\\n\\n> _‡_ Melbourne Institute of Applied Economic and Social Research, University of Melbourne; Email: diana.contreras@unimelb.edu.au\\n\\n> _§_ Institute for Economic and Social Research (LPEM), Faculty of Economics and Business, University of Indonesia; Email: diahhadi.s@ui.ac.id.\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **1 Introduction**\\n\\nOver the last thirty years, many countries across the globe have made large investments in the education and health of girls and women. Such investments have, however, not systematically translated into increased participation of women in the labor market. The continuing low economic participation of women in many places is likely due to social norms that emphasize the role of women as mothers and carers. These gender norms - informal societal rules about appropriate or acceptable behavior for women and men - hinder female opportunities, choices, and achievements across the globe. (See for example Alesina et al. (2013); Fernandez (2013); Bertrand et al. (2015); and Jayachandran (2021).) While governments and non-government initiatives can try to change norms through public messaging, there is relatively little known about how effective such campaigns are and how best to influence them. Lack of information on the benefits of women working is one plausible explanation for such norms, but interventions providing information to families to change social norms related to married women’s labor force participation have had mixed success, Dean and Jayachandran (2019) and McKelway (2021).<sup>1</sup> Inaccurate perceptions of support among peers is another possible explanation. Bursztyn et al. (2020) show that in Saudi Arabia correcting underestimates of the extent to which male peers support women working outside the home increased men’s support for working women and increased women’s labor force participation.<sup>2</sup> Aloud et al. (2020), also in Saudi Arabia, find that informing female students of the labor market aspirations of their peers increases expectations about their own labor force participation. Cort ́es et al. (2012) find that individuals in the US also systematically overestimate gender conservativeness in relation to labor supply of mothers with young children, and that once information on peer beliefs is provided they are more likely to donate towards organizations advocating for women in the workplace.\\n\\nIn this paper we build on the literature examining the role of misperceptions of support among peers. We report the results of two data collection exercises. First, an online survey which was designed to measure and enhance our understanding of social norms around women’s work in Indonesia. Specifically, we collected information from approximately 500 female and 500 male respondents in metropolitan areas across Indonesia on their behavior\\n\\n> 1Dean and Jayachandran (2019) and McKelway (2021) evaluate interventions in India which provided information on female employment opportunities and/or the benefits of female employment. Both found very little change in the acceptability of women working. Makino (forthcoming) however found that havingparents of young women in Pakistan attend a two-hour lecture which provided information on the safe,female-friendly working environment in large garment factories made parents more positive about theirdaughters working in factories. Interventions targeting adolescents in India have also had some successin changing gender attitudes (see Dhar et al. (2022)).\\n\\n> 2While not designed to change social norms, Field et al. (2021) show that male attitudes towards female work and their beliefs about community acceptance of women working were shifted by an intervention that resulted in women receiving wages into their own bank account rather than their husband’s account. This shift was hypothesized to come about due to the consequent increase in women’s household bargaining power.\\n\\n1\", \"datasets\": [{\"mention_name\": {\"text\": \"online survey\", \"confidence\": 0.509739339351654, \"start\": 2242, \"end\": 2255, \"clean_text\": \"online survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2242, \"end\": 2255, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2242, \"end\": 2255, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2242, \"end\": 2255}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2242, \"end\": 2255, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9997987151145935, \"start\": 665, \"end\": 678}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.509739339351654, \"start\": 665, \"end\": 678}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8164626359939575, \"start\": 665, \"end\": 678}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8164626359939575, \"start\": 665, \"end\": 678}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"(whether female respondents and wives of male respondents work outside the home), personal attitudes (level of support for women working) and injunctive norms (incentivized estimates of the extent to which others are supportive of women working). To better understand the motivations that underpin such norms, we also collected information on relevant reference groups (people whose opinion is important to respondents), the concerns that men and women have about women working, and the sanctions they may face, if the wife works outside the home.\\n\\nWe find that respondents’ estimates of the level of support among men for married women with children working outside the home for pay are relatively accurate (unlike Bursztyn et al. (2020)), but both men and women significantly underestimate the extent of women’s support.<sup>3</sup> Men and women also underestimate the extent of support among men for husbands sharing day-to-day childcare responsibilities with wives.<sup>4</sup> Information on which reference groups’ opinions are most important to respondents when deciding whether to work (for women) or, to support their wives working (for men) shows that both men and women are highly concerned about the views of their mothers and mothersin-law when deciding whether they (female respondents) or their wives (male respondents) work. These results underpin the online intervention which we subsequently conducted with more than 4,000 respondents in our second online survey. In our three treatment arms we expose male and female participants to information on the extent of support:\\n\\n1. among women for women with children working;\\n\\n2. among men for parents sharing childcare; and 3. among older women (in respondents’ mother’s and mother-in-law’s generation) for women with children working.\\n\\nAll treatment groups receive information on the extent of women’s support for women with children working; treatment group 2 also receives information on the extent of men’s support for shared child care; and treatment group 3 receives all three types of information listed above.\\n\\n> 3Bursztyn et al. (2023) examine misperceptions of social norms using nationally representative datasets from 60 countries, including Indonesia. They find that Indonesia has one of the lowest shares of support for women having the right to work outside the home (supported by 59% of men and 73% of women). These figures are very similar to the levels of support found in our study - 62% and 76% for men and women respectively. They further find that Indonesians underestimate the level of support for women having the right to work outside the home. The underestimation of women’s support for women working outside the home is 7.9 percentage points (very similar to the average underestimation of 9.4 percentage points in our sample); and of men’s support is 6 percentage points (slightly larger than the 2.5 percentage point underestimate in our sample).\\n\\n> 4Changing men’s behavior in relation to participating in home production and childcare goes hand-inhand with increasing female labor force participation. Since childcare is still largely a gendered task, changing this norm has the potential to increase women’s ability to search for employment opportunities outside the home.\\n\\n2\", \"datasets\": [{\"mention_name\": {\"text\": \"nationally representative datasets\", \"confidence\": 0.7112029790878296, \"start\": 2155, \"end\": 2189, \"clean_text\": \"nationally representative datasets\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2155, \"end\": 2189, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2155, \"end\": 2189, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2155, \"end\": 2189}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2155, \"end\": 2189, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.7380211353302002, \"start\": 544, \"end\": 578}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7112029790878296, \"start\": 544, \"end\": 578}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8891000151634216, \"start\": 544, \"end\": 578}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8891000151634216, \"start\": 544, \"end\": 578}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"The interventions significantly increase both men’s and women’s support for working – women 43% of participants in the treatment arms, compared to 34% in the control group, chose to select an online career mentoring course for themselves (female respondents) or their wives (male respondents) instead of a shopping voucher of equal value. The vast majority of vouchers for the career mentoring course were redeemed, signifying respondents’ genuine interest in women’s labor force participation. The provision of the additional information on men’s childcare sharing norms and older women’s attitudes towards working women further increased support for women working for some groups, although not strongly so. The two additional pieces of information together increased the probability of choosing the career mentoring course increased by 5.7 percentage points (16%) for working women (p=0.11). For male respondents whose wives were not working, the additional information about the extent of support among men for parents sharing childcare increased their likelihood of selecting the career mentoring course by 6.5 percentage points (21%) but not significantly so (p=0.18). The information on the level of support among women in their mothers’ and mothers-in-laws’ generation did not further increase men’s support. Taken together, the impact of the interventions across all treatment arms was about twice as large for respondents with school-aged children (aged over 6 years) than for those with pre-school aged children, reflecting that young children in an environment with very restricted access to childcare are a significant barrier to women’s economic participation.\\n\\nHeterogeneity analysis rejects a model of norm formation in which the further one’s own perceptions of the social norm are from that which is revealed to be true, the greater the adjustment in one’s attitude. Rather, being made aware that there is a high level of support for working women appears to uniformly increase support for working women, regardless of initial expectations. The level of support increases both for those who underestimated and overestimated the level of support and doesn’t vary with the magnitude of the perception error. This result is contrary to that found by Bursztyn et al. (2020), Aloud et al. (2020) and Cort ́es et al. (2012).\\n\\nOur paper contributes to the small but growing literature which uses field experiments to examine the impacts of interventions that address misperceptions about others’ beliefs in relation to women’s employment. While interventions to “correct” misperceptions about others’ beliefs (pluralistic ignorance) have been frequently used to study voting behavior and preferences for income redistribution (see Bursztyn and Yang, 2022 for a review), to our knowledge only Bursztyn et al. (2020), Aloud et al. (2020) and Cort ́es et al. (2012) have experimentally examined inaccuracies in perceptions of support for women’s – work first two studies in the context of Saudi Arabia and among male neighbors and female students, respectively, and the third in the US.<sup>5</sup> We build on this evidence and test\\n\\n> 5Aloud et al. (2020)) examine the impact of information on what percentage of female students expect to be working for pay when they are twenty-five years old alongside information on monthly wages and a\\n\\n3\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"– its effectiveness in a different cultural context Southeast Asia which is home to almost 700 million people and is very distinct from the context of previous work. While still largely more socially conservative than Western nations, female labor force participation in Southeast Asia is considerably higher than in the Middle East and women generally have more freedoms, even in Muslim majority Indonesia.\\n\\nWe introduce two innovations to the literature on “pluralistic ignorance” in relation to women’s employment. First, ours is the first paper of which we are aware which explores whether information on women’s attitudes can be harnessed to change men’s attitudes.<sup>6</sup> Our data suggest that both men and women have relatively accurate perceptions of the extent of support among men for women working but significantly underestimate the support among women for working women, consistent with women not being able to voice their opinions. We explore whether making participants (men and women) aware of their misperceptions of women’s support for working women changes attitudes. Second, we examine complementary norms that drive decision making for women’s employment outside the home. In addition to norms around the acceptability of women working, we seek to address the strong norms around women being responsible for childcare. We do this by expanding the type of information provided to include information on attitudes to the sharing of childcare between husbands and wives. Our intervention also incorporates information on mothers and mothers-in-law’s support for working women as these groups were identified by respondents as being the most important reference groups whose views are important to the female labor force participation decision.\\n\\nWe also contribute to the literature on light touch interventions aiming to change norms. Our results demonstrate that an inexpensive online intervention, as opposed to a more costly in person intervention, can impact social norms. Previous literature had cast doubt on the effectiveness of light touch interventions. For example, Dean and Jayachandran (2019) and McKelway (2021) found that interventions which showed videos to family members highlighting female employment opportunities and the nonmonetary benefits of women’s employment had no effect on families’ support for female members’ employment. That, in our case, such a light touch online intervention was able to significantly increase the extent of support for working women is good news for governments and other policy bodies seeking to increase women’s labor force participation. If accurately targeted to address existing misperceptions, our results suggest that less costly online campaigns can successfully sway norms. Although we are unable to provide evidence on the longevity of such a change, work by Field et al (2021) suggests that norms changed in the short run can have longer run impacts and Bursztyn et al. (2020) found\\n\\n> job assistance program. Gauri et al. (2019) also find evidence of misperceptions on attitudes to women’s work in Jordan but do not implement an intervention to address these misperceptions.\\n\\n> 6Bursztyn et al. (2020) study the effect of men’s attitudes on men and women. Aloud et al. (2020) study the effect of women’s attitudes on women. Cort ́es et al. (2012) examine the effect of revealing own-gender attitudes on both men and women.\\n\\n4\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"that the change in perceptions of norms evoked by their intervention continued to affect perceptions three to five months after exposure. That we were able to change norms in the short run suggests that continued exposure to such messaging (for example, in an ongoing public information campaign) is likely to change norms over the longer term.\\n\\nExtrapolating from Bursztyn et al. (2020) suggests that our intervention could result in an increase in female labor force participation as large as 6 percentage points (12%) from the current FLFP rate of 53%. That is, an extra 3.5 million women working.\\n\\nA further, and interesting, corollary of our findings is that in contexts where women’s support for working women (or any other social phenomenon) is underestimated - and the social norm is formed on the basis of observations of people’s behavior rather than stated preferences - empowering women to express their opinions would likely encourage adoption of a less conservative norm. Hence, empowerment of women’s voices is an alternative policy prescription to public information campaigns for changing social norms.\\n\\nFinally, the paper makes a methodological contribution by demonstrating that real stakes outcome measures are needed to evaluate changes in norms. Twenty percent of our respondents were asked to make a hypothetical choice between the career-mentoring course and the shopping voucher. Unlike the real reward choices, hypothetical choices were significantly and overwhelmingly swayed by social-desirability bias. When the choice was hypothetical, individuals who were the most prone to social desirability bias - measured – via a 5-point Crowne and Marlowe (1960) scale were 28 percentage points (72% of the control mean) more likely to choose the socially-desirable career-mentoring course than those with the lowest social desirability bias scores. In contrast, when the stakes were real, social desirability bias had an insignificant effect on respondent choices. In the absence of real stakes, the interventions would have appeared ineffective.\\n\\n# **2 Data Collection**\\n\\nData was collected in two online surveys. The first online survey covers 1,050 respondents (50% male, 50% female) residing in large urban centers throughout Indonesia.<sup>7</sup> The aim of the first survey was to measure social norms and people’s perceptions of these norms. We targeted respondents who were 18 to 40 years old, with at least a junior secondary education and who were married with at least one child aged under 18 years\\n\\n> 7We sample from urban areas as previous research has shown the negative impact of marriage and childbirth on women’s labor force participation to be larger in urban areas where women are more educated, there are more job opportunities and women’s productivity is potentially higher, Cameron et al. (2019). Both surveys were conducted using Qualtrics’ online platform and with members of Qualtrics panel respondents who met our eligibility criteria, i.e., they lived in a metropolitan area, were married, aged 18-40 years, living with their partner, had children under the age of eighteen, and had at least junior secondary education. Metropolitan areas are those areas defined as such by the Centre for Urban Development and include Lampung, Bandung, Batam, Bekasi, Bogor, Depok, Makassar, Medan, Palembang, Pekanbaru, Semarang, Tangerang, and Jakarta. See the Regional Infrastructure Development Agency’s Strategic Plan 2020-2024: https://bpiw.pu.go.id/product/get ~~i~~ ndex/1.\\n\\n5\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"and living with their spouse. We focus on respondents with at least junior secondary school education as research has shown that women with this level of education have the most discretion over whether they work or not and so are more likely to be able to have their behavior influenced.<sup>8</sup> The sample was constructed such that 75% of both male and female respondents were high school educated and 25% tertiary educated (to roughly reflect the coverage of these groups in the Indonesian population). The survey collected demographic information (including age, gender, number of children, own and spousal work status) and information on respondents’ attitudes and perceptions of social norms.\\n\\nSpecifically, we collected information on personal attitudes in relation to the extent of support for a) female labor force participation: “Are you supportive of married women with children under 12 working for pay outside the home?”; and b) sharing of childcare between husband and wife: “How supportive are you of husbands sharing day-to-day childcare duties with their wives?”.\\n\\nWe also elicited injunctive norms on what proportion of married men and married women are supportive of the above behaviors. These estimates were incentivized by paying approximately USD17 to the respondents who made the most accurate estimate. The difference between the mean of the sample’s reported level of support and the individually reported perception of social norms in both domains generates a measure of the extent of misperceptions about the social norm.\\n\\nWe also collected information on which people’s opinions are important to respondents when making decisions about female household members working and sharing childcare. Respondents were asked to indicate the extent to which they were concerned about the opinions of husbands, parents, parents-in-law, extended family members, people in their social networks and people in their religious community. Finally, we enquired about the potential sanctions individuals would face if they decided to deviate from the social norm.\\n\\nOur second online survey collected similar information from 4,478 similarly selected respondents. The information intervention was embedded in the survey just prior to survey exit. The intervention is detailed further below in Section 4.<sup>9</sup>\\n\\n# **3 Social Context**\\n\\nFemale labor force participation in Indonesia has remained relatively constant over the – past two decades with around 50% of women working even with high economic growth, concomitant large increases in women’s educational attainment and the service sectors\\n\\n> 8Women with upper secondary education have the lowest female labor force participation in Indonesia, Cameron et al. (2019).\\n\\n> 9Ethics approval for the surveys and interventions was obtained from the University of Melbourne (202223161-28577-5) and University of Indonesia (LPEM FEB - 14/UN2.F6.D2.LPM/PPM.KEP/2022). The trial and pre-analysis plan were registered with the AEA RCT Registry (AEARCTR-0009493). Unless otherwise indicated, the analysis presented follows that laid out in the pre-analysis plan.\\n\\n6\", \"datasets\": [{\"mention_name\": {\"text\": \"second online survey\", \"confidence\": 0.41512247920036316, \"start\": 2081, \"end\": 2101, \"clean_text\": \"second online survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2081, \"end\": 2101, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2081, \"end\": 2101, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2081, \"end\": 2101}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2081, \"end\": 2101, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9997616410255432, \"start\": 389, \"end\": 409}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.41512247920036316, \"start\": 389, \"end\": 409}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5913382172584534, \"start\": 389, \"end\": 409}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5913382172584534, \"start\": 389, \"end\": 409}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"and a decrease in fertility rates.<sup>10</sup> Previous research in Indonesia has shown that women’s economic participation is hindered by marriage and childcare responsibilities (for example, Cameron et al. 2019, 2023).<sup>11</sup> To the extent that data on Indonesian social norms were available prior to our study, they suggest that norms in relation to women’s work are conservative. For example, Indonesian men have a similar tolerance to women working as Saudi Arabian men and are less tolerant than Indian men.<sup>12</sup> Data on attitudes to women’s work from the World Values Survey 2018 shows that 76% of Indonesian men agree with the statement that men have more right to a job than women, with women being only slightly (2 ppts) less likely to agree with this statement than men.<sup>13</sup> Women and men, even young adults, continue to strongly conform to social norms that emphasize women’s childcare and domestic responsibilities because women are perceived as being better at care-giving (YouGov and Investing in Women, 2020). In a qualitative study of 40 young adults in Greater Jakarta and Surabaya, Setyonaluri et al. (2021) found that such persistent social norms often stem from perceptions of kodrat, or God’s will when defining gender roles. Participants in their study saw women working for pay as positive, but only if it was done to support husbands and women did not ‘neglect’ their responsibilities at home.\\n\\nAs a result of our sample being comprised of respondents in large urban centers who have access to the internet, 81% of the women in our sample worked, significantly above the national participation rate of 53%.<sup>14</sup> However, only 53% of wives of male respondents worked.\\n\\nOf the female respondents who were working, only 41% worked exclusively outside the home (45% of the wives of male respondents). Of female respondents who worked, 47% were wage workers, with the remainder being self-employed with no employees (37%), self-employed with employees (7%) or casual or family workers (9%). If we define the formal sector to consist of wage workers and self-employed businesses with staff, 54% of working women in our sample are employed in the formal sector. The formal sector is not particularly family friendly with flexible and part-time work rarely being available (Cameron et al., 2023).<sup>15</sup>\\n\\n> 10Of Indonesian women aged over 25 years in 2018, 31% had completed upper secondary school, compared to only 3.4% in 1980. World Bank Databank. Accessible at https://databank.worldbank.org. The service sector accounted for about 29% of employment in 1991 and 49% in 2019. See https://data.worldbank.org/indicator/SL.SRV.EMPL.ZS?locations=ID&view=chart.\\n\\n> 11Halim et al. (2023) show that the expansion of public pre-school provision from 2003 has not changed this situation.\\n\\n> 12A Gallup survey found that 43% of Indonesian men prefer women to not engage in paid work outside the home, the same as Saudi Arabia and more than India (35%), (Gallup and ILO, 2017).\\n\\n> 13Gender norms are becoming more conservative in Indonesia over time. In the 2006 World Values Survey 65% and 42% of men and women agreed with the above statement respectively.\\n\\n> 14As reported in the World Bank data bank, see https://data.worldbank.org/indicator/SL.TLF.CACT.FE.ZS?locations=ID/.\\n\\n> 15Of the wives of male respondents who worked, 58% were wage workers and 62% were employed in the formal sector. Cameron et al. (2023)) find that women who were employed in the formal sector prior to having their first child were 20 percentage points less likely to be working than other women in the year following the birth.\\n\\n7\", \"datasets\": [{\"mention_name\": {\"text\": \"data on Indonesian social norms\", \"confidence\": 0.72159743309021, \"start\": 254, \"end\": 285, \"clean_text\": \"data on Indonesian social norms\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 254, \"end\": 285, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 254, \"end\": 285, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 254, \"end\": 285}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 254, \"end\": 285, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.8420897126197815, \"start\": 254, \"end\": 285}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.72159743309021, \"start\": 254, \"end\": 285}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5800343155860901, \"start\": 254, \"end\": 285}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5800343155860901, \"start\": 254, \"end\": 285}}, {\"mention_name\": {\"text\": \"Data on attitudes to women’s work\", \"confidence\": 0.31444987654685974, \"start\": 534, \"end\": 567, \"clean_text\": \"Data on attitudes to women’s work\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 534, \"end\": 567, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 534, \"end\": 567, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 534, \"end\": 567}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 534, \"end\": 567, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9981003403663635, \"start\": 534, \"end\": 567}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.31444987654685974, \"start\": 534, \"end\": 567}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.75181645154953, \"start\": 534, \"end\": 567}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.75181645154953, \"start\": 534, \"end\": 567}}, {\"mention_name\": {\"text\": \"World Values Survey\", \"confidence\": 0.5742650032043457, \"start\": 577, \"end\": 596, \"clean_text\": \"World Values Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 577, \"end\": 596, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 577, \"end\": 596, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 577, \"end\": 596}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 577, \"end\": 596, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9981003403663635, \"start\": 577, \"end\": 596}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5742650032043457, \"start\": 577, \"end\": 596}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.75181645154953, \"start\": 577, \"end\": 596}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.75181645154953, \"start\": 577, \"end\": 596}}, {\"mention_name\": {\"text\": \"Gallup survey\", \"confidence\": 0.6094074249267578, \"start\": 2844, \"end\": 2857, \"clean_text\": \"Gallup survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2844, \"end\": 2857, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2844, \"end\": 2857, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2844, \"end\": 2857}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2844, \"end\": 2857, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9986490607261658, \"start\": 695, \"end\": 708}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6094074249267578, \"start\": 695, \"end\": 708}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6519917249679565, \"start\": 695, \"end\": 708}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6519917249679565, \"start\": 695, \"end\": 708}}, {\"mention_name\": {\"text\": \"2006 World Values Survey\", \"confidence\": 0.8021218776702881, \"start\": 3105, \"end\": 3129, \"clean_text\": \"2006 World Values Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3105, \"end\": 3129, \"clean_text\": \"\"}, \"producer\": {\"text\": \"World Bank data bank\", \"confidence\": 0.49648284912109375, \"start\": 3229, \"end\": 3249, \"clean_text\": \"World Bank data bank\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3105, \"end\": 3129}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3105, \"end\": 3129, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9988259673118591, \"start\": 147, \"end\": 171}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8021218776702881, \"start\": 147, \"end\": 171}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7523847818374634, \"start\": 147, \"end\": 171}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7523847818374634, \"start\": 147, \"end\": 171}}, {\"mention_name\": {\"text\": \"World Bank data bank\", \"confidence\": 0.3268444836139679, \"start\": 3229, \"end\": 3249, \"clean_text\": \"World Bank data bank\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3229, \"end\": 3249, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3229, \"end\": 3249, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3229, \"end\": 3249}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3229, \"end\": 3249, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"report\", \"confidence\": 0.9761022329330444, \"start\": 271, \"end\": 291}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3268444836139679, \"start\": 271, \"end\": 291}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6271424293518066, \"start\": 271, \"end\": 291}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6271424293518066, \"start\": 271, \"end\": 291}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"Among female respondents who are not working or looking for work, 76% reported that they were unable to do so because of childcare - they had either chosen to look after their children or could not find anyone else to look after them.<sup>16</sup> A further 20% reported that they were not working as their husband did not want them to. These findings support the conjecture that there is considerable scope for interventions to increase female labor force participation, either through the provision of childcare or, as is our focus below, by changing social norms around women’s work and childcare responsibilities.\\n\\n# **3.1 Social Norms around Women’s Work**\\n\\nConservative social norms aside, our data reveal that a majority of Indonesian men and women support women with children under 12 working outside the home for pay - 76% of female respondents reported being supportive and 62% of male respondents.<sup>17</sup> All respondents were asked to nominate up to three reasons against supporting women with children under 12 working for pay outside the home. Figure 1 shows that the most often reported reason men give (reported by 22% of male respondents) is that women’s role is to care for their children, whereas women most often report that finding someone to look after their children is difficult. Of both male and female respondents, 16% report that working will result in a mother neglecting her family duties.\\n\\nFemale (male) respondents were also asked who, among a list of family members and social contacts, would not be supportive of themselves (their wife) working for pay outside the home. The most oft-cited category was mothers (16% for female respondents; 15% for male respondents). Mothers-in-law came a close second at 12% for female respondents and 14% for male respondents. Around 85% stated they were sure of their mother’s and mother-in-law’s attitude. Further, 98% (80%) of women (men) report that it is important to them to have their mother’s support. Having their mothers-in-law’s support was important to 84% of both men and women. Mothers and mothers-in-law are hence an important reference groups for both men and women. Of female respondents, 91% reported that having their husband’s support is important to them.\\n\\nFigure 2 shows what women who report that their husband’s support is important to them would be most concerned about in relation to their husband if they worked. The concern women most often reported is that their husband would be worried that others will think that he is not able to financially provide for his family (34%, or 45% if we include those who report that others will think the family is in financial need); that the husbands will view them (the women) as neglecting their family (21%); and that others will think that they do not respect their husband (20%). Only 5% reported that they themselves will not be respected by their husbands. Interestingly, for this better educated\\n\\n> 16Currently Indonesian preschool sessions are very short and so allow women to increase unpaid work but not paid activities outside the home (Halim et al., 2023).\\n\\n> 17 Defined as people who reported they were very supportive, supportive or somewhat supportive.\\n\\n8\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"group of women in Muslim-majority Indonesia, only 4% reported being concerned that others will view them as not following their religious traditions.<sup>18</sup> Our findings are similar to those of Bernhardt et al. (2018) who found that in India the husband’s social status was the main concern for husbands if their wives worked outside the house. The main concern for women in India, however, was that they were perceived as being disobedient to their husband.\\n\\n# **3.2 Misperceptions about Social Norms Pertaining to Married Women Working**\\n\\nFigure 3 shows the level of support reported for women with children under 12 working for pay outside the home and the estimated (perceived) level of support amongst peers. Female (male) respondents were asked to think of women who are similar to themselves (their wife) in terms of education level and to estimate “Out of 100, how many of these women do you think are supportive of wives with children under 12 years working for pay outside the home?”. The right hand set of bars show that while the actual level of support reported among female respondents was 76%, the mean (incentivized) level of women’s support as estimated by female respondents was 67% - an underestimate of 9 percentage points; and men estimated that only 59% of women are supportive - an even larger underestimate of 17 percentage points.\\n\\nThe left hand set of bars in the same figure show men’s support for women working for pay outside the home. Both men and women estimate this level of support among men (62%) relatively accurately. The average estimate of the level of male support among male respondents is 59% of men, an underestimate of just 3 percentage points. Women overestimate the level of support among men by 2 percentage points.\\n\\nThese findings are consistent with men’s views being more widely known and hence more accurately perceived, while women’s views are more closely held. They suggest that there is scope to influence gender norms in favor of working women by providing information about the greater than expected support among women. There is however little scope for an intervention to influence gender norms by providing information on men’s support for working women (a la Bursztyn et al. (2020). This result underscores the importance of formative research to ensure intervention design reflects the cultural context.\\n\\n> 18A similar pattern was found for respondents who reported that their mothers were not supportive of women working outside the home - 50% (46%) of women (men) reported that if the woman worked their mothers would think that the husband was not able to provide for the family or that the family was in financial need. Of the female respondent, 21% were worried that their mother would think they were neglecting their family.\\n\\n9\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"## **3.3 Misperceptions about Social Norms Pertaining to Sharing of Childcare**\\n\\nA similar range of questions were asked about the level of support among men and women for husbands and wives sharing day-to-day childcare duties. The professed support for husbands sharing day-to-day childcare responsibilities with wives - for example, feeding the child, bathing and dressing the child, taking the child from/to school, as well as monitoring the child’s nap times, playtimes, and other activities - is very high with 90% of men and 96% of women being supportive.<sup>19</sup> Figure 4 shows that both men and women substantially underestimate the high level of support among their peers for shared childcare (e.g. men estimate that 65% of men support sharing childcare duties). This points to scope for an effective information intervention that tells people about the level of support in the community for shared childcare.\\n\\n# **4 Evaluation Design**\\n\\n## **4.1 Theory of Change**\\n\\nThe theory of change that underpins the intervention is illustrated in Figure A.1. Social norms reflect people’s perceptions of the attitudes of others in society. Informing individuals of inaccuracies in their perceptions of others’ attitudes causes updating of their perceptions and as individuals’ behavior is theorized to be determined, in part, by social norms, the updating of such norms results in behavioral change.<sup>20</sup> Over time this will create a self-reinforcing loop in which others observe the changed behavior and update their perceptions and change their behavior.\\n\\n## **4.2 Intervention Design**\\n\\nRespondents in the second survey were randomly divided into four groups – a control group and three treatment groups with the treatment groups receiving information designed to influence their social norms. The design of the interventions reflects the results of the first survey and, as discussed above, consists of the provision of the following information:\\n\\n> 19In practice most of the childcare is, however, performed by wives. Our data show that men overestimate their share of childcare relative to what wives report. On average they report they undertake 34% of childcare duties with their wives doing 54%, while women report their husbands only undertake 23% of childcare duties, compared with their 63%.\\n\\n> 20Bursztyn et al. (2020) present a simple theoretical model to show how social norms affect labor force decisions in a world where husbands makes the decision as to whether to allow their wives to work so as to maximize utility. Utility is modelled as a positive function of the income from wife’s employment but with costs to utility associated with the stigma of acting against societal norms and the psychic cost of making a decision incompatible with one’s own beliefs. They show that if a sufficiently large shift in beliefs occurs (reducing the perceived probability of being stigmatized), the new equilibrium will see an increase in the number of women working.\\n\\n10\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"- 1) The extent of women’s support for women with children under 12 years working for pay outside the home (reflecting the underestimates among men and women of the level of women’s support found in the first survey)\\n\\n- 2) The extent of men’s support for childcare being shared among husbands and wives (reflecting the underestimates among men and women of the level of men’s support in the first survey); and\\n\\n- 3) The extent of older women’s (from the respondents’ mothers’ generation) support for women with children working for pay outside the home (reflecting the concern about mothers’ and mothers-in-law’s support found in the first survey).<sup>21</sup>\\n\\nThe control group received no such information. Figure A.2 in the appendix shows the format in which this information was presented to each treatment group.\\n\\nOur main outcome measure is respondents’ choices as to whether to receive payment for their participation in the form of an online career mentoring course for women or an online shopping voucher of equal value. Respondents were told that about one in every three participants in the survey would be randomly selected to receive a reward for participating in the research project (and they would be told at the end of the survey whether they had been selected for the reward). They were then asked to indicate whether they would prefer to receive:\\n\\n- Free access to an online career mentoring course for female participants or the wives of male respondents. The course provides practical career advice from HR professionals and was valued at Rp100,000 (USD6.50). It equips participants with the skills to create a CV, write a cover letter, prepare for a job interview, and create a LinkedIn profile; or\\n\\n- a Rp100,000 convenience store shopping voucher.<sup>22</sup>\\n\\nChoosing the career mentoring course is taken to indicate support for their own (for female respondents) or their wife’s (for male respondents) labor force participation. A comparison of this variable across the control and treatment groups provides an estimate of the interventions’ impacts.<sup>23</sup>\\n\\n> 21In the first survey we did not collect information on the support of mothers or mothers-in-law for married women with children working outside the home. The information provided on norms in this domain is constructed from 2018 Indonesian World Values Survey data on the proportion of women aged 40 to 60 who disagree or strongly disagree with the statement “When a mother works for pay, the children suffer”. For more information on the World Values Survey, see Haerpfer et al. (2022).\\n\\n> 22For information on the career mentoring course, see https://skillacademy.com/p/career-mentoringregular-bimbingan-untuk-dapat-kerja?courseType=SingleCourse. The shopping vouchers were redeemable at Indomaret and Alfamart stores which are ubiquitous throughout urban Indonesia.\\n\\n> 23We also conducted a list experiment which generates an alternative outcome measure. At the end of the survey (just prior to the choice of reward), respondents were asked how many of the following statements they agreed with (in randomized order): a) The minimum wage should be kept at its current\\n\\n11\", \"datasets\": [{\"mention_name\": {\"text\": \"2018 Indonesian World Values Survey data\", \"confidence\": 0.6140225529670715, \"start\": 2321, \"end\": 2361, \"clean_text\": \"2018 Indonesian World Values Survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2321, \"end\": 2361, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2321, \"end\": 2361, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2321, \"end\": 2361}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2321, \"end\": 2361, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9794685244560242, \"start\": 773, \"end\": 813}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6140225529670715, \"start\": 773, \"end\": 813}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5506511926651001, \"start\": 773, \"end\": 813}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5506511926651001, \"start\": 773, \"end\": 813}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"# **5 Results**\\n\\n## **5.1 Summary Statistics and Tests of Balance**\\n\\nThe demographic characteristics of respondents in the second survey are similar to the first. The average age of respondents is 30.5 years and 92% of respondents are Muslim. Almost all men work (99%), while 83% of women work. Wives of male respondents are considerably less likely to be working (53%) than female respondents.\\n\\nRespondents were randomly allocated across treatment arms (with stratification by gender; education; and whether the voucher choice was real or hypothetical).<sup>24</sup> Table 1 provides summary statistics and tests of balance. Importantly, we found that the personal beliefs in this sample are almost identical to those in the first survey which were used for the treatment design. Of female respondents, 75% reported that they were supportive of married women with children under the age of 12 working for pay outside the home (compared to 76% in the first survey), and 63% of male respondents were supportive (compared to 62% in the first survey). Further, 95% (90%) of female (male) respondents reported being supportive of shared day-to-day childcare, compared to 96% (90%) in the first survey. Respondents reported that 75% of their mothers are supportive of the above (this information was not collected in the first survey).\\n\\nThe control and treatment arms are well balanced. Only two variables differ across arms, with the differences being relatively small and statistically significant only at the 10% level.\\n\\n### **5.1.1 Social Desirability Bias**\\n\\nOne concern with reporting of attitudes is that the reports may reflect experimenter demand effects. That is, respondents might be more likely to report friendlier attitudes towards working women, to experimenters who are likely to be in favor of women working. The effect of social desirability bias is lessened in online surveys due to the lack of personal interactions, nevertheless, to ascertain the extent to which this is a problem, we collected information on a 5-item social desirability scale following Crowne and Marlowe (1960)\\n\\n> level; b) It is currently difficult to find a good job in Indonesia; c) Unemployment is a big problem in Indonesia; and d) Women with young children should be supported to work outside the home. The list experiment enables the researcher to identify whether respondents in the treatment groups are more likely to agree with the statement about women with young children being supported to work outside the home (as it is the only statement that should be affected by the information interventions) from a comparison of answers across the control and treatment groups, without being able to identify whether individual respondents agreed with that particular statement. It thus has the advantage of not being affected by social desirability bias. However, it produces imprecise estimates of intervention impacts. The treatments did not have a significant impact on choices in the list experiment. Results available on request.\\n\\n> 24Two of Qualtrics’ panel partners would not allow us to provide vouchers to respondents. These panel partners contributed 20% of respondents. These respondents were asked which they would choose if given a choice i.e. a hypothetical choice.\\n\\n12\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"and Hays et al. (1989). This module asked respondents whether they have several toogood-to-be-true traits such as being always courteous even to people who are disagreeable, never taking advantage of others, being always forgiving, being never resentful and being always a good listener. We sum these variables to construct a social desirability index (SDI) which ranges in value from 0 (least subject to social desirability bias) to 5 (most subject to social desirability bias) with an average score of 3.36.\\n\\nTable A.1 in the appendix shows that social desirability is not an important driver of reported levels of support for women working, shared childcare, nor for whether respondents’ mothers are supportive of women working. The coefficients on the social desirability bias when regressed on these variables are small and two out of three are not statistically significant. Respondents’ reported perceptions of other’s support (columns 4 to 6) are again not heavily influenced by the SDI.<sup>25</sup> Below we examine whether social desirability – bias affects our outcome measure the choice of the career mentoring course.\\n\\n# **5.2 Intervention Impacts**\\n\\nTo estimate the impact of the provision of information on social norms, we estimate regressions of the following form: where _Yi_ is the outcome variable (choice of online career mentoring); _Ti_ is a vector of treatment arm indicators (relative to the omitted control group); _Xi_ is a vector of control variables (gender, education, social desirability bias index) and _εi_ are robust standard errors.\\n\\nTable 2 shows the results of estimating equation 1 over the real rewards sample.<sup>26</sup> Column 1 presents results for the pooled sample of male and female respondents. The treatments significantly increase the probability of respondents choosing to be compensated for their time by receipt of the online career mentoring course, rather than the shopping voucher. This is the case in all treatment arms (relative to control). Respondents are between 7 and 10 percentage points more likely to choose the career mentoring course. This is a 20% to 29% increase over the control mean of 0.343. Columns 2 and 3 present results separately for female and male respondents, respectively. All treatment arms have significant effects for both men and women.\\n\\nAlthough the point estimates of the interventions’ impacts differ across treatment arms, increasing from 0.067 in treatment 1 through to 0.98 in treatment 3 (column 1),\\n\\n> 25Interestingly, respondents’ perceptions of the extent of support are heavily influenced by their own level of support.\\n\\n> 26We report results for the real rewards sample only and go on to explore the effect of using a hypothetical choice as the outcome variable further below. Table A.2 in the appendix reports results without controls. The inclusion of controls has very little effect on the results.\\n\\n13\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"they are not statistically significantly different from one another. The only difference across treatment arms that approaches statistical significance is the difference between treatments 1 and 3 for women. The point estimate for the impact of treatment arm 3 is 5.1 percentage points larger than for treatment arm 1 (p=0.12). This is suggestive that information on men’s support for shared childcare and mothers support for working women may have additional salience for women.\\n\\nColumns 4 to 9 present results separately by whether the woman (female respondents and wives of male respondents) works or not. These results are informative as to whether the treatments are likely to increase women’s labor force participation. The point estimates are uniformly larger in the sample where the woman is not working at the time of the survey (column 4 vs column 7). The interventions increase the probability of the career mentoring course being selected by 8.2 to 12.1 percentage points. These are very large impacts (27% to 39%). Column 5 reports the results for non-working female respondents and Column 6 for male respondents with non-working wives. The results are stronger for male respondents who have a non-working wife (possibly due to the smaller sample size of non-working female respondents leading to imprecise estimates). Men with non-working wives are 8.2 (26%) to 14.7 (47%) percentage points more likely to choose the career mentoring course for their wife than a shopping voucher. The point estimates on the treatment indicators are, however, very similar in magnitude across Columns 5 and 6, with the exception of the coefficient on the indicator for Treatment 2. The point estimates are largest for male respondents in treatment 2 (where respondents receive information on social norms about men’s support for sharing childcare responsibilities in addition to information on women’s support for married women with children working), suggesting that the information on the extent of support among their male peers for shared childcare increased their support for their wife working beyond the impact of the information on women’s support for working women, but not statistically significantly so.\\n\\nColumns 7 to 9 report results for the sub-sample where the women are working. The point estimates are a similar magnitude for both female and male respondents, but more strongly significant for female respondents. Being exposed to treatments 2 and 3 increases the probability of a female respondent who works selecting the career mentoring course by 8.6 and 10.8 percentage points, respectively, relative to the control group. The difference between the impact of treatments 1 and 3 again approaches statistical significance (p=0.11).\\n\\nIn terms of the control variables, tertiary education appears to play little role in respondents’ choices, apart from tertiary educated men who are married to a working woman being more likely to choose the career monitoring course than non-tertiary educated men with working wives (7 ppts). Social desirability bias does not affect the reward choice in these real reward interventions.<sup>27</sup>\\n\\n> 27Interactions of the social desirability index with the treatment indicators are also not statistically\\n\\n14\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"Panel B in Table 2 reports results where the treatment impacts are restricted to be equal across treatment arms. Across the entire sample of male and female respondents (column 1), treatment increases the probability of choosing the career mentoring course by 8.6 percentage points (25%). Large treatment impacts are detected for all groups other than non-working women (the point estimate on treatment is similar in magnitude to other groups but not statistically significant).\\n\\nTo address the issue that our sample oversamples working women, Table A.3 in the appendix reports results of regressions in which the sample is reweighted to be representative of the population in the large urban centers from which our respondents are sampled, i.e., with lesser weight being applied to working women and greater weight to non-working women. The point estimates are largely unchanged.<sup>28</sup>\\n\\n# **5.2.1 Intensity of the Treatment**\\n\\nIn this section we examine whether spending more time reading the information presented in the treatments – arguably a sign of more deeply engaging with the information - is associated with a greater likelihood of choosing the career mentoring voucher. If so, this provides further evidence that the information provided is driving participants’ choices. Table 3 includes a variable for the total time spent on the intervention pages.<sup>29</sup> Column 1 shows that for each additional 10 seconds spent looking at the intervention screen, the respondent’s probability of selecting the career voucher increased by 0.9 percentage points (2.6%). For respondents in a couple in which the woman does not work (Column 2), ten more seconds spent reading the intervention material increases the probability of the respondent choosing the career mentoring course by 1.6 percentage points (5.2%). Reading time has no effect for respondents in a couple in which the woman already works (Column 3).\\n\\n# **5.2.2 Heterogeneity Analysis**\\n\\nWe now explore the heterogeneity of treatment effects. One might expect larger treatment effects in households which are more open to attitudinal change, possibly through greater exposure to role models of working women, and who have real prospects of being able to\\n\\n> significant. We also estimated specifications including controls for whether the respondent is supportive of women working outside the home for pay. The coefficients are similar to the ones reported here as the level of support prior to treatment is balanced across control and treatment groups. Results available upon request.\\n\\n> 28Weights for (non-)working women were calculated as the number of (non-)working women in the population over the number of (non-)working women in the sample. The population numbers were calculated from the 2020 National Socioeconomic Survey (SUSENAS).\\n\\n> 29On average people spent 19 seconds reading the information in treatment 1, 26 seconds in treatment 2 and 35 seconds in treatment 3 (as the format of the information is similar in each treatment, the time spent accelerates). No one spent more than 2.5 minutes reading the information. We also estimated a specification with a quadratic term but its coefficient was not significant.\\n\\n15\", \"datasets\": [{\"mention_name\": {\"text\": \"2020 National Socioeconomic Survey\", \"confidence\": 0.8306622505187988, \"start\": 2768, \"end\": 2802, \"clean_text\": \"2020 National Socioeconomic Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.4495883285999298, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2768, \"end\": 2802, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2768, \"end\": 2802}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2768, \"end\": 2802, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.5380287766456604, \"start\": 398, \"end\": 432}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8306622505187988, \"start\": 398, \"end\": 432}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.925832986831665, \"start\": 398, \"end\": 432}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.925832986831665, \"start\": 398, \"end\": 432}}, {\"mention_name\": {\"text\": \"SUSENAS\", \"confidence\": 0.35984542965888977, \"start\": 2804, \"end\": 2811, \"clean_text\": \"SUSENAS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2804, \"end\": 2811, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2804, \"end\": 2811, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2804, \"end\": 2811}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2804, \"end\": 2811, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.5380287766456604, \"start\": 434, \"end\": 441}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.35984542965888977, \"start\": 434, \"end\": 441}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.925832986831665, \"start\": 434, \"end\": 441}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.925832986831665, \"start\": 434, \"end\": 441}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"find and retain a job. To this end, we examine whether treatment effects are larger for respondents: 1) whose mothers worked when they were age 12; 2) in regions with more job opportunities (a larger formal sector); and 3) with less binding childcare responsibilities (those with children aged 6 or older).\\n\\nTable 4 shows that when we compare individuals whose mothers worked (column 1) with those whose mother did not (column 2), we find little difference in treatment impact, suggesting that the absence of positive role models in childhood does not hamper the treatment’s effectiveness. The intervention impact is also similar in areas where there is high and low availability of formal employment opportunities (columns 3 and 4), showing that labor market opportunities do not affect the effectiveness of the intervention. There is however significant heterogeneity in intervention impact across families with and without pre-school children. Much larger effects are found for families in which women are not constrained by the presence of young children (p=0.05). For individuals who are in a household where the youngest child is over 6, participation in any of the treatments increases the probability of choosing the career mentoring course by 16 percentage points. This is a 55% increase relative to the control group. The treatment is effective, but less so, for respondents with younger children – 7.1 percentage point impact (20%). Once children start primary school, women are more likely to be able to look for work, reflecting the limited access to pre-primary school childcare in Indonesia.<sup>30</sup>\\n\\n# **5.2.3 The Role of Misperceptions**\\n\\nTo further explore the way in which the interventions affected participant choices, following the previous literature, we examine whether those who underestimated the level of support in the community were more greatly impacted by the interventions (than those who estimated the level of support correctly or over-estimated it). Figure 5 shows the extent of misperceptions in the three domains of the interventions. The red line shows the actual level of support in the community (as measured in our first survey or the World Values Survey for older women’s attitudes). All three figures show that there was substantial, and varying, misperceptions in all three domains.\\n\\nTable 5 presents results where we interact treatment with the extent to which respondents underestimated or overestimated the extent of community support for the social norms relevant to their treatment arm. For respondents in treatment arm 1 we allow the treatment impact to differ with the extent of over or underestimation of support for working women. For respondents in treatment 2 we allow for participants’ under/overestimation of support for working women and for shared childcare to affect the treatment\\n\\n> 30These heterogeneity tests were not pre-specified. The pre-analysis plan specified tests of heterogeneity of treatment effects in relation to gender (presented in the main results above); education; religiosity; female respondents’ and male respondents’ wives’ pre-intervention labour force participation; preintervention attitudes to working women and shared childcare; and reasons the woman in the couple is not working. No heterogeneity was found along these dimensions. Results available on request.\\n\\n16\", \"datasets\": [{\"mention_name\": {\"text\": \"World Values Survey\", \"confidence\": 0.8411363959312439, \"start\": 2181, \"end\": 2200, \"clean_text\": \"World Values Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2181, \"end\": 2200, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2181, \"end\": 2200, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2181, \"end\": 2200}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2181, \"end\": 2200, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9970923662185669, \"start\": 585, \"end\": 604}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8411363959312439, \"start\": 585, \"end\": 604}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5776203274726868, \"start\": 585, \"end\": 604}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5776203274726868, \"start\": 585, \"end\": 604}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"impact. We do the same for respondents in treatment 3 but also allow their under/overestimate of the extent of mothers’ support for working women to affect the treatment impact.\\n\\nTable 5 shows that there is little relationship between misperceptions of the social norm and the impact of the treatment. The interactions between treatment and misperceptions are insignificant in all cases, except for men in treatment 3 where the coefficient is small, counterintuitively signed and significant only at the 10% level.\\n\\nThat the extent of misperceptions had little impact is a surprising result, contrary to the theory of change outlined above, and in contrast to the results in Burzstyn et al. (2020), Aloud et al. (2020) and Cortes et al. (2022). The demonstration that women’s support for working women, men’s support for shared childcare and mothers’ support for working mothers is greater than many perceive shifted participants’ views to be more supportive but the mechanism does not seem to be via the correction of misperceptions. Rather, the program impact was similar for people with different estimations of the extent of support.<sup>31</sup> This result suggests that merely highlighting majority community support can change social norms and behavior across the community. This is a positive finding in the sense that it suggests one does not need to worry about the provision of information on a norm reducing support among those who estimate there are higher levels of support than the prevailing norm. Rather it seems that a demonstration of strong community support for a behavior (i.e., women with children working) encourages increases in support among those with lower levels of personal support while at the same time reaffirming the views of those who already believed that there was strong community support.\\n\\n# **5.2.4 Potential Long-Run Effects**\\n\\nTo test if the intervention is likely to have long lasting effects, we track whether female respondents who selected the career mentoring course and wives of male respondents who selected the course actually enrolled in the course. Seventy-two percent of those who selected the course enrolled within the one-month period after the intervention for which the offer was valid. Table A.4 shows the effect of the interventions on the probability of choosing the voucher and using it.<sup>32</sup> The results remain strongly statistically significant and are very similar to the main results presented in Table 2. This suggests that the majority of respondents were sincere in their interest in the women’s career mentoring course\\n\\n> 31Similar results were found when using indicator variables of whether someone overestimated versus underestimated the extent of support.\\n\\n> 32A total of 86% of the vouchers for the career mentoring course were redeemed. Of the redeemed vouchers - all female respondents used the voucher themselves and 67% of male respondents enrolled their wife in the course, with 33% claiming the course for themselves rather than giving it to their wives as intended. In the latter case, we code the voucher as “not being used”. (The career mentoring course was presented as being for their wives when respondents made their choice but in practice we were unable to enforce this.)\\n\\n17\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"and that these light-touch interventions are capable of generating meaningful behavioral change.\\n\\n# **5.2.5 Is It Necessary to Use Real Reward Payoffs?**\\n\\nWe now examine whether similar results are observed when the choice over the career mentoring course for women is hypothetical. Two of Qualtrics panel partners who supply Qualtrics with respondents (constituting 20% of respondents) would not allow us to randomly select respondents to receive a reward as this went against their contract with respondents. As those making the hypothetical choice were not randomly selected, Table A.5 in the appendix compares the characteristics of respondents in these panels with the other respondents. The respondents who are given the hypothetical choice are demographically very similar to the other respondents. Demographic differences are small in magnitude and not strongly statistically significant. Respondents in the hypothetical sample however report that they and their family members (spouses and mothers) are more supportive of women working and husbands sharing daily childcare. The magnitude of the differences are relatively small but we nevertheless control for these variables in the specifications below (doing so does not affect the results).\\n\\nColumn (1) of Table 6 reports results for the entire sample (men and women). Although participants faced with a hypothetical choice were 7.9 percentage points more likely to choose the career mentoring course than participants who were making a real choice (48.6% versus 40.7%), this result is driven by the interaction between making a hypothetical choice and treatment, rather than the choice being hypothetical, per se. Respondents in treatment arms where the choice was hypothetical were 22 percentage points less likely to choose the career mentoring course over the shopping voucher. This is observed for both men and women (columns 2 and 3). The hypothetical choice also interacts with social desirability bias. If their choice is hypothetical, participants who are most concerned about appearing to behave in a socially desirable way (social desirability index = 5) are 28 percentage points (72%) more likely to choose the career mentoring course than those who are the least concerned (social desirability index=0).\\n\\nColumns (2) and (3) present the results of estimating the same models on the sample of female respondents and male respondents, respectively. The results are broadly similar for men and women. One difference is that women are more likely to choose the career mentoring course per se when the choice is hypothetical, whereas men are more susceptible to social desirability bias (possibly because male respondents may feel more social pressure to support working women than women who would be seen to be acting in their selfinterest).<sup>33</sup> Columns (4) to (6) present the results estimated over only the sub-sample of respon-\\n\\n> 33The interaction between social desirability bias and treatment and the triple interaction between treatment, hypothetical choice and the social desirability index were both insignificant in all specifications.\\n\\n18\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"dents who made a hypothetical choice. In the absence of a real reward choice, we would have concluded that treatment _decreases_ the probability of female participants choosing the career mentoring course and has no effect on male respondents. Hence, the use of a real, meaningful outcome choice is essential to the identification of intervention impacts. The choice need not have real world consequences with 100% probability. In our case offering a one-third probability of the choice being real created sufficient salience for the decision to be taken more seriously.<sup>34</sup> The greater susceptibility of men to social desirability bias in this setting is again evident.\\n\\n# **6 Discussion and Conclusions**\\n\\nOur results demonstrate that a light-touch, inexpensive intervention that seeks to change gender norms around women’s work can be effective in a middle-income, majority moderateMuslim nation like Indonesia. We have thus broadened the scope of the evidence of the impact of interventions that seek to address correct misperceptions beyond studies in the developed world and the very specific context of the Middle East. The ability of such interventions to change men’s attitudes is especially important as in our sample 20% of women who were not working at the time of the survey reported that this was because their husband does not wish them to, while the percentage of women who reported that they were not working because they do not wish to is very low at 1%.\\n\\nGiven the percentage increase in men’s support for working women attributed to the intervention (25%) and assuming the same elasticity of female labor supply to men’s support as found in Bursztyn et al. (2020), we estimate that our intervention could increase Indonesian female labor supply by as much as 6 percentage points (12%).\\n\\nThe finding that this light touch, low-cost, easily scalable online intervention was able to change behavior so that participants made choices consistent with an aspiration for either themselves (for female respondents) or their wives (for male respondents) to work is promising in terms of the likely effectiveness of public information campaigns that demonstrate community support for working women. These could be in the form of TV commercials, billboards or social media posts. Mass media has been shown to influence norms in a variety of contexts and on a variety of issues (Kearney and Levine, 2015; La Ferrara et al., 2012; La Ferrara, 2016; Banerjee et al., 2019). The greater understanding of existing social norms towards working women in Indonesia generated here will also be useful for the formulation of such campaigns.\\n\\n> 34This result is seemingly at odds with the findings of Cort ́es et al. (2012)) who find that incentivizing their outcome measure of how much of $100 to donate to a non-profit organization that advocates for women in the workplace (with the remainder being forfeited) has no effect. They find no difference in results when this measure is incentivized by respondents being informed that five survey participants would be randomly selected and their choices implemented. The difference in their results may reflect the relatively low value of the incentives they employ (only applying to five people in their sample of over 1500 participants and there being no direct payoff to the participant).\\n\\n19\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"We show that perceptions of others’ attitudes play a substantial role in husband’s attitudes to women working and to shared childcare, especially perceptions about husbands’ status. Of women who viewed their husbands as not being supportive of them working, 65% report that if they were to work, their husband would be worried what other people would think about him - either about his capacity to provide for his family (49%) or that his wife does not respect him (20%). These concerns are likely to be alleviated if he knows that there is wide-spread support in the community for women working. Campaigns that further try to dispel the link between wives working and a husband’s financial capacity are likely to be especially successful.\\n\\nFinally, that the extent of women’s support for working women is underestimated suggests an alternative policy approach. Empowering women to speak up and voice their opinions in the home, and more widely, may influence social norms just as effectively as a public information campaign. This is an area worthy of further research.\\n\\n20\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099424001232435449/pdf/IDU1edba8d7d10b7214eb11bfe31bb7ad3731c0c.pdf\", \"pages\": [22]}}]}"}}},{"rowIdx":87,"cells":{"doc_id":{"kind":"string","value":"34244968"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34244968\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"authors\": {\"0\": {\"author\": \"Feriga,Moustafa Amgad Moustafa Ahmed Moustafa\"}, \"1\": {\"author\": \"Lozano Gracia,Nancy\"}, \"2\": {\"author\": \"Serneels,Pieter Maria\"}}, \"count\": \"World\", \"count_key\": \"517191\", \"docna\": {\"0\": {\"docna\": \"The Impact of Climate Change on Work :\\n Lessons for Developing Countries\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"LAC Regional Director (SLCDR)\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34244968\"}, \"repnb\": \"WPS10682\", \"docdt\": \"2024-01-24T00:00:00Z\", \"datestored\": \"2024-01-24T06:42:00Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"What is the impact of climate change on\\n labor? Reviewing the evidence, this paper finds five areas\\n of potential impact. Climate change may have an immediate\\n effect on labor demand, labor supply and time allocation,\\n on-the-job productivity, and income and vulnerability among\\n the self-employed. In the medium term, climate change may\\n lead to a reallocation of labor across economic activities\\n and across space. Impact estimates typically rely on fixed\\n effect estimation. These estimates require care when\\n interpreted as they typically reflect the short-term direct\\n impact of past events and abstract from potential\\n adaptation. The paper discusses emerging work trying to\\n address this, analyzing the responses by firms, farms,\\n households, and workers. Together, the existing evidence\\n points toward six potential areas of government response.\\n Potential labor policies include green jobs, green skills,\\n labor-oriented adaptation, flexible work regulation, labor\\n market integration, and social protection. The paper\\n concludes by setting out avenues for future research in this field.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10682; PLANET\", \"display_title\": \"The Impact of Climate Change on Work\\n : Lessons for Developing Countries\", \"disclosure_date\": \"2024-01-24T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099842001242436996/text/IDU13fb2bcd51c7f1140371b61811281e0957f86.txt\", \"listing_relative_url\": \"/research/2024/01/34244968/impact-climate-change-work-lessons-developing-countries\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099842001242436996/IDU13fb2bcd51c7f1140371b61811281e0957f86\", \"new_url\": \"2024/01/34244968/IDU13fb2bcd51c7f1140371b61811281e0957f86\", \"ext_pub_date\": \"2024-01-24T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Labor Markets m1326488 855,Climate Change m1326829 567,Climate Change Adaptation m1327767 574,Labor Supply m1504055 2751\", \"docm_id\": \"IDU-3fb2bcd5-c7f1-4037-b618-1281e0957f86\", \"chronical_docm_id\": \"IDU-3fb2bcd5-c7f1-4037-b618-1281e0957f86\", \"publishtoextweb_dt\": \"2024-01-24T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10682\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"guid\": \"099842001242436996\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"56\", \"author_keys\": {\"0\": {\"author_key\": \"2028568\"}, \"1\": {\"author_key\": \"611031\"}, \"2\": {\"author_key\": \"320793\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099842001242436996/IDU13fb2bcd51c7f1140371b61811281e0957f86\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10682 The Impact of Climate Change on Work Lessons for Developing Countries _Moustafa Feriga Nancy Lozano Gracia Pieter Serneels_ Sustainable Development Practice Group January 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10682\\n\\n# **Abstract**\\n\\nWhat is the impact of climate change on labor? Reviewing the evidence, this paper finds five areas of potential impact. Climate change may have an immediate effect on labor demand, labor supply and time allocation, on-thejob productivity, and income and vulnerability among the self-employed. In the medium term, climate change may lead to a reallocation of labor across economic activities and across space. Impact estimates typically rely on fixed effect estimation. These estimates require care when interpreted as they typically reflect the short-term direct impact of past events and abstract from potential adaptation. The paper discusses emerging work trying to address this, analyzing the responses by firms, farms, households, and workers. Together, the existing evidence points toward six potential areas of government response. Potential labor policies include green jobs, green skills, labor-oriented adaptation, flexible work regulation, labor market integration, and social protection. The paper concludes by setting out avenues for future research in this field.\\n\\nThis paper is a product of the Sustainable Development Practice Group. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The author may be contacted at nlozano@worldbank.org.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"The Impact of Climate Change on Work: Lessons for Developing Countries Moustafa Feriga, Nancy Lozano Gracia, Pieter Serneels<sup>+</sup> JEL Codes: Q54, J01, O1 Key words: Climate Change, Labor, Development\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **1. Introduction**\\n\\nWork is the main source of income for most people around the world, and in particular for the poor. As a primary factor of production, labor also plays a key role in economic growth. Yet our understanding of how climate change impacts labor remains scattered. This paper takes stock of what we know, distinguishing five ways in which climate change may affect labor. We discuss limitations and challenges with existing estimates, underlining the need for careful interpretation. A central issue of concern is the omission of adaptation, and we discuss recent insights on the responses by firms, farms, households and workers. Together, the evidence suggests several potential areas of government response. We discuss possible labor policies and ways forward for research.\\n\\nA growing literature establishes the impact of climate change on general economic performance. Weather variation – measured in terms of temperature, precipitation, or climate change related weather events – is estimated to lower economic output, with the largest expected impacts for low- and middle-income countries.<sup>i</sup> Changes in weather also increase or deepen poverty, in particular in rural and exposed areas, and among vulnerable groups.<sup>ii</sup> Labor represents a possible key channel through which climate change affects economic performance and poverty. Reviewing the most rigorous existing evidence, we distinguish five potential areas for the impact of climate change on labor: labor demand, labor supply and time allocation, on-the-job productivity, and income and vulnerability among the self-employed. In the medium-term climate change may lead to a reallocation of labor across economic activities and across space.\\n\\nThe studies we consider follow a common estimation strategy, with impact estimates typically obtained through fixed effect estimation. These estimates generally capture the short-term, immediate effects of past events and do not account for potential adaptation. They might therefore overestimate long-term effects, in particular if substantial adaptation is anticipated. Conversely, they might underestimate long2\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"term impacts, when they fail to consider effects across the economy, or if climate change worsens, i.e. realized past events are a poor predictor for future occurrences. While we mention research aimed at tackling the latter two concerns, mainly through general equilibrium analysis and simulation of future outcomes, where appropriate, our primary interest lies in the emerging adaptation research that examines responses by firms, farms, households and workers.\\n\\nThis leads us to the question how governments can, and do, respond. There has been limited research investigating the impact of general climate change policies on labor outcomes, highlighting the need for more thorough examination. In the realm of labor-targeting climate change policies, we distinguish six potential areas where governments can take action: green jobs, green skills, labor-oriented adaptation, flexible work regulation, labor market integration, and social protection.\\n\\nOur review demonstrates the impact of climate change on labor to be a growing area of research, where much is to be discovered. We conclude by setting out avenues for future research.\\n\\nSection 2 reviews the evidence, starting with a note on methods used and challenges of identification, before distinguishing the five ways in which climate change may impact labor. Section 3 discusses how agents respond, commencing with an examination of analytical methods, followed by a discussion of how firms, farms, households and workers respond. A subsequent section examines government response. The concluding section discusses ways forward for research and the evaluation of policies in this field.\\n\\n# **2. The impact of climate change on labor**\\n\\n## **2.1. Methods and identification**\\n\\nThis overview focuses on studies that estimate the impact of climate change on labor outcomes using regression analysis, typically a variation of the following form: 3\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"Where i is the labor variable of interest, and is the impact of climate change – often temperature – ii in area YY _j_ at time _t_ , to which worker _i_ (in micro studies) or sector or country ββ _i_ (in meso or macro studies) is exposed ( i ). Some papers look beyond temperature, at precipitation, or other climate change events. ee ii i reflects other climate change events that may be included or not in the regression. TT i are individual ee ii ii CCworker (or sector or country) characteristics that affect i . and are unobserved individual and XX ii ii ii location fixed effects respectively (which are combined in some studies), and YY μμ μμ i is a time trend.<sup>iii</sup> This estimation typically combines longitudinal data on economic outcomes<sup>iv</sup> with climate data from weather γγ station, gridded, satellite, or other secondary sources and exploits short-run, _plausibly_ exogeneous variation in climate to estimate impact on labor outcomes of interest.\\n\\nStudies differ in the variables they include on the right-hand side. Research focusing on temperature typically control for other climate events like precipitation. The observed subject characteristics that are included as control variables also vary. Including fixed effects and time trends yields a more robust estimate.\\n\\nThe next sections zoom in on specific left hand side labor variables of interest, one at a time: demand, supply, productivity, income, allocation, and migration. The studies on allocation compare estimates from across different sectors. Throughout the estimations, challenges remain in obtaining estimates of that reflect causal impact, given the non-experimental nature of the data. We see five issues of concern. ββ First, the temperature (or other climate event) to which workers are actually exposed remains unobserved. Locally measured temperature (or climate change event) serves as a proxy.<sup>v</sup> This will affect precision, and can lead to biased estimation, if the exposed temperature is very different from the observed one.<sup>vi</sup> In general, studies seem to be aware of this concern and pay attention to avoid major mismeasurement of this nature.\\n\\n4\", \"datasets\": [{\"mention_name\": {\"text\": \"longitudinal data\", \"confidence\": 0.39499378204345703, \"start\": 718, \"end\": 735, \"clean_text\": \"longitudinal data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 718, \"end\": 735, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 718, \"end\": 735, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 718, \"end\": 735}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 718, \"end\": 735, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8837684392929077, \"start\": 718, \"end\": 735}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.39499378204345703, \"start\": 718, \"end\": 735}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6719201803207397, \"start\": 718, \"end\": 735}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6719201803207397, \"start\": 718, \"end\": 735}}, {\"mention_name\": {\"text\": \"climate data\", \"confidence\": 0.5650633573532104, \"start\": 775, \"end\": 787, \"clean_text\": \"climate data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 775, \"end\": 787, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 775, \"end\": 787, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 775, \"end\": 787}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 775, \"end\": 787, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8837684392929077, \"start\": 775, \"end\": 787}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5650633573532104, \"start\": 775, \"end\": 787}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6719201803207397, \"start\": 775, \"end\": 787}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6719201803207397, \"start\": 775, \"end\": 787}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"A second limitation is that the studies estimate historic and short term, immediate, impacts. Firms, workers, markets and regulators, may adapt to higher temperature and other climatic changes, possibly reducing their impact on labor. The extent to which adaptation takes place is a topic of increasing attention and the subject of a nascent literature, which we discuss in the section on response. In general, our understanding in this area remains limited. Adaptation will also be heterogenous in form and in the time it takes to respond. In some countries (low income), areas (rural), and sectors (agriculture, outdoor) adaptation is likely to be slower. Nevertheless, because accounting for adaptation remains a challenge, most of today’s estimates need to be seen as short-term impacts. They may be an upper bound if there is increased adaptation in the future; they may be a lower bound if changes in climate become more severe, leading to higher impacts.\\n\\nTo address the uncertainty about future trends, a separate strand of papers looks at projections for the future, using simulation-based approaches starting from predicted temperature and current economic outcomes. They use a distinct type of analysis from the one set out above. Their inference is informative, providing a sense of the general trends, albeit under specific assumptions (Stern, 2008). These estimates are not the focus of attention in this paper, although we include the results from a few overview studies because of their informative nature – we mention these explicitly as simulation studies.\\n\\nA third shortcoming is that the estimates are obtained for partial equilibrium. General equilibrium effects – changes across the economy that come about from linkages between economic activities – are neglected. Specific studies try to shed light on GE effects using distinct methodologies, including Computable General Equilibrium (CGE) and Dynamic General Equilibrium models or Input-Output (I/O) analysis. While these studies are useful to get some idea of data patterns, their results are quite sensitive to the underlying assumptions.<sup>vii</sup> We mention these studies where useful, although they are not central to our overview.\\n\\n5\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"Fourth, while some research discusses potential omitted variables, other studies don’t. Yet in the absence of experimental data, estimates may be biased due to omitted variables. One key unobservable of high importance is the capacity for adaptation. Firms or workers with some characteristics – e.g. more profitable, wealthier – may be quicker at taking adaptive measures, thereby reducing impact. Other omitted variables, including climate variables, may also matter.<sup>viii</sup> Fifth, while location fixed effects help to obtain a more robust impact estimate, they do not shed light on the salient aspects of context. Which characteristics of the local economy and labor market matter most, is important for policy design. One way to address this is to carry out subsample analysis, breaking down the results for different subgroups or localities along pre-meditated dimensions of interest. Another approach is to focus a study on specific setting, which allows explicitly taking into account salient characteristics, leading to richer interpretation, and highlighting the gaps, identifying key directions for future work.<sup>ix</sup> Despite these limitations, the estimated impacts are the best we have for now. Comparison across the most rigorous studies can help assess the robustness of these estimates. Nevertheless, critical assessment and meticulous interpretation remain key, as discussed throughout the next sections.\\n\\n# **2.2. Labor demand**\\n\\nClimate change is expected to affect the demand for labor through diverse channels. Where climate change reduces output, as the evidence suggests for specific sectors, including agriculture, derived demand for labor will typically go down. At the same time, as the world moves toward greener production, demand for labor in carbon intensive activities, like nonrenewable energy, is anticipated to decline, while that in green activities is set to increase. As it stands, we possess an incomplete comprehension of the net effect of climate change on labor demand. To gain a clearer understanding, sectoral analysis is needed. In what follows we provide a summary of expected climate impacts on labor demand in selected sectors.\\n\\n6\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"# **2.2.1. Agriculture**\\n\\nAgricultural output can be affected in several ways. Growing evidence documents how increased heat reduces agricultural output in low- and middle-income countries, varying between 5% and 25%, depending on the crop, region, and time span considered.<sup>x</sup> Increased rainfall is often positively related to agricultural output.<sup>xi</sup> Exposure to cyclones bears negative impacts.<sup>xii</sup> Direct impacts on the demand for labor have been observed across several developing country settings. In Mexico extreme heat is associated with a 1.4% decline in rural employment over a 28-year period, driven by weather-impacted agricultural yield losses. Reductions are largest for non-farm labor, as a result of the reduced demand for non-agricultural goods (Jessoe et al., 2018).<sup>xiii</sup> In El Salvador total corn production declined by as much as 2.8% for an additional week of high temperature during the harvest season (Ibanez et al., 2022). Agricultural producers adjusted by using more household labor and fewer hired workers, who either migrate or change to a non-agricultural job. In Brazil a 1-standard deviation increase in excess aridity reduced planted and harvested areas by 1.6% and 2.7%, respectively, during 2000-2010. Agricultural employment fell by around 11% in line with the negative impact on agriculture output (Albert et al., 2021). In China, a 1<sup>_◦_</sup> C increase in temperature is associated with 7% reduced farm work across rural communities (Huang et al., 2020).\\n\\nReduced demand for labor may especially affect women. In India, after a drought shock, women were 7% less likely to be in employment compared to men (1.2 percentage points for women versus 0.5 percentage points for men), spend 29% more days seeking work, and have 20% lower non-farm workdays; those in farm work see their real earnings fall by 38% compared to 3.4% for men (Afridi et al., 2022).\\n\\nMost studies in agriculture suffer from at least two of the limitations mentioned in section 2.1. First, they abstract from adaptation behavior of farmers, which is likely to temper long term impacts A handful of studies accounts for adaptation and consider multiple seasons, but the findings are disparate. Chinese 7\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"farmers use more machines to compensate for the shortage of labor arising from extreme hot weather, and this offsets nearly 47% of the short-run effects on agriculture yield over a 35-year period (Chen and Gong, 2021). One attractive method to study adaptation exists of comparing short-term panel estimates to long term difference estimates. Exploiting large variation in weather across time and counties in the US, one study finds minimal long-run adaptation in terms of unchanged agricultural inputs and staying with same crop, while the cultivated area declined (Burke and Emerick, 2016). The limited adaptation is understood to arise from farmers either not being fully aware of a changing climate, or recognizing the need for adaptation but unable to implement adjustment.\\n\\nA second shortcoming is that these studies rarely look at labor outcomes beyond agriculture, and in doing so, ignore labor mobility as a margin of adjustment. When labor is mobile (across sectors or geographies) long-run impacts may be considerably smaller compared to the estimated short-run effects. Section 2.6 zooms in on the reallocation of labor.\\n\\n# **2.2.2. Industry**\\n\\nClimate impacts on non-agricultural output tend to be more pronounced for developing countries, and stem primarily from hot weather and extreme heat. Overall, industrial value-added is found to shrink by 2% for every 1<sup>_◦_</sup> C warming, but only in low-income countries (Dell et al., 2012; Hsiang, 2010). Precipitation effects are less robust, though there is variation by region (Dell et al., 2012). These impacts seem to affect industrial exports from low-income to high-income economies, which decline by 2.4% for a 1<sup>_◦_</sup> C temperature rise but do not vary with changes in precipitation (Jones and Olken, 2010).\\n\\nEmployment in heat exposed industries has received special attention in the literature. The few studies for developing countries typically concentrate on output, rather than employment, and find that output elasticity of labor is a primary mechanism for the observed negative effects of higher temperature. In India a 1<sup>_◦_</sup> C warming in daily temperature depresses annual firm output by 2.1% across manufacturing nationwide over 15 years; this seems to be driven by reductions in the output elasticity of labor, rather 8\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"than capital or other factors of production (Somanathan et al., 2021). Firm-level evidence from China confirms high temperature to affect sectoral output due to heat exposure: an additional hot day ( > 90<sup>_◦_</sup> F _,_ 32<sup>_◦_</sup> C) negatively affects output of almost half of the two-digit industrial sectors in China, including both labor-intensive and capital-intensive sectors, with notable heterogeneity in the magnitude of impact. For example, in timber manufacturing, output is expected to fall by 1.26% due to an additional hot day, while in nonmetal mining it would reduce by 1%. Impact is insignificant in medicine manufacturing and smelting of nonferrous metals, among others. High-tech industry output is insensitive to temperatures between 80<sup>_◦_</sup> F–90<sup>_◦_</sup> F (27<sup>_◦_</sup> C–32<sup>_◦_</sup> C) but sensitive to increase beyond that, while low-tech industry output is negatively impacted above any of these temperatures. These effects are large in economic terms: when assuming no additional adaptation, Chinese manufacturing output would be expected to fall by 12% annually (Zhang et al., 2018). These findings are in line with in-depth research for the US on employment loss in heat-exposed industries (Behrer and Park, 2017).13F<sup>xiv</sup> The impact of cyclones on industrial output varies by sector and over time. In the Caribbean, one standard deviation increase in the exposure to cyclones is negatively related to output in mining and utilities (-0.9%), but positively related to construction after the event (+1.4%) (Hsiang, 2010). In India,<sup>xv</sup> these events reduced 3% of firm sales, destroyed 2% of firm assets, but had no impact on salaries or labor inputs across a panel of manufacturing firms between 1995 and 2006. Overall, these impacts tend to be short-lived and disappear after a year, with better-performing industries recovering faster (Pelli et al., 2023).\\n\\nAs with the research on agriculture, the above studies have two important caveats. First, they tend to concentrate on immediate, short term, impact. Second, they typically do not account for impacts outside the sector, which may be considerable. Section 2.6 9\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"# **2.2.3. Energy**\\n\\nDemand for energy is expected to increase with climate change, reflecting consumers’ short-term response and long-term adaptation.<sup>xvi</sup> In Mexico, each additional day > 90<sup>_◦_</sup> F (32<sup>_◦_</sup> C) increases monthly electricity consumption by 3.2%. As incomes rise, AC ownership increases with temperature by 3 and 27 percentage points, in cool and warm places, respectively, for a $10,000 increase in annual household income (Davis and Gertler, 2015).16F<sup>xvii</sup> Most research on the energy sector concentrates on the expected impact of the shift towards green energy, following the global move towards low carbon, rather than the impact of climate change events themselves. Replacing fossil fuel with renewable energy will lead to job destruction in carbon-intensive industries and job creation in low-carbon sectors.<sup>xviii</sup> Existing evidence suggests a positive net employment effect in the energy sector for this climate transition. Because changes in energy have immediate effects across multiple sectors, many of these studies take a wider sector approach and focus on simulation of future outcomes.\\n\\nAn ex-ante I/O simulation indicates that if 139 countries were to convert 100% to clean energy from wind, water, and solar by 2050, 27.7 million jobs could be lost in high-carbon sectors such as oil, gas, coal, and biofuel, while approximately 52 million full-time jobs could be created (25.4 million construction jobs and 26.6 million operation and maintenance jobs), resulting in a net gain of 24.3 million full-time jobs in the sector (Jacobson et al., 2017). Another study simulates the direct employment capability of power plants across the world and finds that renewables can generate between 0.1 to 4 jobyears per GWh on average versus 0.1 to 2.4 job-years per GWh for non-renewable power plants (Barros et al., 2017).\\n\\nA regional CGE model for South Africa that analyzes the impact of a smaller share of coal in the energy supply mix finds an anticipated decline in employment in coal-producing regions, accompanied by an increase in other regions due to the expansion of non-fossil fuel power generation, assuming labor 10\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"mobility is relaxed (Bohlmann et al., 2019). For the Middle East region, an Integrated Assessment Model (IAM) simulation finds that the large-scale deployment of renewables required to meet the region’s GHG emissions reduction targets, needs a direct workforce of up to 180,000 and indirect workforce of up to 115,000 jobs, in wind, solar Photovoltaic (PV), and Concentrated Solar Power (CSP), depending on how much of the technology is manufactured locally—with the number of direct jobs created per year nearly reaching 18,000 for wind and well over 90,000 for both PV and CSP (Van der Zwaan et al., 2013).\\n\\nWhile these simulation-based studies produce useful insights, their findings are highly sensitive to the model’s assumptions and specifications. Recent meta-analysis confirms the sensitivity of estimated net employment effects of the energy transition to the specific method applied.<sup>xix</sup> Rigorous econometric estimation, following a specification like equation (1), for the energy sector itself remains scarce. Simulation methods themselves are also open to improvement.<sup>xx</sup>\\n\\n# **2.2.4. Service**\\n\\nServices exposed to extreme weather conditions will be strongly affected. The transport sector is particularly vulnerable to heat extremes, heavy rains and flooding. Extensive adaptation of infrastructure, operations and service provision will likely be needed to mitigate impact (Sims et al., 2014). The net employment effect in this sector remains understudied globally. Evidence from the US indicates that transportation payroll expenses declined significantly in hotter than usual years, though it remains unclear whether this effect stems from reduced labor supply or decreased demand for transportation (Behrer and Park, 2017).\\n\\nTourism is another service exposed to climate risks. In South Africa, informal workers in tourism-related subsectors in provinces with high tourist activity face the greatest vulnerability to drought: overall employment in these areas declined by 1.2 percentage points, and transport sector employment declines by 0.3 percentage points, for a 1-standard deviation increase in measured drought (Gray et al., 2023).\\n\\n11\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"Similar to the energy sector, the transport industry is vulnerable to both climate-related conditions and the risks associated with climate transition – think economywide carbon pricing, climate-smart transport policies, new transportation solutions, among others. As one of the leading contributors to carbon emissions, responsible for 24% of CO2 emissions from fuel combustion on a global scale, the transport sector is expected to undergo a significant transformation (IEA, 2020). These long term and far reaching effects are often overlooked in analysis, but are crucial for shaping effective policies. Section 3.3.1 discusses examples of government response, including carbon tax and energy pricing, and their impact on employment.\\n\\n# **2.3. Labor supply and time allocation**\\n\\nChanges in climate may affect labor supply and time use in various ways. First, weather conditions may encourage worker absenteeism. In India, a rise in the 10-day average temperature by 1<sup>_◦_</sup> C WBGT<sup>xxi</sup> increased the probability of absenteeism among steel workers by 2% and among garment workers by 10%, but not for piece rate workers in cloth-weaving factories (Somanathan et al., 2021). Workers in Chinese manufacturing, however, are not more absent on hotter days (Cai et al., 2018).<sup>xxii</sup> Weather conditions may also encourage shirking. The quality of reported data among Demographic and Health Survey (DHS) interviewers declined significantly on hot days, with more missing responses per interview on average, while the total number of interviews conducted per interviewer remained unchanged (LoPalo, 2023).<sup>xxiii</sup> Second, workers may adjust the distribution of work time in response to weather shocks. In one example work time is reallocated from farm to off-farm work. Indian households that experienced a persistent annual decline in farm income due to drying up wells (caused by climate change) are 12 percentage points more likely to derive income from off-farm employment. These added earnings compensate for income lost, leaving total household revenue unaffected (Blakeslee et al., 2020). Neither work related migration nor employment in nearby villages, nor changes in agricultural practices, played a similarly large role in 12\", \"datasets\": [{\"mention_name\": {\"text\": \"Demographic and Health Survey\", \"confidence\": 0.4973125159740448, \"start\": 1389, \"end\": 1418, \"clean_text\": \"Demographic and Health Survey\"}, \"acronym\": {\"text\": \"DHS\", \"confidence\": 0.439634770154953, \"start\": 1420, \"end\": 1423, \"clean_text\": \"DHS\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1389, \"end\": 1418, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1389, \"end\": 1418}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1389, \"end\": 1418, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9973738193511963, \"start\": 550, \"end\": 579}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4973125159740448, \"start\": 550, \"end\": 579}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5981824994087219, \"start\": 550, \"end\": 579}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5981824994087219, \"start\": 550, \"end\": 579}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"offsetting the fall in income. Identical shifts to non-farm activity have been observed in rural China.<sup>xxiv</sup> Rainfall shocks induce analogous behavioral patterns. In Brazil more work hours are allocated to nonagricultural activities at times of negative rainfall shocks, with stronger effects for households in lowincome municipalities (Branco and Feres, 2021).\\n\\nIn some cases, workers seem to reallocate time between leisure and labor. Workers in heat-exposed industries in the US reduce their daily labor supply by as much as 1 hour at temperatures above 85<sup>_◦_</sup> F (29<sup>_◦_</sup> C); this reduced work time is primarily replaced by indoor leisure (Graff Zivin and Neidell, 2014).<sup>xxv</sup> In another US study male workers labor 30 minutes longer on bad weather or rainy days (up to 48 minutes in dry regions), and reduce work time on the day following rainy day to make up for the previous day’s lost leisure (Connolly, 2008). It is unclear whether shifting activities across time is as common in developing countries, as the existing evidence is mixed. In rural China, more growing-season Harmful Degree Days – defined as days with temperatures above 32<sup>_◦_</sup> C – increases the probability of a time shift to leisure for men (Huang et al., 2020). The earlier mentioned DHS interviewers work more hours per day during hotter than usual periods.<sup>xxvi</sup> Labor supply response may also vary by gender. Women tend to have less opportunity to diversify into other tasks at times of weather shocks, possibly exacerbating existing gender gaps in the labor market.<sup>xxvii</sup> Together these studies suggest that workers may adapt their labor supply in the short term in response to changes in climate. Most of the current studies abstract from adaptation, which remains understudied.<sup>xxviii</sup> A growing literature studies the possible mechanisms behind the observed short-term changes in labor supply. Early work, primarily based on macro analysis, conjectured the decline in the temperature sensitive agricultural sector as a major factor (Jones and Olken, 2010; Dell et al., 2012). Recent evidence underlines the importance of health (Deschenes, 2014).\\n\\n13\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"Exposure to extreme temperature has a direct impact on the body: it strains cardiovascular, respiratory, and cerebrovascular systems. This also leads to increased mortality risks. An additional hot day with a mean temperature above 36<sup>_◦_</sup> C increases. annual mortality in India by 0.75% (Burgess et al., 2011). In the US, the number of days with temperature above 90<sup>_◦_</sup> F ( > 32<sup>_◦_</sup> C) is associated with 0.11% increase in mortality, after accounting for adaptation at home (Deschenes and Greenstone, 2011), while annual mortality is estimated to increase by up to 3% by 2100.28F<sup>xxix</sup> Hot-day related fatalities fall with access to residential air conditioning (Barreca et al., 2016). Average temperature is also positively related to suicide rates and negatively related to mental well-being, with adaptation (like air conditioning) having little impact. A 1<sup>_◦_</sup> C increase in the monthly average temperature increases suicide rates in Mexico 2.1%, compared to 0.7% in the US (Burke et al., 2018).29F<sup>xxx</sup> The health effects of temperature are further aggravated by air pollution (Graff Zivin and Neidell, 2013) and humidity (Barreca, 2012), both of which negatively impact health directly. Not controlling for humidity in equation 1, leads to 0.9% lower mortality rates for the US, underestimating the potential cost of climate change.\\n\\nNatural disasters also lead to death. Those who were less physically strong, like children and older adults, and those who could not find help from a physically strong person, were less likely to survive the 2004 Indian Ocean Tsunami in Indonesia (Frankenberg et al., 2011).\\n\\nA sub-strand of the literature looks at long term effects on human capital, and finds that climate change represents a major threat to health and education during childhood, which are likely to affect employment outcomes later in life.<sup>xxxi</sup>\\n\\n# **2.4. On-the-job productivity, wages**\\n\\nSeveral studies examine how climate change affects on-the-job productivity. These papers tend to follow the approach set out in equation (1), but the level of observation differs, leading to estimation at the 14\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"micro, meso, or macro level, which each measure and define worker productivity in their own way. Micro studies focus on observed individual worker performance, output or wages as the left-hand side variable.<sup>xxxii</sup> Studies at the sector or economy wide level typically consider output per worker, using an aggregate economic indicator such as output, value-added, or GDP, divided by reported number of workers employed.xxxiii Increased temperature generally has an adverse effect on labor productivity, and this seem to hold across settings, specifications, levels of observation, and types of output. This is confirmed by a recent review of a subset of studies (Lai et al., 2023). Micro-level analysis in developing countries finds a negative relationship between productivity and temperature. In India’s manufacturing sector, a 1<sup>_◦_</sup> C increase on a hot day reduces labor productivity by 2–4% (Somanathan et al., 2021). For garment factories in India, productivity effects of relatively hot days are in the magnitude of -2 pp/+1<sup>_◦_</sup> C for temperatures above 27<sup>_◦_</sup> C-28<sup>_◦_</sup> C (Adhvaryu et al., 2020). Workers’ data from non-climate-controlled manufacturing firms in China suggest a U-shaped relationship: temperature is positively correlated with productivity up to 78<sup>_◦_</sup> F (~25<sup>_◦_</sup> C), after which productivity declines with further warming (Cai et al., 2018). Call center workers in India see a productivity decline of 1.8% for a 1<sup>_◦_</sup> C rise over a temperature range from 21 _._ 9<sup>_◦_</sup> C to 28 _._ 5<sup>_◦_</sup> C (Niemela et al., 2002).\\n\\nBecause individual worker productivity can be hard to measure, some research focuses on the change in firm Total Factor Productivity (TFP) obtained from estimating a production function. Firm-level analysis of 500,000 Chinese manufacturing plants finds heat stress to be strongly related to economic output (Zhang et al., 2018): a day with a temperature above 90<sup>_◦_</sup> F (32<sup>_◦_</sup> C) lowers TFP of the average manufacturing firm by 0.56% relative to a day with a temperature between 50◦ -60◦F (10<sup>_◦_</sup> C-15<sup>_◦_</sup> C), resulting in 0.45% lower output. Productivity losses tend to be larger for workers in heat-exposed industries.<sup>xxxiv</sup> In Chinese manufacturing, the largest losses in value added per worker on hotter summer days occur in industries where workers are highly exposed to heat while air conditioning is largely infeasible; these include ferrous 15\", \"datasets\": [{\"mention_name\": {\"text\": \"Workers’ data\", \"confidence\": 0.5824936032295227, \"start\": 1152, \"end\": 1165, \"clean_text\": \"Workers’ data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1152, \"end\": 1165, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1152, \"end\": 1165, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1152, \"end\": 1165}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1152, \"end\": 1165, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.3286136984825134, \"start\": 340, \"end\": 353}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5824936032295227, \"start\": 340, \"end\": 353}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5260384678840637, \"start\": 340, \"end\": 353}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5260384678840637, \"start\": 340, \"end\": 353}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"metal mining (-48.7%/+1<sup>_◦_</sup> C) and timber (-41.2%/+1<sup>_◦_</sup> C) (Chen and Yang, 2019). Losses in agricultural TFP in China are estimated at 2.6% for every additional day with a temperature of more than 33<sup>_◦_</sup> C during the year (Chen and Gong, 2021).\\n\\nThese findings are consistent with county-level payroll evidence from the US indicating that labor productivity in highly-exposed non-agricultural industries, such as utilities, mining, and manufacturing, declines by as much as 50% on a workday above 95<sup>◦</sup> F (35<sup>◦</sup> C), a ninefold impact relative to less exposed industries, although there seems to be adaptation when evaluating cross-county differences (Behrer and Park, 2017).<sup>xxxv</sup> In indoor controlled environments, productivity tends to be similarly negatively associated with temperature, often in a non-linear, U-shaped way. A meta-analysis of indoor office, laboratory, and classroom settings finds task performance to increase until reaching a ceiling at 21<sup>_◦_</sup> C- 22<sup>_◦_</sup> C, with further increasing temperature from 23<sup>_◦_</sup> C to 30<sup>_◦_</sup> C resulting in a 9% productivity loss (Seppanen et al., 2006). Student performance among college entrants in China declined by 2.9% of a standard deviation for a 1<sup>_◦_</sup> C increase in mean temperature of 23<sup>_◦_</sup> C during the exam period—with a more pronounced effect for highperforming students (Graff Zivin et al., 2020). These impacts are larger than those found among students of similar age in New York (-1.60%/+1<sup>_◦_</sup> C) (Park, 2022), possibly due to lower unobserved adaptation, like access to air conditioning.\\n\\nHealth is increasingly seen as an important mediating factor. When heat exceeds certain thresholds, negative physiological responses such as respiratory disease and cardiovascular strain, higher levels of fatigue and exhaustion, heat strokes, and cognitive impairment can affect body and brain functioning, which typically reduce work capacity while on-the-job (Zander et al., 2015). One adaptive response to cope with these effects is to reduce labor effort. DHS interviewers reduce their labor effort on hotter and more humid than usual days, completing 13.6% fewer interviews per hour (LoPalo, 2023). Rice harvesters in India self-report heat exhaustion, pain, signs of cardiovascular strain, during work on hot days, which 16\", \"datasets\": [{\"mention_name\": {\"text\": \"county-level payroll evidence\", \"confidence\": 0.5255524516105652, \"start\": 312, \"end\": 341, \"clean_text\": \"county-level payroll evidence\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 312, \"end\": 341, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 312, \"end\": 341, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 312, \"end\": 341}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 312, \"end\": 341, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.33728644251823425, \"start\": 312, \"end\": 341}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5255524516105652, \"start\": 312, \"end\": 341}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5802154541015625, \"start\": 312, \"end\": 341}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5802154541015625, \"start\": 312, \"end\": 341}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"caused a reduction in the number of rice bundles collected per worker – the equivalent of 5% lower hourly productivity per worker per 1<sup>_◦_</sup> C at WBGT > 26<sup>_◦_</sup> C (Sahu et al., 2013). Emerging evidence suggests that workers may acclimatize to heat exposure as the body may develop heat tolerance over time, which may mute productivity loss, but more research is needed to understand this relationship.35F<sup>xxxvi</sup> An alternative response is to increase the number of rest breaks. In Indonesia, workers in deforested settings that lack natural cooling services are 8% less productive and take 44% more breaks; they are found to have a 0.14<sup>_◦_</sup> C higher median core body temperature and experience 39% chance of moderate hyperthermia (Masuda et al., 2021). Taking more breaks was found to be the key mechanism through which productivity declined, as work effort measured by physical activity using an accelerometer showed no differences between the two groups.\\n\\nThese micro impact estimates provide important complementary insights to macro-level estimates. First, they suggest that the latter may underestimate the economic costs of climate change when not taking on the job productivity losses into account. Second, they present a channel through which output losses occur, formally testing a hypothesis made by early research.<sup>xxxvii</sup> The primary focus on short-term effects, and limited attention for adaptation, may mean that these impacts will be smaller in the long run. Nevertheless, magnitudes of productivity losses may be substantial.<sup>xxxviii</sup>\\n\\n# **2.5. Income, vulnerability among self-employed**\\n\\nChanging weather patterns increase vulnerability and variation in income of the self-employed, especially in agriculture, increasing the risk of poverty. The poorest populations, which typically include smallholder subsistence farmers, are likely to be hit hardest and most frequently by climate change events (Morton, 2007). Their livelihoods depend more on natural assets like land and livestock, tend to be concentrated 17\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"in climate-sensitive activities, such as farming, forestry and capture fisheries, and they are more likely to live in hotter and drier locations (Nordhaus, 2006).\\n\\nThe rural poor in developing countries typically have limited capacity to cope with and manage the risks arising from negative income shocks, as shown by a large body of research, including shocks stemming from climate variability. For instance, households’ main liquid assets hardly recovered to pre-famine levels a decade after the famine of the mid-1980s in the Horn of Africa, even though they had used multiple coping strategies. Cattle holdings were, on average, two-thirds in value terms of what they were before the famine (Dercon, 2002). Additionally, we have limited understanding of long-term adaptation in agriculture in developing countries. Existing work suggests that adaptation is often sub-optimal, both across space and time, as we discuss in Section 3.2, possibly due to poor incentives or high cost of adaptation, and more research is needed.<sup>xxxix</sup> Rigorous research on the distribution of impacts – for instance across income levels or occupations – remains mostly absent for developing countries.<sup>xl</sup> The most advanced insights in this area are those related to reallocation and migration, which is what we turn to next.\\n\\n# **2.6. Reallocation of labor**\\n\\nThe impacts of climate change on labor demand, supply and productivity may in turn lead to a reallocation of labor. Climate migration has received much attention in the literature.<sup>xli</sup> Change in the sector of work – which may coincide with migration – has been studied less. We discuss each in turn.\\n\\n## **2.6.1. Sectoral reallocation**\\n\\nRecent research, primarily conducted in India, demonstrates how agricultural productivity shocks caused by changing weather conditions can lead to a reallocation of labor from agriculture to manufacturing and services. Between 2001 and 2007, a rise of 1<sup>_◦_</sup> C in the daily average temperature corresponded to a decrease of 7.1 percentage point in agricultural employment, an increase of 2.0 and 3.4 percentage point in 18\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"manufacturing and services employment, respectively, and a 0.7 percentage point increase in district-level unemployment (Colmer, 2018). Rural households that have experienced depletion of their first borewell within the past ten years are four percentage points more likely to see their adult members engaging in non-agricultural off-farm work (Blakeslee et al., 2020).\\n\\nFlexible and integrated labor markets can facilitate reallocation and expansion in the receiving sector. In more flexible local labor markets in India, an increase in the daily average temperature is linked to a 10% growth in output and a 14.6% increase in the number of contract workers in formal manufacturing firms. Conversely, in the less flexible labor markets, there was a decrease in employment, marked by a 12.9% reduction in firm output and a 14.8% decrease in the proportion of contract workers (Colmer 2018). The majority of workers who are reassigned in the less flexible labor markets transition to smaller, informal manufacturing companies.\\n\\nOther research confirms that climate change may accelerate structural change. In Brazil, economic activity in localities with higher incidence of droughts during 2000-2010 shifted more rapidly towards manufacturing, where employment increased nearly 8% while agricultural employment reduced by 11% (Albert et al., 2021). Among Indian rural households whose first borewell failed, the capacity to diversify into non-agricultural employment hinged on the structure of the local economy (Blakeslee et al., 2020). Individuals were more likely to leave their agricultural jobs in villages with more and larger firms and a more industrialized rural economy.\\n\\nLocal demand also matters. Falling agricultural employment and productivity can lead to lower demand for labor in the non-tradeable sector through backward and forward linkages, while manufacturing employment may at the same time expand with labor becoming cheaper (Liu et al., 2023; Albert et al., 2021). Likewise, an increase in local demand can promote the shift of labor from agriculture to nonagricultural sectors in response to positive rainfall shocks, as higher incomes stimulate greater demand for local products and services. A 1-standard deviation above local average rainfall during the main 19\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"growing season in rural India increased the probability of households having a primary occupation in the non-agricultural sector by 1.1 percentage point (Emerick, 2018).\\n\\nThe influence of climate change on structural transformation is garnering growing interest and presents a promising avenue for future research.<sup>xlii</sup>\\n\\n# **2.6.2. Climate migration**\\n\\nVariations in climate can affect people's choices to migrate in search of employment opportunities.<sup>xliii</sup> Several aspects of this relationship warrant investigation.\\n\\nMultiple studies observe a strong correlation between migration and rising _temperatures_ . Higher temperature induces within-country migration from rural to urban areas in Mexico, as well as crossborder migration to the US (Jessoe et al., 2018; Nawrotzki et al., 2015) . An additional Harmful Degree Day – defined as a day with an increase in temperature from 32 _._ 5<sup>◦</sup> C to 33 _._ 5<sup>◦</sup> C – increases the probability of migrating to urban areas and to the US by 1.4% and 0.3%, respectively. In El Salvador, an extreme rise in temperature decreases agricultural productivity and total production. To compensate the loss in income and escape poverty, farmers cut back their demand for hired workers, which in turn leads to outmigration, including to the US (Ibanez et al., 2022). Heat stress is found to be strongly related to longterm migration in Pakistan, particularly for men who mostly move long distances, including abroad, while women often move within geographical areas and villages (Mueller et al., 2014).<sup>xliv</sup> In Indonesia, higher temperature raises the probability of permanent household migration in non- linear ways, by 0.8% when the mean temperature increases from 26<sup>_◦_</sup> C to 27<sup>_◦_</sup> C _,_ and by 1.4% when it increases from 27<sup>◦</sup> C to 28<sup>◦</sup> C (Bohra-Mishra et al., 2014).\\n\\nMigration is also related to decreases in _rainfall_ . A one-standard-deviation decline in rainfall, signifying a negative rainfall shock compared to the long-term average, leads to a 4.5% reduction in agricultural labor in China and a 5% increase in migration to urban areas. This effect is particularly pronounced among younger individuals (Minale, 2018). In Bangladesh, the likelihood of internal migration rises when a higher 20\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"percentage of households face crop failure due to rainfall (Gray and Mueller, 2012). Similarly, in Indonesia, agricultural households affected by unanticipated low rainfall are more likely to have a household member migrate to a nearby area (Kleemans, 2015; Kleemans and Magruder, 2018).\\n\\nStudies examining fluctuations in both temperature and precipitation indicate that the relationship between precipitation and migration is less robust and less pronounced compared to that with heat and rising temperatures. While rainfall and flooding can prompt temporary and short-distance movement in Pakistan and Indonesia, permanent migration tends to be predominantly linked to temperature increase (Mueller et al., 2014; Bohra-Mishra et al., 2014; Kleemans, 2015).\\n\\nResearch on migration impact of natural disasters tend to find a strong relationship, though evidence for developing countries remains scarce. In the US, regions affected by tornadoes in the early 1900s exhibit increased net emigration, whereas areas subject to flooding experience greater inward migration, likely due to the government’s reconstruction efforts in flood-prone regions, rendering them more appealing for economic activity (Boustan et al., 2012).\\n\\nIn a similar vein, the Dust Bowl of the 1930s in the US, characterized by severe dust storms and drought that ravaged the US Great Plains, was found to trigger substantial emigration from impacted regions, leading to decline in the local population (Hornbeck, 2012).<sup>xlv</sup> Migration prompted by natural disasters may mediate lower local economic growth. On average, annual economic growth fell by 0.45 percentage points in hurricane-exposed US coastal counties, of which 28% are attributed to the relocation of wealthy individuals (Strobl, 2011). Scarce evidence for developing countries indicates that international migration to OECD countries is positively related with natural disasters in the migrants’ home country. However, the intensity of this relationship fluctuates depending on the nature of the disaster and the geographical region in question (Drabo and Mbaye, 2015).\\n\\nClimate migration seems to accelerate urbanization up to some limit. Precipitation is positively related to urbanization in Sub-Saharan Africa (Barrios et al., 2006).<sup>xlvi</sup> But moisture, measured as a function of 21\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"precipitation and evapotranspiration, is negatively associated with the absorption of farm workers into the urban labor force across 29 African countries, especially in cities with a strong manufacturing presence (Henderson et al., 2014).\\n\\nCausal identification remains a challenge when analyzing climate migration. Impact estimates may reflect pull as well as push factors, and attribution to a single element is often impossible (Mueller et al., 2020).<sup>xlvii</sup> The implication for policy is that, in many instances, it may be more effective to promote overall labor market integration rather than focusing on specific groups or measures. However, there is a unique case when it comes to conflict, which is frequently overlooked in the analysis.xlviii In general, climate migration tends to be seen as a coping strategy to alleviate liquidity constraints stemming from climate-related shocks (Kleemans 2015; Bazzi 2017), but at the same time frequently necessitates access to capital (Cattaneo and Peri, 2016).\\n\\n# **3. Response to the impact of climate change on labor**\\n\\nEconomic actors respond to the labor impact of climate change both through adaptation and mitigation. This section examines on the observed responses of firms, farms, households and workers to the first order labor impacts of climate change discussed earlier, with subsequent focus on government policy responses.<sup>xlix</sup>\\n\\n## **3.1. Methods and identification**\\n\\nHow do economic agents respond to a manifested impact of climate change on labor? The nascent literature in this area follows one of three approaches when analyzing the response of firms and workers.\\n\\nA first group of studies investigates _the extent of adaptation_ , regressing observed adaptive behavior on temperature. A common example is a firm’s adoption of air conditioning to improve worker productivity.\\n\\n22\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"The analysis is similar to the one presented in equation (1), with the adaptation behavior as left-hand side variable.<sup>l</sup> A second family of papers focuses on how adoption moderates the impact of climate on labor. In this approach adaptation enters as a right-hand side variable in equation (1) where it is interacted with the climate variable. For instance, one can study whether the adoption of AC alters the impact of temperature on worker productivity. While this does not amount to causal testing of the underlying mechanism, the obtained estimates allow assessing whether the impact of climate – in this case temperature – on worker productivity varies with adoption.<sup>li</sup> The same approach can be used to evaluate the adaptation capacity of specific subgroups (e.g. gender and income in the case of workers, size and profitability for firms; firms; regions; or industries).<sup>lii</sup> A third approach compares the short-term impact estimates obtained from equation (1) with estimates obtained from a long-term-difference equation (which differences out short-term unobserved changes) to get an approximation for adaptation over time (see Dell et al., 2014; Hsiang, 2016). This approach can be used to evaluate if short-term impacts are offset in the long term.<sup>liii</sup> These approaches suffer from some of the limitations discussed earlier, in particular that actual temperature (or other climate change variable) to which workers are exposed remain unobserved, and that the estimates are for partial equilibrium only.\\n\\nHow governments respond is typically assessed using the classic tools from impact evaluation. Few studies are able to report rigorous estimates based on exogenous variation (e.g. a discrete cut off, or a roll out over time).<sup>liv</sup> Other thorough work in this area uses panel data fixed effects estimation of the form presented in equation (1). While the policy target of interest serves as left hand-side variable, only a limited number of studies evaluate the impact of policies interventions on labor outcomes. Most studies in this field concentrate on the impact of policies on output or emissions as the left-hand side variable.\\n\\n23\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"# **3.2. Firm, farm, household and worker response**\\n\\n_Firms_ may mitigate the negative impact of higher temperature by cooling the workplace. Firm adaptation effort depends on the relative cost of adopting climate controls such as air-conditioning, passive cooling systems, or specific heat saving technology, versus the gains in output stemming from the resulting improvements in labor supply and productivity.\\n\\nSome adaptation measures may therefore be less attractive for low-productivity jobs. In India, the probability of investing in climate control is substantially lower in cloth-weaving plants compared to diamond plants (Somanathan et al., 2021). Air conditioning for the average cloth-weaving firm costs nearly 23% of the total wage bill, while estimated productivity loss is about 2%–4% per 1<sup>_◦_</sup> C rise, making the investment unattractive for low-markup firms. In contrast, high value-added diamond plants do provide air conditioning, especially to polishing units, which carry out work central to diamond quality and value.<sup>lv</sup> For other technologies, benefits may exceed costs more rapidly. Energy-efficient LED lighting was reduced heat, thereby increasing worker productivity, and at the same time lowering energy costs (by almost a third), lowering the break-even period from 3.5 years to 8 months (Adhvaryu et al., 2020).<sup>lvi</sup> Our understanding of how _farmers_ respond to climate change impacts is growing. Rice farmers in China use a range of adaptation strategies including upscaling irrigation, reseeding, fixing and cleaning seedlings, changing crop varieties, as well as increasing the use of fertilizers and pesticides. While these measures tend to improve mean yield, only a third of farms employ these at times of flood and drought (Huang et al., 2015). In Ethiopia farmers report using increased irrigation, soil conservation to preserve moisture content, drought-tolerant crop varieties, adjusting of planting time, and planting trees (Deressa et al., 2009). Famers in Bangladesh adopted heat-tolerant rice crops, switching in the process from rain-fed to irrigation-based varieties (Moniruzzaman, 2015).<sup>lvii</sup> Dryer and hotter climate encouraged farmers in South 24\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"America to adopt more suitable crop varieties including squash, fruits, and vegetables (Seo and Mendelsohn, 2008).\\n\\nWhile some of these adaptation strategies overlap, others vary by country, setting, and income group. Whether they reflect an optimal response to local conditions remains an active field of research. In China farmers managed to mitigate 9% of potential yield losses by adapting planting dates and growing-season length in response to contemporaneous changes in temperature. No such or other response is observed in the long term, suggesting suboptimal behavior (Cui and Xie, 2022). In India, hotter districts were found to adapt for temperature increases in the range of 18<sup>_◦_</sup> C to 27<sup>_◦_</sup> C, by adopting new agricultural practices and switching to heat-tolerant crops, resulting in lower yield losses than colder districts that faced the same increase.<sup>lviii</sup> But for higher temperature rises (>30<sup>_◦_</sup> C), their losses were the same as in cold districts, suggesting non-optimal adoption (Taraz, 2018).\\n\\nThis is broadly consistent with findings in high-income countries. In the US temperature affects maize yields in hot (Southern) and cool (Northern) states almost identically and there is little variation in the temperature-crop relationship across time, despite technological advances in farm-level adaptation for warmer climate (Schlenker and Roberts, 2009). In France, equal short- and long-term losses in wheat yields (10% and 7%, respectively), related to a 2<sup>_◦_</sup> C warming also indicated limited adaptation over time (Merel and Gammans, 2021).\\n\\nRestricted information, incomplete insurance, lack of access to extension services, and low adoption of technology itself have been named as barriers to optimal adaptation (Huang et al., 2015; Karlan et al., 2014; Suri and Udry, 2022).\\n\\nBeyond the farm-level response, rural _households and workers_ are found to use multiple strategies to reduce exposure and mitigate impact of climate shocks. These include common individual and collective strategies to reduce the impact of income shocks, like diversification, precautionary savings, liquidation of assets, and risk-sharing (Dercon, 2002). Among Eastern and Southern African households, about a 25\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"third adopted alternative forms of employment and one in six reduced consumption to adapt to climate shocks. In Ethiopia and Tanzania over 50% of households used savings or borrowing as a primary response to climate shocks. In Kenya, two third sold livestock as part of changing farming practices (Rahut et al., 2021).<sup>lix</sup> A large literature on income shocks shows that, despite the variety of risk-coping strategies they adopt, rural households do not manage to completely insure against income loss.<sup>lx</sup> How households and workers can insure themselves against climate related shocks is receiving increased attention, and we discuss some new directions as part of social protection policies below.\\n\\n# **3.3. Government policy response**\\n\\nA comprehensive overview of policies designed and implemented in response to climate change is beyond the scope of this paper. Our main interest is to what extent these policies focus on, or take into account, labor outcomes. Our examination of the relevant literature reveals two issues. First, evaluations of general climate change interventions seldom include an assessment of their labor impact. Second, climate change policies that target labor remain understudied. We discuss each in turn.\\n\\n## **3.3.1. General climate change policies and their impact on labor**\\n\\nA wide range of policies aim to address challenges related to climate change. It is a very broad set; even for the highest impact is observed: from mitigation policies like fossil fuel subsidy reform, over carbon pricing, alternative power generation, change in land use, and altered transport policies, to adaptation policies like the use of increased crop variety, disaster preparedness and adaptive social protection. Table A1 in Appendix 1 provides examples of policies that have been evaluated. These appraisals pay virtually no attention to the impact on labor. On the rare occasion when they do, they tend to focus on employment, showing sizeable – typically short-term – impacts, and neglect potential effects on other labor outcomes.\\n\\n26\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [27]}}, {\"page\": 28, \"chunk\": 28, \"input_text\": \"As an illustration, consider the example of carbon pricing, which is seen as an attractive policy lever to bring down carbon emissions, reduce global warming, and accelerate the transition to a low carbon economy.lxi A simulation for US manufacturing industries finds that a 1% increase in the energy price associated with carbon tax (resulting in $15/ton carbon) reduces output in the short run by as much as 5%, particularly in energy-intensive sectors, like iron, steel and aluminum (Aldy and Pizer, 2015). Although the study extensively covers the impact on output, it neglects the impact on labor. Research in this strand typically concentrates on the policy’s impact on economic output and emissions. But there is much to be gained from examining the labor-related consequences. One study, exploiting exogenous variation in eligibility for carbon tax discount across a panel of UK manufacturing firms, finds no causal evidence of policy impact on plant-level employment, and this result holds across differences in firm size, energy intensity, or trade intensity (Martin et al., 2014). In contrast, an economy wide simulation study for the US between 1976 and 2007 finds a sizeable labor impact of higher energy prices in the short term, including a negative employment effect between -0.10% to -0.16% (Deschenes, 2012).<sup>lxii</sup> These shortterm impacts differ among sectors, with agriculture and transportation experiencing the largest employment losses, with estimated cross-elasticities of -0.43% and -0.29%, respectively.lxiii A study for British Columbia finds that annual employment rose by 0.74% over the subsequent 6 years, compared to the rest of Canada, when _unilaterally_ introducing a state level revenue-neutral carbon tax in 2008 (Yamazaki, 2017). This masks large heterogeneity across industries with employment falling in emissionintensive and trade-exposed industries but rising in services.<sup>lxiv</sup> Impact evaluation of a $40/ton CO2 carbon tax for a 30% emission coverage in Europe (EU+ countries) find positive, but generally not statistically significant, effects on employment, either immediately or in the subsequent 5 years of policy implementation (Metcalf and Stock, 2020).\\n\\nStudies evaluating employment effects for developing countries remain scarce. A CGE model for the transport sector in Malaysia suggests that carbon tax leads to a fall in employment across all subsectors – echoing the negative effect on sectoral output, demand, and investment – as factors of production shift 27\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [28]}}, {\"page\": 29, \"chunk\": 29, \"input_text\": \"to less energy-intensive, more labor-intensive sectors in the economy (Solaymani et al., 2015). A recent IMF simulation for manufacturing in Asia and the Pacific utilizes national I/O tables to estimate the potential short-term impact of a $25/ton carbon tax (Dabla-Norris et al., 2021). Sectors that rely on carbon-intensive inputs, typically the downstream, extractive and energy-producing industries, are most affected in terms of output and employment. The amount of projected job loss is primarily driven by the industry’s energy dependence and share of employment. This varies substantially across countries. In Malaysia the largest expected job losses are concentrated in electrical and electronics manufacturing while in Mongolia the vast majority projected job losses are in wood and paper manufacturing.\\n\\n# **3.3.2. Labor targeting climate change policies**\\n\\nOur review of the literature points to six potential policies to target the impact of climate change on labor. We refer to these as: green jobs, green skills, labor-oriented adaptation, flexible labor regulation, labor market integration, and social protection. As shown in Table 1, each of these relate to one or more of the above discussed potential labor impacts of climate change. In depth or formal evaluation of these policies has so far remained limited. We discuss each in turn.\\n\\n**Table 1. Labor targeting climate change policies**\\n\\n|Impact of climate change on labor|Labor oriented policy response<br>|\\n|---|---|\\n||H|\\n|Demand for labor|<br>Green jobs|\\n|Supply of labor, time use|Labor-oriented adaptation, regulation for flexible work|\\n|Productivity|Labor-oriented adaptation, regulation for flexible work|\\n|Self-employed income, vulnerability, poverty|Green jobs, social protection|\\n|Reallocation of labor|Labor market integration, social protection|\\n|Skills and human capital|Green skills, social protection, education and health policies|\\n\\nPolicies that promote “green jobs” aim to increase net labor demand while transitioning to a low carbon economy.<sup>lxv</sup> Although comparing the employment impact of green employment policies remains 28\", \"datasets\": [{\"mention_name\": {\"text\": \"national I/O tables\", \"confidence\": 0.3872346878051758, \"start\": 171, \"end\": 190, \"clean_text\": \"national I/O tables\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 171, \"end\": 190, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 171, \"end\": 190, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 171, \"end\": 190}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 171, \"end\": 190, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9895229935646057, \"start\": 171, \"end\": 190}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3872346878051758, \"start\": 171, \"end\": 190}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6855534911155701, \"start\": 171, \"end\": 190}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6855534911155701, \"start\": 171, \"end\": 190}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [29]}}, {\"page\": 30, \"chunk\": 30, \"input_text\": \"relatively rare, a small literature is emerging. Table 2 provides an overview of key studies in this area. Comparison of impact across studies is complicated by the use of different measures and methods. The emerging work on energy transition, predominantly focusing on high-income countries, provides an illustration. One study finds that an additional state-level green policy, like renewable portfolio standards and energy efficiency resource standards, is associated with 1% additional green jobs on average across 361 metropolitan areas in the US (Yi, 2013). These results corroborate the findings from simulationbased studies that dominate this literature. A multi-regional I/O analysis of the transformation of the EU energy sector over the period 1995–2009 estimates a net employment effect of +0.2.4%, an equivalent of 530,000 jobs by 2009, with one-third attributable to spillovers between EU countries, i.e. employment generated in one country due to changes in another (Markandya et al., 2016). Another EU study finds that substantial job creation can be realized in more flexible labor markets (Blazejczak et al., 2014). A simulation study for the US power sector suggests that implementing more aggressive energy efficiency measures and renewable energy standards can create over 4 million FTE green job-years by 2030, while clean nuclear energy and a rise in carbon capture and storage (CCS) to 25% and 10% of total energy generation, respectively, can create an additional 500,000 job-years (Wei et al., 2010). Since green energy tend to be more labor-intensive than high-carbon, these sectors generate more jobs for a given dollar invested. In the US, a $1 million invested in fossil fuels is expected to generate 2.65 FTE jobs versus 7.49–7.72 FTE jobs when invested in green energy, such as renewables or energy efficiency (GarrettPeltier, 2017).\\n\\n29\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [30]}}, {\"page\": 31, \"chunk\": 31, \"input_text\": \"**Table 2. Examples of expected impact of decarbonization policies on the demand for labor**\\n\\n|Policy|Expected impact<br>C|ountryor region<br>Method|Reference|\\n|---|---|---|---|\\n|State-level green policy|1% more green jobs|US<br>Historical data|Yi (2013)|\\n|Regional-level energy|0.24% increase in|EU<br>Simulation|Markandya et|\\n|shift|employment||al. (2016)|\\n|Energy efficiency<br>measures and renewable<br>energy targets|4 million+ green job-years by<br>2030|US<br>Simulation|Wei et al.<br>(2010)|\\n|Investment in<br>renewables and energy<br>efficiency|Around 5 more FTE jobs<br>compared to similar<br>investment in fossil fuels|US<br>Simulation|Garrett-<br>Peltier, (2017)|\\n\\nState-level green policy refers to market-based tools that provide support for renewable energy and energy efficiency. Energy shift refers to a change from carbon intensive sources toward gas and renewables.\\n\\nThe new jobs that come with the transition to low carbon may require new skills, often referred to as _green skills_ . There is, by our knowledge, very little rigorous evaluation in this area using historical data. One US study finds that changes in environmental regulation are positively related to demand for green skills, in particular for technical and engineering tasks (Vona et al., 2018).<sup>lxvi</sup> Cooling technology in the workplace is an important example of _labor-oriented adaptation_ that may improve worker productivity and labor supply, as well as worker well-being. As mentioned previously, firm adaptation may be less likely for low-productivity jobs (Somanathan et al., 2021). Government subsidies can be beneficial in situations where a change in behavior is considered desirable but faces obstacles, for instance when the cost of early adoption is high but anticipated to fall substantially over time, potentially due to increased demand (Stern, 2022).\\n\\nPolicy makers have been motivated to introduce regulation regarding _occupational health and heat management to protect workers_ from heat-related stress. Preventative measures, such as frequent rest breaks, and hydration and high-sodium intake, can protect workers from heat-related illness and work accidents. These regulations are unlikely to be followed in informal firms and settings with weak enforcement. In China’s manufacturing industries protective measures knew limited implementation by private firms during periods of elevated temperature (Zhang et al., 2018).\\n\\n30\", \"datasets\": [{\"mention_name\": {\"text\": \"Historical data\", \"confidence\": 0.6066300868988037, \"start\": 229, \"end\": 244, \"clean_text\": \"Historical data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 229, \"end\": 244, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 229, \"end\": 244, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 229, \"end\": 244}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 229, \"end\": 244, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.526787519454956, \"start\": 229, \"end\": 244}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6066300868988037, \"start\": 229, \"end\": 244}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5810202956199646, \"start\": 229, \"end\": 244}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5810202956199646, \"start\": 229, \"end\": 244}}, {\"mention_name\": {\"text\": \"historical data\", \"confidence\": 0.6960301399230957, \"start\": 1083, \"end\": 1098, \"clean_text\": \"historical data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1083, \"end\": 1098, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1083, \"end\": 1098, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1083, \"end\": 1098}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1083, \"end\": 1098, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.9152629375457764, \"start\": 584, \"end\": 599}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6960301399230957, \"start\": 584, \"end\": 599}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4601942300796509, \"start\": 584, \"end\": 599}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4601942300796509, \"start\": 584, \"end\": 599}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [31]}}, {\"page\": 32, \"chunk\": 32, \"input_text\": \"_Labor regulation for flexible working_ may provide an alternative policy. Optimized working hours allow firms and workers to flexibly adjust work schedules to hot temperatures (Connolly, 2018). In one example workers can choose to operate more on cooler days or cooler hours within a day. In another approach, workers decide their days off, relaxing the “consecutive vacation days” requirements in some sectors; this can also reduce absenteeism. Increased flexibility is particularly attractive for those with dual job holdings, for instance working in agriculture while having a second job in trade, as they tend to adjust working hours toward less exposed activities during extreme temperatures (Li and Pan, 2020; Huang et al., 2020; Branco and Feres, 2021).\\n\\n_Labor market integration_ increases labor mobility across the economy, which in turn facilitates the reallocation of labor. In India, reallocation in response to increased temperature takes place primarily _within_ districts (Colmer, 2018; Blakeslee et al., 2020).<sup>lxvii</sup> In other settings, integration with wider labor markets is found to stimulate climate-induced migration within and across regions, within the country, and internationally (see, for example, Bohra-Mishra et al. (2014), Mueller et al. (2014), Jessoe et al. (2018), Kleemans and Magruder (2018), Ibanez et al. (2022)).<sup>lxviii</sup> _Social protection_ can help reduce vulnerability and smoothen transition. One active area of research is shockresponsive social protection. Experimental evidence from Nicaragua suggests that conditional cash transfers combined with vocational training or a productive investment grant can help households exposed to weather variability smoothen their consumption and diversify their economic activities (Macours et al., 2022). Other research focuses on anticipatory crisis financing. Findings from a recent large-scale evaluation of immediate cash transfers during the 2020 monsoon floods in Bangladesh show that _timely_ and _quick_ release of cash support leads to more effective evacuation, better child food consumption outcomes, and less costly borrowing in the aftermath of the floods (Pople et al., 2021).\\n\\n31\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [32]}}, {\"page\": 33, \"chunk\": 33, \"input_text\": \"# **4. Ways forward**\\n\\nThis paper aims to stimulate research related to the impact of climate change on labor. A better understanding of this impact is needed for at least three reasons. First, much like the increased attention for other primary factors of production, such as land and finance, a better grasp of the role of labor will inform policies that aim to promote green growth and facilitate the transition to zero carbon.<sup>lxix</sup> Secondly, labor remains one of the key channels through which climate change impacts individuals’ livelihoods, affecting both their income from employment and their opportunities for work. Recognizing these effects can guide the development of policies aimed at poverty reduction and help steer clear of strategies that exacerbate poverty. Thirdly, the shift toward low carbon requires political, and thus citizen support. Ignoring the influence of climate change on labor could undermine such backing; acknowledging it has the potential to bolster it.\\n\\nWhile evidence is emerging across the five areas of labor impact that we identified, its availability remains constrained, particularly for developing countries. Four observations stand out from reviewing the evidence, for which summary is visualized in Table A.2 in Appendix 2.\\n\\nTo begin with, there exists at least some evidence pertaining each of the different labor themes. In certain instances, this evidence originates predominantly from research conducted in high income countries, notably the US. Even so, it frequently serves as an initial reference for subsequent studies, offering valuable insights into both research methods and the expected scale of the relationships under investigation, providing an empirical benchmark. Recent efforts have been increasingly directed towards the study of developing countries. Secondly, changes in temperature have garnered the greatest amount of scrutiny, with changes in precipitation ranking second in terms of attention. Other climate change events such as rising sea levels, tropical cyclones, and drought, have received considerably less concentrated effort. One contributing factor to this disparity may be the relatively rare occurrence of these events, while the scarcity of high-quality data also plays a part. Thirdly, certain geographical regions 32\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [33]}}, {\"page\": 34, \"chunk\": 34, \"input_text\": \"are characterized by a dearth of available evidence. This holds true for low-income countries in general, and specifically for Sub-Saharan Africa, as well as the Middle East.<sup>lxx</sup> Fourthly, even though we maintained our emphasis on the highest-quality evidence throughout our review, there are several limitations with the existing analysis. The omission of adaptation in most estimations imply that the majority of estimates reflect short term impact; they also typically focus on partial equilibrium effects, and primarily are about the influence of past, already occurred events. As a result, these estimates are prone to bias when it comes to the long term. Ongoing efforts to address these shortcomings involve conducting economy wide analysis, and more precise forecasting future of climate change events, both of which remain challenging. Of particular interest is the research examining how climate change could expedite structural transformation. Essential is the work that incorporates adaptation, typically using now established methods like explicitly modelling its adoption, examining how it moderates the estimated impacts of climate events, or comparing short term impacts with those of the long term. Emerging evidence suggests that adaptation is frequently suboptimal, and we currently have limited understanding why this is the case.\\n\\nIn the realm of government policy much remains to be done. The assessment of general policy interventions often overlooks their impact on labor. Formal evaluation of any of the six labor-targeting policies that we identified – green jobs, green skills, labor-oriented adaptation, flexible work regulation, labor market integration, and social protection – remains scarce overall, and in particular for developing countries.\\n\\nBased on these observations, we see three promising avenues for future research. A first route is to deepen research in areas where we have an initial understanding. Additional effort is required to incorporate adaptation, utilizing either existing methodologies or developing new ones. Conducting more causal analysis of historical data, whenever feasible, can further enrich our understanding. Sector specific examination serves as a solid starting point, and can be further enriched by studying impact beyond the 33\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [34]}}, {\"page\": 35, \"chunk\": 35, \"input_text\": \"primary activity of interest, including an investigation of both backward and forward linkages. This will serve as both a supplement to and a source of information for approaches examining economy wide effects, which remain sensitive to their underlying assumptions. Further progress can also be made in future-oriented analysis, with two avenues of work opening up. First, research comparing and integrating historical and simulation analyses would significantly enhance our knowledge base. Forward-looking projections of the likelihood of climate events have become more accurate, thanks to insights from science. Simultaneously, there has been a substantial increase in research devoted to estimating impact through historical data. Work that bridges insights from these two traditions, while considering anticipated adaptation, would be highly valuable. Second, a newer strand of work takes into account social adaptive behavior at scale, further extending the type of socioeconomic variables that are being considered, which is a promising way forward. Society wide behavior is hard to predict, and much can be gained from theoretical work, which remains limited in this area. Recent efforts underline how theory can push our understanding, for instance by recognizing endogenous preferences, the role of technology, and changing consumption patterns society wide ( <mark>Mattauch et al.,</mark> 2018; Besley and Persson, 2020; van der Ploeg and Venables, 2022).\\n\\nSecondly, in the few themes where we have advanced evidence, replication and comparison are needed. This will help us better understand the generalizability of the results, or, failing that, the possible aspects of context that matter. Similar lessons can be drawn for evaluation of government policy. Greater utilization of evaluations examining the impact of general climate change policies on labor, along with the replication of labor-targeting climate change policies, would enable a systematic comparison of their economic and social impacts. This would help determine the relative effectiveness of these policies and enable more informed policy making.\\n\\n34\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [35]}}, {\"page\": 36, \"chunk\": 36, \"input_text\": \"Third, themes for which evidence is lacking warrant a thorough assessment to identify the most fundamental and pressing questions that need addressing. This process will, in turn, highlight the need for improved data.\\n\\nWe conclude this review with a more optimistic outlook than when we began. While there is still much work ahead, clear pathways have emerged. Combined, these research and policy making efforts will help determine the optimal paths for achieving an equitable transition to a low-carbon economy.\\n\\n35\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099842001242436996/pdf/IDU13fb2bcd51c7f1140371b61811281e0957f86.pdf\", \"pages\": [36]}}]}"}}},{"rowIdx":88,"cells":{"doc_id":{"kind":"string","value":"34247621"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34247621\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"authors\": {\"0\": {\"author\": \"Gascoigne,Jon\"}, \"1\": {\"author\": \"Baquie,Sandra\"}, \"2\": {\"author\": \"Vinha,Katja Pauliina\"}, \"3\": {\"author\": \"Skoufias,Emmanuel\"}, \"4\": {\"author\": \"Calcutt,Evie Isabel Neall\"}, \"5\": {\"author\": \"Kshirsagar,Varun Sridhar\"}, \"6\": {\"author\": \"Meenan,Conor\"}, \"7\": {\"author\": \"Hill,Ruth\"}}, \"count\": \"Africa\", \"count_key\": \"517192\", \"docna\": {\"0\": {\"docna\": \"The Welfare Cost of Drought in Sub-Saharan Africa\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"EFI-Poverty and Equity-GE (EPVGE)\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34247621\"}, \"repnb\": \"WPS10683\", \"docdt\": \"2024-01-29T00:00:00Z\", \"datestored\": \"2024-01-29T01:25:32Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"This paper quantifies the impact of\\n drought on household consumption for five main\\n agroecological zones in Africa, developing vulnerability (or\\n damage) functions of the relationship between rainfall\\n deficits and poverty. Damage functions are a key element in\\n models that quantify the risk of extreme weather and the\\n impacts of climate change. Although these functions are\\n commonly estimated for storm or flood damages to buildings,\\n they are less often available for income losses from\\n droughts. The paper takes a regional approach to the\\n analysis, developing standardized hazard definitions and\\n methods for matching hazard and household data, allowing\\n survey data from close to 100,000 households to be used in\\n the analysis. The damage functions are used to quantify the\\n impact of historical weather conditions on poverty for eight\\n countries, highlighting the risk to poverty outcomes that\\n weather variability causes. National poverty rates are\\n 1–12 percent higher, depending on the country,\\n under the worst weather conditions relative to the best\\n conditions observed in the past 13 years. This amounts to an\\n increase in the total poverty gap that ranges from US$4\\n million to US$2.4 billion (2011 purchasing power parity).\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10683; PLANET\", \"display_title\": \"The Welfare Cost of Drought in\\n Sub-Saharan Africa\", \"disclosure_date\": \"2024-01-29T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099325301292478621/text/IDU1ae0eac0e145d214c6218002156b672eb8155.txt\", \"listing_relative_url\": \"/research/2024/01/34247621/welfare-cost-drought-sub-saharan-africa\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099325301292478621/IDU1ae0eac0e145d214c6218002156b672eb8155\", \"new_url\": \"2024/01/34247621/IDU1ae0eac0e145d214c6218002156b672eb8155\", \"ext_pub_date\": \"2024-01-29T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Poverty m1326687 695,Drought m1503313 3162\", \"docm_id\": \"IDU-ae0eac0e-45d2-4c62-8002-56b672eb8155\", \"chronical_docm_id\": \"IDU-ae0eac0e-45d2-4c62-8002-56b672eb8155\", \"publishtoextweb_dt\": \"2024-01-29T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10683\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"guid\": \"099325301292478621\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"49\", \"author_keys\": {\"0\": {\"author_key\": \"NA\"}, \"1\": {\"author_key\": \"1767911\"}, \"2\": {\"author_key\": \"435289\"}, \"3\": {\"author_key\": \"177660\"}, \"4\": {\"author_key\": \"1717790\"}, \"5\": {\"author_key\": \"810339\"}, \"6\": {\"author_key\": \"NA\"}, \"7\": {\"author_key\": \"663673\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099325301292478621/IDU1ae0eac0e145d214c6218002156b672eb8155\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10683\\n\\n# The Welfare Cost of Drought in Sub-Saharan Africa\\n\\n_Jon Gascoigne Sandra Baquie Katja Vinha Emmanuel Skoufias Evie Calcutt Varun Kshirsagar Conor Meenan Ruth Hill_ Poverty and Equity Global Practice January 2024 A verified reproducibility package for this paper is available at http://reproducibility.worldbank.org, click **here** for direct access.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10683\\n\\n# **Abstract**\\n\\nThis paper quantifies the impact of drought on household consumption for five main agroecological zones in Africa, developing vulnerability (or damage) functions of the relationship between rainfall deficits and poverty. Damage functions are a key element in models that quantify the risk of extreme weather and the impacts of climate change. Although these functions are commonly estimated for storm or flood damages to buildings, they are less often available for income losses from droughts. The paper takes a regional approach to the analysis, developing standardized hazard definitions and methods for matching hazard and household data, allowing survey data from close to 100,000 households to be used in the analysis. The damage functions are used to quantify the impact of historical weather conditions on poverty for eight countries, highlighting the risk to poverty outcomes that weather variability causes. National poverty rates are 1–12 percent higher, depending on the country, under the worst weather conditions relative to the best conditions observed in the past 13 years. This amounts to an increase in the total poverty gap that ranges from US$4 million to US$2.4 billion (2011 purchasing power parity).\\n\\nThis paper is a product of the Poverty and Equity Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at rhill@worldbank.org. A verified reproducibility package for this paper is available at http://reproducibility.worldbank. org, click **here** for direct access.\\n\\n<!-- Start of picture text --> TRANSPARENT<br>ANALYSIS<br>G<br>R E<br>N<br>S<br>I P<br>Y E<br>K A<br>C A<br>R P<br>LI R<br>O E<br>O C<br>W R<br>H<br>S<br>P<br><!-- End of picture text --> _The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [{\"mention_name\": {\"text\": \"survey data\", \"confidence\": 0.6908695697784424, \"start\": 705, \"end\": 716, \"clean_text\": \"survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 705, \"end\": 716, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 705, \"end\": 716, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 705, \"end\": 716}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 705, \"end\": 716, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9460253119468689, \"start\": 705, \"end\": 716}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6908695697784424, \"start\": 705, \"end\": 716}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.638032078742981, \"start\": 705, \"end\": 716}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.638032078742981, \"start\": 705, \"end\": 716}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# The Welfare Cost of Drought in Sub-Saharan Africa\\n\\nJon Gascoigne (Centre for Disaster Protection) r Sandra Baquie (World Bank) r Katja Vinha (Consultant) r Emmanuel Skoufias (National University of Singapore and World Bank) r Evie Calcutt (World Bank) r Varun Kshirsagar Consultant) r Conor Meenan (Centre for Disaster Protection) r Ruth Hill (World Bank and Centre for Disaster Protection)<sup>1</sup> Originally published in the Policy Research Working Paper Series on January 2024. This version is updated on July 2025. To obtain the originally published version, please email prwp@worldbank.org.\\n\\nJEL Codes: Q54, I32 Project Code: P177432 Keywords/topics: Shocks and vulnerability to poverty, drought\\n\\n> 1 r indicates randomized author order using the American Economic Association author randomization tool, https://www.aeaweb.org/journals/policies/random-authororder/search?RandomAuthorsSearch%5Bsearch%5D=XxSZtJmQOh5k.\\n\\nThis paper started as a result of a conversation with Jon Gascoigne in early 2020. Jon understood better than any of us the need for a multidisciplinary team to understand how to model the welfare impacts of drought. Collaborating across disciplines can be a challenge, and negotiating those discussions requires good grace, humility, a strong focus on relationship, and a genuine interest in understanding another’s view and expertise. Jon exemplified these characteristics in all of his work, and true to form, brought them to this project too. We think it is very fitting that the randomization tool placed him as first author on this paper early in its drafting. We are saddened he didn’t get to see the final product and to know just how grateful we all were to him and the example he set us.\\n\\nWe thank Stefanie Brunelin and Odyssia Ng for being part of the early conversation on this paper and we thank Bhavin Thakrar and Paul Wilson for working on parts of the simulations data exercise.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# 1. Introduction\\n\\nClimate disasters reverse gains and limit progress in efforts to reduce poverty. Global extreme poverty is increasingly concentrated in Sub-Saharan Africa (SSA) where drought is a major risk, a risk that is increasing in many parts of the continent because of climate change. Investments to reduce the impact of drought on welfare are thus essential to poverty reduction efforts in the region. Quantifying the welfare needs that arise when climate shocks occur is an important part of targeting investments to reduce their impact (Clarke and Dercon 2016). This requires vulnerability or damage functions that relate losses in welfare to climate conditions (Auffhammer 2018). While these are widely available for the impact of floods and cyclones on buildings, this is not the case when it comes to quantifying the impact of drought on household consumption in Sub-Saharan Africa.\\n\\nThis paper pilots a regional approach to estimating vulnerability functions for drought and consumption using methods widely employed in the microeconomic literature (Dell, Jones, and Olken 2014). The analysis uses survey data from close to 100,000 households across nine countries and five livelihood zones, allowing considerable variation in climatic conditions to be utilized in estimation. The result is stable estimates that are not driven by any one climate event. We find that consumption is reduced by 10-20 percent in the worst weather observations in our surveys (corresponding to about a 1 in 10-year drought event). We use the estimated vulnerability functions and the full hazard distribution in each household location to calculate the risk drought poses to welfare outcomes. The results indicate that risk is large. Poverty is 1–12 percent higher under the worst weather conditions relative to the best conditions observed in the past 13 years. This amounts to an increase in the total national poverty gap that ranges from US$4 million to US$2.4 billion (2011 PPP) across countries. To our knowledge, these are the first crosscountry estimates of the poverty impact of drought in Africa.\\n\\nThere are several ways in which damage functions for rainfall deficits have been estimated in the climate economics literature. Auffhammer (2018) provides a review and examples of the approaches applied to US farm incomes. The one used for estimating the impact of extreme weather events uses observed data to identify the causal impact of weather changes on outcomes of interest (Dell, Jones, and Olken 2014). Data on hazards are matched to household observations in survey data using information on the geographic location of a household, and a fixed effects regression is run to identify the impact of weather in determining economic outcomes.\\n\\nAlthough not undertaken with the purpose of estimating vulnerability functions, several studies have used this method to estimate the impact of drought on welfare. For example, in Africa, Hill and Porter (2017) look at the impact of a water requirement satisfaction index (WRSI) on consumption and poverty, and Hill and Mejia-Mantilla (2017) look at the impact of WRSI on household income and consumption in Uganda. In Malawi, Baquie and Fuje (2020) look at the impact of rainfall on consumption and poverty. Wineman, Mason, Ochieng and Kirimi (2019) examine the relationship between rainfall and income, calorie consumption and poverty in Kenya. Baez, Kshirsagar and Skoufias (2019) look at the impact of rainfall defined dry spells on nutrition outcomes. Outside Africa, Kochhar and Knippenberg (2023) look at the impact of NDVI, rainfall and temperature on consumption and poverty in Afghanistan. Studies that use estimated impacts of drought to predict the welfare cost of other weather events include Hill and Porter (2017), Porter and White (2016), Kochhar and Knippenberg (2023).<sup>2</sup> This study uses the same methods as these papers but takes a regional approach (as in Baez, Kshirsagar and Skoufias 2019) with the objective of developing robust estimates of the relationship between drought conditions and consumption that is not specific to one context or event. This relationship can then be used to predict the impact of drought for other events or locations. The estimates presented in this paper are within the range of results found in country-specific papers. For example, a moderate (about a 1 in 10 year) drought is predicted to reduce consumption by 15 percent and 9 percent in Uganda and Ethiopia, akin to the 10 percent loss we estimate for crop cultivating households in maize and highland zones in this paper.\\n\\n> 2 Anttila-Hughes and Sharma (2015) take the same approach for storms.\\n\\n2\", \"datasets\": [{\"mention_name\": {\"text\": \"survey data\", \"confidence\": 0.7883277535438538, \"start\": 1115, \"end\": 1126, \"clean_text\": \"survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1115, \"end\": 1126, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1115, \"end\": 1126, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1115, \"end\": 1126}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1115, \"end\": 1126, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9891263246536255, \"start\": 1115, \"end\": 1126}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7883277535438538, \"start\": 1115, \"end\": 1126}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9020770192146301, \"start\": 1115, \"end\": 1126}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9020770192146301, \"start\": 1115, \"end\": 1126}}, {\"mention_name\": {\"text\": \"observed data\", \"confidence\": 0.5088502764701843, \"start\": 2398, \"end\": 2411, \"clean_text\": \"observed data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2398, \"end\": 2411, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2398, \"end\": 2411, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2398, \"end\": 2411}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2398, \"end\": 2411, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9546656608581543, \"start\": 654, \"end\": 667}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5088502764701843, \"start\": 654, \"end\": 667}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9586206078529358, \"start\": 654, \"end\": 667}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9586206078529358, \"start\": 654, \"end\": 667}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"Taking a regional approach requires the use of standardized hazard definitions and methods for matching hazard and household data. This is done by considering the main hazard measures and techniques used in comparable country studies, and systematically assessing the choice of hazard, the hazard-household matching approach, and the functional form used in analysis. Specifically, we use gridded historical data on precipitation, vegetation, evapotranspiration, and soil moisture, using measures such as the Normalized Difference Vegetation Index (NDVI), the Standardized Precipitation Evapotranspiration Index (SPEI), and WRSI. These measures reflect the hazard to livelihoods from lack of rain in different ways. Some are direct measures of climate conditions (e.g., precipitation, soil moisture, and SPEI), others aggregate climate data through crop models (WRSI), and others measure vegetation outcomes (NDVI). The gridded hazard measures are then merged to harmonized household surveys from nine countries (Ethiopia, Lesotho, Malawi, Mauritania, Mozambique, Niger, Nigeria, Zambia, and Zimbabwe) with location coordinates for either the household or the community. The surveys used are the ones that are also used for official poverty estimates. A consistent approach to temporal matching was used to take into account the timing of the survey and agricultural seasons. Different approaches to spatial merging were tested.\\n\\nThere are important methodological insights from the analysis that emerge. First, we find that soil moisture and greenness measures provide more consistently reliable results than evapotranspiration or rainfall measures. This is an important finding given most of the extant literature is based on evapotranspiration or rainfall metrics. Second, we find that a linear specification performs adequately for the types of moderate shocks that are well-represented in the survey data used. Third, we find that a hazard measure based on 20 km or 50 km radius around each household performs better than those that use smaller radii, perhaps because a larger radius allows local market effects to also be considered.\\n\\nThe relationship between hazard measures and welfare outcomes is estimated separately for different livelihood zones across SSA. Separate estimation provides needed flexibility, given that drought conditions manifest themselves differently across livelihood zones. Important differences in impacts between agroecological zones and between land and livestock owners are observed. However, the downside of a regional approach is a limited ability to explore the heterogeneity of impacts across households in detail without significant investments in data harmonization, and we leave additional work on the heterogeneity of impacts as something to explore further in future analysis.\\n\\nThere are limitations to the analysis undertaken that should be borne in mind when interpreting results. First, estimates capture the direct impact of local drought conditions and not impacts beyond the local area, such as those that might arise from disruption of markets or price impacts. The degree to which this is a concern depends on the relative importance of direct and indirect impacts. The evidence is mixed on this. Hallegatte et al. (2016) suggest food prices are an important channel of the welfare impacts of climate change. However, Artuc et al. (2023) model trade between locations and find the main impact is the direct production effect, not the impact on prices. To the extent this is the case the local estimates presented in this paper may be quite close to aggregate impacts.\\n\\nSecondly, the empirical estimates are subject to “survivor bias” in that consumption is only measured for households still in existence after the shock has occurred. If households move or disintegrate (perhaps due to the death of one or more household members) because of drought conditions, the estimated impacts could be over or underestimated. This is much more likely to be the case for severe drought conditions, whereas those that are captured in our surveys tend to be less extreme events that are unlikely to cause movement or disintegration of households.\\n\\nThis brings us to the third potential source of bias in the results. Although the analysis uses survey data across many weather conditions in the analysis, there are very few observations in the survey data that experience drought conditions below two standard deviations from the mean. This limits the degree to which non-linearities can be estimated. It also means there is limited empirical support for predicting the welfare costs of extreme events.\\n\\nA final source of bias comes from measurement error in the hazard data. Although care was taken to use the best available data and match hazard and household data carefully, any measurement error in the hazard variable will attenuate the estimated coefficients, thereby underestimating the welfare impacts. Additional uncertainty is 3\", \"datasets\": [{\"mention_name\": {\"text\": \"gridded historical data\", \"confidence\": 0.42629241943359375, \"start\": 389, \"end\": 412, \"clean_text\": \"gridded historical data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 389, \"end\": 412, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 389, \"end\": 412, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 389, \"end\": 412}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 389, \"end\": 412, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.5051462054252625, \"start\": 389, \"end\": 412}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.42629241943359375, \"start\": 389, \"end\": 412}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9090306162834167, \"start\": 389, \"end\": 412}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9090306162834167, \"start\": 389, \"end\": 412}}, {\"mention_name\": {\"text\": \"Normalized Difference Vegetation Index\", \"confidence\": 0.7295650839805603, \"start\": 509, \"end\": 547, \"clean_text\": \"Normalized Difference Vegetation Index\"}, \"acronym\": {\"text\": \"NDVI\", \"confidence\": 0.34257081151008606, \"start\": 549, \"end\": 553, \"clean_text\": \"NDVI\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 509, \"end\": 547, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 509, \"end\": 547}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 509, \"end\": 547, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.5051462054252625, \"start\": 509, \"end\": 547}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7295650839805603, \"start\": 509, \"end\": 547}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9090306162834167, \"start\": 509, \"end\": 547}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9090306162834167, \"start\": 509, \"end\": 547}}, {\"mention_name\": {\"text\": \"Standardized Precipitation Evapotranspiration Index\", \"confidence\": 0.6637430787086487, \"start\": 560, \"end\": 611, \"clean_text\": \"Standardized Precipitation Evapotranspiration Index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 560, \"end\": 611, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 560, \"end\": 611, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 560, \"end\": 611}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 560, \"end\": 611, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.5051462054252625, \"start\": 560, \"end\": 611}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6637430787086487, \"start\": 560, \"end\": 611}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9090306162834167, \"start\": 560, \"end\": 611}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9090306162834167, \"start\": 560, \"end\": 611}}, {\"mention_name\": {\"text\": \"WRSI\", \"confidence\": 0.42541730403900146, \"start\": 624, \"end\": 628, \"clean_text\": \"WRSI\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 624, \"end\": 628, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 624, \"end\": 628, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 624, \"end\": 628}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 624, \"end\": 628, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.5051462054252625, \"start\": 624, \"end\": 628}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.42541730403900146, \"start\": 624, \"end\": 628}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9090306162834167, \"start\": 624, \"end\": 628}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9090306162834167, \"start\": 624, \"end\": 628}}, {\"mention_name\": {\"text\": \"harmonized household surveys from nine countries\", \"confidence\": 0.6044202446937561, \"start\": 963, \"end\": 1011, \"clean_text\": \"harmonized household surveys from nine countries\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 963, \"end\": 1011, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 963, \"end\": 1011, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 963, \"end\": 1011}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 963, \"end\": 1011, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8998010158538818, \"start\": 90, \"end\": 138}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6044202446937561, \"start\": 90, \"end\": 138}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4792098104953766, \"start\": 90, \"end\": 138}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4792098104953766, \"start\": 90, \"end\": 138}}, {\"mention_name\": {\"text\": \"survey data\", \"confidence\": 0.9434069991111755, \"start\": 1898, \"end\": 1909, \"clean_text\": \"survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1898, \"end\": 1909, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1898, \"end\": 1909, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1898, \"end\": 1909}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1898, \"end\": 1909, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.99122554063797, \"start\": 122, \"end\": 133}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8450183868408203, \"start\": 122, \"end\": 133}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7853732109069824, \"start\": 122, \"end\": 133}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7853732109069824, \"start\": 122, \"end\": 133}}, {\"mention_name\": {\"text\": \"hazard data\", \"confidence\": 0.7067624926567078, \"start\": 4702, \"end\": 4713, \"clean_text\": \"hazard data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4702, \"end\": 4713, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4702, \"end\": 4713, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4702, \"end\": 4713}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4702, \"end\": 4713, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.48796629905700684, \"start\": 312, \"end\": 323}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7067624926567078, \"start\": 312, \"end\": 323}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.668321430683136, \"start\": 312, \"end\": 323}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.668321430683136, \"start\": 312, \"end\": 323}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"present given the distribution of hazards is changing over time with climate change and historical data forms the basis of all simulations presented in this paper.\\n\\nThe models built in this paper are an important first step toward constructing vulnerability or damage functions as used in the catastrophe risk modeling literature, and can inform disaster response planning and financing. However, these limitations mean that care is needed in interpreting the results of the analysis. We envisage that the set of tools developed herein will provide the basis for refining the analysis as more data become available, and expanding coverage to other regions. Accompanying this paper is code designed to extract a large set of potential shock measures and determine which of these are consistent with changes in the welfare measures (outof-sample). The modular code used is intended to be transparent, to aid in replicability, and to be available for others to improve upon. The goal is to encourage both learning and common standards, which in turn contribute to the implementation of drought-responsive policies that are objective and transparent.\\n\\nThe paper is structured as follows. Section 2 presents the empirical framework for assessing the relationship between welfare and hazards in rural areas where rainfall-dependent agricultural production is dominant, and where rainfall deficits (droughts) thus may have very serious welfare consequences. It sets out the approach used to estimate the vulnerability function and an exceedance probability curve that describes the welfare losses likely to occur because of rainfall deficits of increasing severity. Section 3 discusses in detail the historical hazard database and the household survey data used, the merging of the hazard data to the household survey data, and the specifications used for the hazard measures. Section 4 presents estimates of vulnerability functions. Section 5 shows how these vulnerability functions can be used to project welfare losses and presents some exceedance probability curves. Section 6 offers some conclusions.\\n\\n# 2. An Empirical Framework for an Ex Ante Assessment of the Welfare Costs of Drought\\n\\nA catastrophe risk modeling framework uses three components to generate an ex ante assessment of the costs of a catastrophe: hazard, exposure, and vulnerability function. Hazards capture data on the possible weather outcomes that may occur in a given place, such as possible tropical cyclones. Exposure captures what could be affected by the hazard in that location, such as the number and type of buildings. A vulnerability function provides an estimate of the cost of damage that will occur as a result of a given hazard and a given exposure. In the example of tropical cyclones and buildings, it would indicate the cost of damage for a given windspeed affecting a given type of building. Combining the three components allows development of a probabilistic impact curve, which indicates how the cost of hazard varies with the probability of the hazard for a specific place, given the expected hazards and the exposed assets.\\n\\nThis section of the paper shows how this framework can be applied to generate a probabilistic impact curve for drought in SSA. For drought in SSA, the hazard is rainfall deficits during the growing season. As noted in the introduction, there are different ways of measuring this hazard, and the next section discusses the relevant data in more detail, including the time period for which historical data are available and use of the data to generate an estimate of the probability distribution of hazards in the near future.\\n\\nExposure in the current case is not about physical infrastructure and its location and characteristics, but about households that may be impacted differently by rainfall deficits depending on their location and characteristics. The data on households are also described in more detail in the next section.\\n\\nThe vulnerability function in this case is the income or welfare losses among households engaged in different livelihoods, which result from lower yields, lower demand for labor, and increased prices that may result from that. The rest of this section focuses on the estimation of the vulnerability function and how it is used to derive a probabilistic impact curve. The estimation of the vulnerability function is widely present in the microeconometric literature, even though it is not referred to as such or used for generating estimates of probabilistic impact.\\n\\n4\", \"datasets\": [{\"mention_name\": {\"text\": \"historical data\", \"confidence\": 0.7671879529953003, \"start\": 88, \"end\": 103, \"clean_text\": \"historical data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 88, \"end\": 103, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 88, \"end\": 103, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 88, \"end\": 103}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 88, \"end\": 103, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.8603382706642151, \"start\": 88, \"end\": 103}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7671879529953003, \"start\": 88, \"end\": 103}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8918560743331909, \"start\": 88, \"end\": 103}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8918560743331909, \"start\": 88, \"end\": 103}}, {\"mention_name\": {\"text\": \"historical hazard database\", \"confidence\": 0.5732136964797974, \"start\": 1693, \"end\": 1719, \"clean_text\": \"historical hazard database\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1693, \"end\": 1719, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1693, \"end\": 1719, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1693, \"end\": 1719}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1693, \"end\": 1719, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8633596897125244, \"start\": 796, \"end\": 822}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5732136964797974, \"start\": 796, \"end\": 822}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7637546062469482, \"start\": 796, \"end\": 822}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7637546062469482, \"start\": 796, \"end\": 822}}, {\"mention_name\": {\"text\": \"household survey data\", \"confidence\": 0.8641612529754639, \"start\": 1794, \"end\": 1815, \"clean_text\": \"household survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1794, \"end\": 1815, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1794, \"end\": 1815, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1794, \"end\": 1815}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1794, \"end\": 1815, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8633596897125244, \"start\": 897, \"end\": 918}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8641612529754639, \"start\": 897, \"end\": 918}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7637546062469482, \"start\": 897, \"end\": 918}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7637546062469482, \"start\": 897, \"end\": 918}}, {\"mention_name\": {\"text\": \"data on households\", \"confidence\": 0.6415009498596191, \"start\": 3874, \"end\": 3892, \"clean_text\": \"data on households\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3874, \"end\": 3892, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3874, \"end\": 3892, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3874, \"end\": 3892}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3874, \"end\": 3892, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.7400457859039307, \"start\": 350, \"end\": 368}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6415009498596191, \"start\": 350, \"end\": 368}, \"usage_context\": {\"text\": \"background\", \"confidence\": 0.3700110614299774, \"start\": 350, \"end\": 368}, \"is_used\": {\"text\": \"False\", \"confidence\": 0.3700110614299774, \"start\": 350, \"end\": 368}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"2.1. Estimating a vulnerability function Household welfare at any given time may be summarized in general by the function below: hk k hk , (1) ff where hk denotes the measure of welfare (in the current analysis, log of per capita expenditures hkk hkk P hk of WW = ff(RR , XX , EE ) + εε household h in locality or cluster ,), k is the measure of the hazard at time at the level of the geographic cluster WW , denotes household observable characteristics, denotes an observable variable summarizing the EE environment of households, and κκhk RR is an error term summarizing the influence of all unobservable factors on tt hkk hkk welfare. κκ<sup>3</sup> For the purposes of this study, welfare is measured by the monetary value of total consumption XX EE expenditures or food and non-food consumption. εε<sup>4</sup> The main objective of a vulnerability function is to summarize the relationship between the level of and the level of . If the function were known a priori it would be simple to plug in values for and to determine the value of the welfare outcome . In the absence of information on the function , methods need to be WW employed to estimate it. RR ff RR, XX, EE WW ff This necessity gives rise to a fundamental trade-off between the prediction accuracy and the interpretability of the model used (James et al. 2013). For example, machine learning methods, such as random forests, bagging, boosting, and support vector machines, treat as a black box. None of these methods is especially concerned about the exact form and interpretability of the function , provided that it yields accurate predictions for the welfare outcome. ff ff On the other hand, approximating the function with the average value of the welfare outcome hk conditional on specific values of the explanatory variables , and — i.e., hk k = or the conditional expected function (CEF)—is typically easier to interpret and is useful for deeper ff WW kk 1hkk hkk hkk insights into the causal relationship between hazard and welfare (Angrist and Pischke 2009). These advantages, RR , XX EE ff(rr, xx, ee) = EE(WW |RR = rr, XX hkk xx, EE however, may come at the expense of predictive accuracy (in-sample or out-of-sample). = ee),<sup>5</sup>\\n\\nWith these considerations in mind, we adopt the CEF approach for estimating vulnerability functions. To keep the analysis simple and tractable, the approach begins with an ad hoc specification of the controls used in the regression model, using a linear approximation to the CEF that is extended to a nonlinear specification for the hazard and allowing for the potential heterogeneity of the hazard’s impacts on welfare based on some easily and commonly observable household characteristics and environmental features. For any given welfare measure h , a log linear approximation to equation (1) above yields hk 0 1 2 k 1 hk , (2) WW where denotes region- or district-level fixed effects and hkk hkk denotes country-survey year effects. The districtkk kk WW = ββ + ββ XX + ββ RR + γγ EE + μμ + ττ + εε level fixed effects control for fixed spatial characteristics, whether observed or unobserved, and thus disentangle kk kk the hazard measure from many possible sources of omitted variable bias. The year effects μμ ττ further control for any common trends and thus help ensure that the relationships of interest are identified from idiosyncratic local kk ττ\\n\\n> 3 The elements of X can include the age and education level of the household head; the ethnicity, religion, total number, age, and gender composition of household members; ownership of assets such as livestock, phone, radio, TV, etc.; and characteristics of the household residence, including type and quality of water and sanitation facilities, main source of energy, and type of floor and material of construction for walls and roof.\\n\\n4 Focusing on non-monetary dimensions of welfare such as the height for age z score of children (HAZ) or the weight for height z-score (WHZ), Skoufias et al. (2023) explore an alternative specification of the vulnerability function that allows consideration of the effects of hazards through the income as well as the environmental channels. In this specification, exposure to weather-related hazards may affect the value of agricultural income earned (the income channel), as well as the health status of household members (environmental channel).\\n\\n5 In data-constrained environments like the ones we study, complex (black-box) machine learning models are unlikely to provide the gains in accuracy that justify the loss in interpretability.\\n\\n5\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"shocks. Thus the empirical approach is very much in line with the strong identification properties of the weathershock approach summarized by Dell, Jones, and Olken (2014). It should be noted that with district fixed effects, , included in the specification, the coefficients 1 can be estimated only if the values of vary within districts. kk Otherwise, the use of the district dummies will absorb the effects of these environmental factors. μμ hkk γγ EE The impact of drought on consumption expenditures is the cumulative effect of a household’s crop loss and a series of entitlement failures in a local economy that are triggered by this initial loss (Sen, 1981; Deveraux, 2007). Devereux (2007) characterizes a sequence of four entitlement failures: first, production fails as a result of the rains failing; then labor markets fail as households are less and less able to find work opportunities on other farms or in off-farm activities; then commodity markets fail as grain prices increase and prices of liquid assets decrease. Finally, transfers fail as households cannot rely on the support of others in their network who face the same constraints in meeting everyday basic needs. The formulation outlined here is agnostic as to how rainfall is impacting consumption. However, the spatial measure of drought used and the necessity of incorporating district fixed effects focuses the analysis more on picking up production failures and failures in the immediate labor and commodity local market in which the drought is measured. The impact of drought through commodity markets that span a larger geographic area or transfer networks over a larger geographic space will not be picked up. If higher levels of k indicate rainfall sufficiency, it is anticipated that 2 > 0 such that lack of rainfall and droughts have a negative effect on welfare (especially when a monetary welfare measure is used, such as consumption expenditures per adult equivalent). RR ββ̂ The extent to which hazards experienced by households in the sample are associated with sufficiently significant declines in their welfare depends in part on the degree to which there is variability in the exposure to and intensity of the shocks experienced. For example, if the value of k is the same for all in the sample and not just for those within a given cluster, then there is no way to identify the parameter 2 . Hence, it is critical to derive estimates of 2 based on data that include significant variability in RRk . This task is a particular challenge when analyzing the impact of drought because these events are spatially correlated and not localized. Increasing the number of years ββ and the number of countries included increases the variation in ββ̂ RR k . For this reason, the analysis prioritizes inclusion of countries where multiple rounds of survey data were available. In addition, by estimating this relationship at the level of the livelihood zone and not the country, the analysis increases the number of years and RR variation in hazards available.\\n\\nBased on the general but unknown function _f_ , the effect of exposure to the hazard on welfare depends on the specific values of the determinants of welfare summarized by the variables k , and , i.e., <u>hk</u> = k d <u>( k hk hk )</u> ddWW k ~~.~~ The linear specification in regression equation (2) assumes a constant effect of the hazard on RR , XXhkk EEhkk ddRR RR =rr,XX =xx,EE =ee welfare (i.e., ddRR <u>hk</u> = 2 ) that is independent of the level of the hazard and of specific household characteristics k ddWW and environmental factors . It is important to relax this assumption in two ways: first, by allowing the ddRR ββ impact of the hazard to vary by the size of the hazard deviation; and second, by allowing the impact of the hazard hkk hkk XX to vary by characteristics of the household. EE To explore the functional form of the hazard, several different approaches were adopted. Two approaches were visual: a simple plot of welfare against hazard and a binned scatterplot of the residualized welfare against the residualized hazard. Another approach estimated linear, quadratic, and cubic hazard specifications, as well as the best fractional polynomial specification, and compared the fit and plotted residuals.<sup>6</sup> A final approach explored further nonlinearities with step and piecewise linear functional forms. This involved segmenting the continuous hazard distributions into 5 (or 10) bins for use as a set of indicator variables to capture the hazard intensity. Results for linear, quadratic and cubic specifications are shown. Parsimonious specifications that best described the underlying relationship, especially in hazard range where 90 percent of the observed events occurred, were used for the simulations.\\n\\n> 6 In Stata, the _binscatter2_ command was used to visualize the overall relationship between hazard and welfare. The _fp_ and _fp plot_ commands were used to explore the fit and residuals of different polynomial specifications.\\n\\n6\", \"datasets\": [{\"mention_name\": {\"text\": \"survey data\", \"confidence\": 0.7948257327079773, \"start\": 2830, \"end\": 2841, \"clean_text\": \"survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2830, \"end\": 2841, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2830, \"end\": 2841, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2830, \"end\": 2841}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2830, \"end\": 2841, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.974418044090271, \"start\": 343, \"end\": 354}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7948257327079773, \"start\": 343, \"end\": 354}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7874933481216431, \"start\": 343, \"end\": 354}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7874933481216431, \"start\": 343, \"end\": 354}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"It is important to bear in mind that sensitivity tests for the specification of the vulnerability function are based on the hazards and extreme events contained within the sample used to estimate the vulnerability function. A maintained assumption is that the functional form estimated based on the observed and probably limited set of historic events is also valid for extrapolated events outside the range of the sample. As King and Zeng (2006) point out, extreme counterfactuals that are far from the range of hazards contained in the estimation sample are more likely to be model dependent and thus quite misleading. This is because different model specifications (e.g., linear versus quadratic) may yield comparable losses within the limited range of the estimation sample, while in reality the welfare losses may be very different for extreme hazard values that are beyond the range of the estimation sample. Some of these concerns are addressed in section 5 of the paper.\\n\\nThe impact of the hazard is allowed to vary by characteristics of the household by estimating a model that includes interaction terms of the hazard with specific household characteristics and/or some or all environmental characteristics, as in equation (3): hk 0 1 2 k 3 ) + 1 2 k hk 1hkk kk (3) 1hkk WW = ββ + ββ XX + ββ RR + ββ (RR ∗XX In this specification, it is expected that hkk hkk kk kk 3 > 0 for a household characteristic or asset that mitigates the negative γγ EE + γγ (RR ∗EE ) + μμ + ττ + εε impact of the shock on welfare (e.g., irrigated land) (or 4 > 0 for an environmental characteristic that mitigates the negative impact of the shock on welfare). Among a number of characteristics that were tried, two were ββ̂ consistently significant across specifications: land for cropping livelihood zones and tropical livestock units (TLUs) γγ� for pastoral livelihood zones. We present results by land ownership and TLUs as appropriate, but future work should include other characteristics.\\n\\nAcknowledging the costs of some out-of-sample predictive accuracy, we opted for the benefits of a simple (or adhoc) specification that allows for a more systematic and transparent comparison of the differences in estimates resulting from the hazard measure used, the specification of the functional form of the vulnerability function, and the potential heterogeneity of impacts. Country-level applications building on the findings of this study would benefit from the adoption of data-driven shrinkage or regularization methods such as Lasso, that are known to improve out-of-sample predictive accuracy.\\n\\n# 2.2. Estimating an exceedance probability curve\\n\\nA common way to communicate risk—as calculated by combining data for hazard and exposure (value at risk) with vulnerability functions—is via exceedance probabilities (EPs) and their inverse measure, return periods. An EP curve describes the probability that various levels of loss will be exceeded, in this case the welfare loss that will occur because of rainfall deficits of increasing severity. Another way to understand the EP is to consider its inverse measure: return period = 1/(exceedance probability). For example, a loss with an annual exceedance probability of 1 percent, which means that there is only a 1 percent chance of facing a worse loss, is equivalent to a loss that gives a return period of 1 in 100 years. A longer return period suggests a lower probability that the impact from an extreme hazard will occur in any single year.\\n\\nFor each hazard value in a given location, the loss is estimated using the vulnerability function and the characteristics of the households in the exposure data. To generate an EP curve, losses are ranked and combined with the probability of these drought events occurring. This calculation can be based on historical information (“as if” such events were to happen again today with current exposures) or many thousands of years of simulated/synthetic events. This paper focuses on the consumption losses that occur for below-average years, and it estimates an EP curve for both total consumption losses and consumption losses that push people into poverty or push those who were already poor further into poverty (i.e., that increase the poverty gap). Section 5 describes how this is done in practice and presents these curves.\\n\\nCatastrophe risk modeling has developed in the insurance and reinsurance sector over the past 30 years in response to potential insolvency-inducing events, such as Hurricane Andrew in 1992 in the US. Such analytics exploit increasing geographical data that captures exposure value at risk, predominantly the built environment, as 7\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"well as improved representation of hazards. Investment in research and technology in the field has been driven by the inadequacies of the historical record, mainly insurance claims data, for extreme events, and specifically by the inability of these data to robustly capture the statistics of catastrophic loss on insured portfolios. Unlike automotive or health claims, which are numerous and lend themselves to more reliable curve fitting for frequency and severity distributions, the small number of records of low-frequency and high-severity loss-making events do not provide sufficient evidence for such actuarial analysis. . While EP curves can be constructed from 20 years of observed historical loss data, they are limited to low-return-period estimation, are highly erratic and sensitive to data addition, and almost certainly underestimate loss levels, especially when considering average annual loss (Verisk 2013). Such records are rarely representative, especially in the extreme tail of the EP curve, which has few if any points to define it and is of primary importance to insurers and reinsurers. That is, history is not a good predictor of all possible future events.\\n\\nMoreover, in low-income regions of the world, there is often acute data scarcity around historical evidence of natural hazards. The resulting short time series of hazards hampers the estimation of the EP curve. The increasing availability of remotely sensed weather information from satellites has helped provide global coverage for hazardrelated data sets that now range from 10 years to 40 years in length. The range of variables inferred from satellite data has also expanded to include new ones such as soil moisture, as new sensors come online and provide reliable training and validation data.\\n\\nA key component of catastrophe modeling is developing statistical and physics-based methodologies, which vary by hazard type, to extend limited historical catalogs of events to many thousands of years of probabilistic events in order to adequately capture frequency, extreme severities, and full geographical location. This process is computationally intensive, and great care must be taken to ensure that resultant event sets are constrained by observed physical processes; for example, definition of hazard parameter distributions of climatological characteristics must maintain spatial and temporal correlations. Model calibration and validation against historical information are a valuable check, but inevitably assumptions are required on data processing and sampling steps, ideally supported by academic reference (ABI 2011). Though the event-by-event loss calculations of catastrophe modeling are computationally intensive, especially for hazards that demand high geographic resolution (which can be less than 10 m for flood), these calculations also allow differing aspects of model uncertainty to be quantified and communicated.<sup>7</sup> Only such a large time series of synthetic but nevertheless realistic events can generate enough samples of losses to robustly define the remote tail of EP curves that are of interest to the insurance industry, such as 1-in-100-year and 1-in-250-year return periods. Taking such a probabilistic loss perspective also ensures the EP is smoothed out by having many thousands of data points. For the purposes of this paper, a 10,000-year stochastic catalog of synthetic simulations was generated across Malawi at a spatial resolution of 0.05° and dekadal temporal resolution for several hazard variables (including soil moisture).<sup>8</sup> These 10,000 years of hazard events are all potential versions of what could happen in the next 12 months and can be combined with vulnerability and exposure information to estimate longer return period tail behaviors) to get a fuller risk perspective. This step requires making assumptions about the relationship between hazard and impact, using historical data as a starting point. In a data-scarce environment, it can be very hard to validate the performance of the vulnerability function in the tails. This paper is a starting point, and more investigation is required.\\n\\nAs noted above, these 10,000 years of hazard events reflect events that are likely to occur under current (and near-current) climate conditions, rather than forecasting climate conditions 100 years (for example) in the future. In this way catastrophe models differ significantly from general circulation models/global climate models (GCMs), 7 LMA (Lloyd’s Market Association), “Understanding Uncertainty in Catastrophe Modelling for Non-Catastrophe Modellers,” <u>https://www.lmalloyds.com/lma/readfile.aspx?iDocumentStorageKey=cc44f6be-b83f-4cf9-903ce802c1f312a8&iFileTypeCode=PDF&iFileName=Understanding%20uncertainty%20in%20cat%20modelling%20for% 20non-cat%20modellers.</u> 8 Given the computational requirements to generate such a stochastic catalog, this exercise was completed for only one country.\\n\\n8\", \"datasets\": [{\"mention_name\": {\"text\": \"insurance claims data\", \"confidence\": 0.9166980981826782, \"start\": 164, \"end\": 185, \"clean_text\": \"insurance claims data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 164, \"end\": 185, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 164, \"end\": 185, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 164, \"end\": 185}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 164, \"end\": 185, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.4672887623310089, \"start\": 164, \"end\": 185}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9166980981826782, \"start\": 164, \"end\": 185}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5262371301651001, \"start\": 164, \"end\": 185}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5262371301651001, \"start\": 164, \"end\": 185}}, {\"mention_name\": {\"text\": \"hazardrelated data sets\", \"confidence\": 0.49492448568344116, \"start\": 1517, \"end\": 1540, \"clean_text\": \"hazardrelated data sets\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1517, \"end\": 1540, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1517, \"end\": 1540, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1517, \"end\": 1540}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1517, \"end\": 1540, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.4497753977775574, \"start\": 587, \"end\": 610}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.49492448568344116, \"start\": 587, \"end\": 610}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5943936109542847, \"start\": 587, \"end\": 610}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5943936109542847, \"start\": 587, \"end\": 610}}, {\"mention_name\": {\"text\": \"satellite data\", \"confidence\": 0.3891794681549072, \"start\": 1630, \"end\": 1644, \"clean_text\": \"satellite data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1630, \"end\": 1644, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1630, \"end\": 1644, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1630, \"end\": 1644}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1630, \"end\": 1644, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.7096701264381409, \"start\": 700, \"end\": 714}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3891794681549072, \"start\": 700, \"end\": 714}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5772883892059326, \"start\": 700, \"end\": 714}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5772883892059326, \"start\": 700, \"end\": 714}}, {\"mention_name\": {\"text\": \"historical data\", \"confidence\": 0.7913169264793396, \"start\": 3923, \"end\": 3938, \"clean_text\": \"historical data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3923, \"end\": 3938, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3923, \"end\": 3938, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3923, \"end\": 3938}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3923, \"end\": 3938, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.8689097166061401, \"start\": 338, \"end\": 353}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7913169264793396, \"start\": 338, \"end\": 353}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5416100025177002, \"start\": 338, \"end\": 353}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5416100025177002, \"start\": 338, \"end\": 353}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"as they have been built to answer different questions about risk. Such simulated hazard event sets focus upon a primary damage discriminator, such as wind speed for storm or ground shaking for earthquake, which is subsequently translated into impact on exposure through damage functions describing vulnerability.\\n\\nA criticism of catastrophe models is that they are not forward looking regarding climate change.<sup>9</sup> Since the consequences of climate change are already present, however, any recent historical record for atmosphericrelated natural hazards comprises a climate change signal, alongside other potential modalities, such as the El Niño–Southern Oscillation (ENSO). Model sensitivity to these climate aspects can be explored in a catastrophe modeling environment and considered as another dimension of uncertainty. In recent years, “climate conditioning” of catastrophe models for differing time horizons and emissions scenarios (such as in Bates [2023]) has received much attention and is an active research frontier, attracting many new entrants to natural hazard risk modeling (see Fielder et al. 2021). Model adjustment can range from a simple shifting of event frequencies in stochastic hazard tables to much fuller incorporation of and alignment with GCM approaches. It must be remembered that there may be considerable uncertainty around future climate patterns (such as rainfall across the Sahel), and it may be challenging to adequately parameterize them without suggesting inappropriate levels of exactitude. Scenario modeling is one typical approach under high uncertainty. Additionally, exposure and vulnerability components of risk (such as urbanization, informal settlements, and flooding) can change much faster than climate trends. Since this work increasingly falls under financial regulation and disclosure regimes (e.g., NGFS 2022), explicit climate modeling for catastrophic risks will see much growth in the near future, with translation of approaches to low-income regions through use of underlying global data sets. This paper’s analysis focuses on past and current weather patterns and does not include an estimation of future changes due to global warming.\\n\\n# 3. Data on Hazards and Welfare\\n\\n## 3.1. Hazard data\\n\\nFor the analysis described in this paper, we constructed a database of historical hazard data for all SSA using remotely sensed satellite data and indices that were reputable and publicly available and that had high levels of spatial and historical coverage. The considered hazard variables include rainfall, vegetation cover, evapotranspiration, soil moisture, and the WRSI. Given there is no single definition of drought, these hazard data were used to develop different measures of climatic conditions as a proxy for drought. These measures range from simple ones based on the amount and timing of precipitation during the growing season in a defined area, to more complex composite measures combining precipitation with other information capturing the impact of increased temperatures on water availability and agricultural stress at a location. Appendix B provides more details on all the drought measures considered and the sources of the data.<sup>10</sup> Table 3.1 provides a brief overview of the measures used in the analysis. These measures by and large focus on the first three months of the growing season.\\n\\n**Table 3.1. Summary of range of hazard measures derived from remotely sensed data**\\n\\n|**Data type **|**Source**|**Hazard measures**|\\n|---|---|---|\\n|Rainfall|CHIRPS global rainfall data set from US<br>Geological Survey (NASA)|Standardized anomaly of cumulative<br>rainfall (based on the first three months of<br>thegrowingseason)|\\n|Vegetation<br>(greenness)|NDVI product of MODIS vegetation<br>indices|NDVI standardized anomaly (based on the<br>average over the first three months of the<br>growingseason)|\\n|Evapotranspiration|SPEI generated by the Climatic Research<br>Unit of the Universityof East Anglia|Level of three-month SPEI at the end of the<br>first three months of thegrowingseason|\\n\\n9 See for example Purdy (2021).\\n\\n10 A workshop was held at the beginning of the work described in this paper to ensure the right measures were included (in addition to those summarized in appendix B).\\n\\n9\", \"datasets\": [{\"mention_name\": {\"text\": \"global data sets\", \"confidence\": 0.8216919302940369, \"start\": 2039, \"end\": 2055, \"clean_text\": \"global data sets\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2039, \"end\": 2055, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2039, \"end\": 2055, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2039, \"end\": 2055}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2039, \"end\": 2055, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.4209376573562622, \"start\": 1144, \"end\": 1160}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8216919302940369, \"start\": 1144, \"end\": 1160}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5143937468528748, \"start\": 1144, \"end\": 1160}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5143937468528748, \"start\": 1144, \"end\": 1160}}, {\"mention_name\": {\"text\": \"historical hazard data\", \"confidence\": 0.4819895625114441, \"start\": 2327, \"end\": 2349, \"clean_text\": \"historical hazard data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2327, \"end\": 2349, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2327, \"end\": 2349, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2327, \"end\": 2349}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2327, \"end\": 2349, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.7094722390174866, \"start\": 545, \"end\": 567}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.4819895625114441, \"start\": 545, \"end\": 567}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4931350648403168, \"start\": 545, \"end\": 567}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4931350648403168, \"start\": 545, \"end\": 567}}, {\"mention_name\": {\"text\": \"remotely sensed satellite data and indices\", \"confidence\": 0.30811169743537903, \"start\": 2368, \"end\": 2410, \"clean_text\": \"remotely sensed satellite data and indices\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2368, \"end\": 2410, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2368, \"end\": 2410, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2368, \"end\": 2410}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2368, \"end\": 2410, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.7094722390174866, \"start\": 586, \"end\": 628}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.30811169743537903, \"start\": 586, \"end\": 628}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4931350648403168, \"start\": 586, \"end\": 628}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4931350648403168, \"start\": 586, \"end\": 628}}, {\"mention_name\": {\"text\": \"hazard data\", \"confidence\": 0.9461194276809692, \"start\": 2686, \"end\": 2697, \"clean_text\": \"hazard data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2686, \"end\": 2697, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2686, \"end\": 2697, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2686, \"end\": 2697}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2686, \"end\": 2697, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.7157454490661621, \"start\": 60, \"end\": 71}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9461194276809692, \"start\": 60, \"end\": 71}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5464814901351929, \"start\": 60, \"end\": 71}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5464814901351929, \"start\": 60, \"end\": 71}}, {\"mention_name\": {\"text\": \"remotely sensed data\", \"confidence\": 0.42285850644111633, \"start\": 3440, \"end\": 3460, \"clean_text\": \"remotely sensed data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3440, \"end\": 3460, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3440, \"end\": 3460, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3440, \"end\": 3460}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3440, \"end\": 3460, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.9807273149490356, \"start\": 5, \"end\": 25}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.42285850644111633, \"start\": 5, \"end\": 25}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7176526784896851, \"start\": 5, \"end\": 25}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7176526784896851, \"start\": 5, \"end\": 25}}, {\"mention_name\": {\"text\": \"CHIRPS global rainfall data set\", \"confidence\": 0.4148541986942291, \"start\": 3536, \"end\": 3567, \"clean_text\": \"CHIRPS global rainfall data set\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3536, \"end\": 3567, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3536, \"end\": 3567, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3536, \"end\": 3567}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3536, \"end\": 3567, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.7315965890884399, \"start\": 101, \"end\": 132}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4148541986942291, \"start\": 101, \"end\": 132}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4789806306362152, \"start\": 101, \"end\": 132}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4789806306362152, \"start\": 101, \"end\": 132}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"|Soil moisture|Soil Water Index from Copernicus Global<br>Land Service|Soil moisture standardized anomaly (based<br>on the average over the first three months<br>of thegrowingseason)|\\n|---|---|---|\\n|Availability of water<br>for crops|FEWS NET|End of season WRSI as anomaly (as a ratio<br>of the WRSI to the median historical WRSI)|\\n\\n_Source:_ World Bank.\\n\\n_Note:_ CHIRPS = Climate Hazards Group InfraRed Precipitation with Station; FEWS NET = Famine Early Warning Systems Network; MODIS = Terra Moderate Resolution Imaging Spectroradiometer; NASA = National Aeronautics and Space Administration; NDVI = Normalized Difference Vegetation Index; SPEI = Standardized Precipitation Evapotranspiration Index; WRSI = Water Requirement Satisfaction Index.\\n\\nThe historical coverage of the data differs. In general, the analysis uses data from the period 2000/01 to 2019/20 at dekadal or monthly intervals to reflect the shocks faced within the context of a changing climate. For example, the NDVI data utilized are a product of the MODIS (Terra Moderate Resolution Imaging Spectroradiometer) satellites and are available for February 2000 and after. However, the precipitation data are from the CHIRPS (Climate Hazards Group InfraRed Precipitation with Station) global rainfall data set and are available from 1981. Only the soil moisture data cover a shorter period, starting in 2007. Despite the limited history, however, soil moisture data are rapidly proving useful in estimating drought conditions, and they have a memory effect that rainfall data lack.\\n\\nThe hazard data were extracted at a uniform grid resolution of 0.05° by 0.05° (about 5 km by 5 km near the equator) and a crop coverage mask (Copernicus Global Land Cover Layers, 2019) was applied to exclude hazard values in areas that would not affect crop yields. Data on seasonal timing were taken from the UN Food and Agriculture Organization (FAO) for each country. To aggregate the hazard data to an appropriate geographic unit for merging with the welfare data, additional data were incorporated in the data set, including country and administrative area boundaries from the World Bank Group, Global Administration Areas (GADM), and Global Administrative Unit Layers (GAUL) of the FAO.\\n\\nThe standardized anomaly measure is defined as _(level – mean for local area)/standard deviation_ for the local area, where the mean and standard deviation for the local area are calculated based on the 20 years of hazard data, or the maximum data available in the case of soil moisture. The local area is defined by aggregating the data for a 20 km buffer of the household location or for the administrative area if the former is not available.\\n\\nThe unconditional correlation between different hazard measures is not particularly high. Table 3.2 presents results for each livelihood zone used in the analysis. The independent variation across these hazard measures highlights their capture of different atmospheric conditions; hence the difficulty of identifying a measure capturing drought events well. But as the analysis will show, in a regression framework the results are more consistent across most hazard measures (with some exceptions). This suggests that despite capturing different atmospheric conditions, the hazard measures are reasonably effective at capturing situations making households worse off.\\n\\n10\", \"datasets\": [{\"mention_name\": {\"text\": \"Soil Water Index\", \"confidence\": 0.40075090527534485, \"start\": 15, \"end\": 31, \"clean_text\": \"Soil Water Index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 15, \"end\": 31, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 15, \"end\": 31, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 15, \"end\": 31}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 15, \"end\": 31, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.6928057670593262, \"start\": 15, \"end\": 31}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.40075090527534485, \"start\": 15, \"end\": 31}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6417284607887268, \"start\": 15, \"end\": 31}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6417284607887268, \"start\": 15, \"end\": 31}}, {\"mention_name\": {\"text\": \"Soil moisture standardized anomaly\", \"confidence\": 0.45776641368865967, \"start\": 71, \"end\": 105, \"clean_text\": \"Soil moisture standardized anomaly\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 71, \"end\": 105, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 71, \"end\": 105, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 71, \"end\": 105}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 71, \"end\": 105, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.6928057670593262, \"start\": 71, \"end\": 105}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.45776641368865967, \"start\": 71, \"end\": 105}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6417284607887268, \"start\": 71, \"end\": 105}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6417284607887268, \"start\": 71, \"end\": 105}}, {\"mention_name\": {\"text\": \"global rainfall data set\", \"confidence\": 0.32338228821754456, \"start\": 1253, \"end\": 1277, \"clean_text\": \"global rainfall data set\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1253, \"end\": 1277, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1253, \"end\": 1277, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1253, \"end\": 1277}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1253, \"end\": 1277, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.3297692537307739, \"start\": 537, \"end\": 561}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.32338228821754456, \"start\": 537, \"end\": 561}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5427637696266174, \"start\": 537, \"end\": 561}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5427637696266174, \"start\": 537, \"end\": 561}}, {\"mention_name\": {\"text\": \"soil moisture data\", \"confidence\": 0.6485539078712463, \"start\": 1415, \"end\": 1433, \"clean_text\": \"soil moisture data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1415, \"end\": 1433, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1415, \"end\": 1433, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1415, \"end\": 1433}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1415, \"end\": 1433, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8934480547904968, \"start\": 699, \"end\": 717}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6485539078712463, \"start\": 699, \"end\": 717}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6502367258071899, \"start\": 699, \"end\": 717}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6502367258071899, \"start\": 699, \"end\": 717}}, {\"mention_name\": {\"text\": \"hazard data\", \"confidence\": 0.8601143956184387, \"start\": 1555, \"end\": 1566, \"clean_text\": \"hazard data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1555, \"end\": 1566, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1555, \"end\": 1566, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1555, \"end\": 1566}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1555, \"end\": 1566, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6233149766921997, \"start\": 0, \"end\": 11}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.44181281328201294, \"start\": 0, \"end\": 11}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7446168661117554, \"start\": 0, \"end\": 11}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7446168661117554, \"start\": 0, \"end\": 11}}, {\"mention_name\": {\"text\": \"Copernicus Global Land Cover Layers\", \"confidence\": 0.48438921570777893, \"start\": 1693, \"end\": 1728, \"clean_text\": \"Copernicus Global Land Cover Layers\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1693, \"end\": 1728, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1693, \"end\": 1728, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1693, \"end\": 1728}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1693, \"end\": 1728, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.47621217370033264, \"start\": 138, \"end\": 173}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.48438921570777893, \"start\": 138, \"end\": 173}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6562626957893372, \"start\": 138, \"end\": 173}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6562626957893372, \"start\": 138, \"end\": 173}}, {\"mention_name\": {\"text\": \"Global Administrative Unit Layers\", \"confidence\": 0.3814873695373535, \"start\": 2191, \"end\": 2224, \"clean_text\": \"Global Administrative Unit Layers\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2191, \"end\": 2224, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2191, \"end\": 2224, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2191, \"end\": 2224}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2191, \"end\": 2224, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.765300452709198, \"start\": 636, \"end\": 669}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3814873695373535, \"start\": 636, \"end\": 669}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.8225960731506348, \"start\": 636, \"end\": 669}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8225960731506348, \"start\": 636, \"end\": 669}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"**Table 3.2. Unconditional correlation between different hazard measures by livelihood zone**\\n\\n|||**Ma**|**ize**|||**Highl**|**and**||**R**|**oots and**|**lowland**|**s**|\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n||Total||||Total||||Total||||\\n||<br>seasonal<br>rainfall<br>|Soil<br>moistu|re<sup>SPEI</sup>|NDVI|<br>seasonal<br>rainfall|Soil<br>moistur|e <sup>SPEI </sup>|<sup>NDVI</sup>|<br>seasonal<br>rainfall|<br>Soil<br>moisture|<sup>SPEI</sup>|NDVI|\\n|Soil moisture|0.43||||0.49||||0.00||||\\n|SPEI|0.47|0.26|||0.44|0.28|||0.30|0.12|||\\n|NDVI|0.48|0.63|<br>0.32||0.32|0.51|0.18||0.22|0.43|0.07||\\n|WRSI|0.21|0.13|<br>0.12|0.16|0.22|0.15|0.18|0.11|0.15|-0.11|0.01|0.02|\\n||||**Pastoral**|||||**Ag**|**ropastora**|**l**|||\\n||Total seas<br>rainfal|onal<br>l|Soil<br>moisture|SPEI|<br>NDVI|To|tal seas<br>rainfall|onal<br> <br>|Soil<br>moisture|SPEI|NDVI||\\n|Soil moisture|0.46||||||0.32||||||\\n|SPEI|0.45||0.23||||0.51||0.39||||\\n|NDVI|0.57||0.54|0.28|||0.33||0.60|0.31|||\\n|WRSI|0.29||0.09|0.24|<br>0.26||0.26||-0.11|0.17|0.10||\\n\\n_Source:_ World Bank. _Note:_ NDVI = Normalized Difference Vegetation Index; SPEI = Standardized Precipitation Evapotranspiration Index; WRSI = Water Requirement Satisfaction Index.\\n\\n# 3.2. Hazard simulations data\\n\\nFor the analysis, we asked Verisk’s AIR Worldwide (AIR) to produce 10,000-year drought-related stochastic catalogs of synthetic simulations for Malawi at a spatial resolution of 0.05° and dekadal temporal resolution. Stochastic catalogs are designed to provide a view of hazard that goes beyond the data available in the historical record. AIR used historical data on rainfall to produce three such catalogs: for precipitation, soil moisture, and vegetation index.\\n\\nTo generate the stochastic catalogs, AIR used data similar to that used by the World Bank team in generating the historical hazard time series. CHIRPS data were used for precipitation; the Soil Water Index at a depth of 40 cm from the Copernicus Global Land Service was used for soil moisture; and the NDVI for vegetation cover was from MODIS. The resulting catalogs were validated by comparing the synthetic time series with the historical data in Malawi. The validation results show that the catalogs preserve the intrinsic properties in terms of spatial and temporal correlations, produce a realistic view of hazard, and are consistent with historical observations while allowing for the creation of new extremes.\\n\\nFor the purposes of this paper, the approach was tested using the stochastic catalog for soil moisture in Malawi. Further details on the methodology used to generate the 10,000-year catalog and the validation conducted are in appendix C.\\n\\n11\", \"datasets\": [{\"mention_name\": {\"text\": \"historical data on rainfall\", \"confidence\": 0.6616079807281494, \"start\": 1601, \"end\": 1628, \"clean_text\": \"historical data on rainfall\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1601, \"end\": 1628, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1601, \"end\": 1628, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1601, \"end\": 1628}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1601, \"end\": 1628, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.8223326802253723, \"start\": 630, \"end\": 657}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6616079807281494, \"start\": 630, \"end\": 657}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7046367526054382, \"start\": 630, \"end\": 657}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7046367526054382, \"start\": 630, \"end\": 657}}, {\"mention_name\": {\"text\": \"CHIRPS data\", \"confidence\": 0.6283562183380127, \"start\": 1862, \"end\": 1873, \"clean_text\": \"CHIRPS data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1862, \"end\": 1873, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1862, \"end\": 1873, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1862, \"end\": 1873}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1862, \"end\": 1873, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.44492605328559875, \"start\": 270, \"end\": 281}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6283562183380127, \"start\": 270, \"end\": 281}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7483195662498474, \"start\": 270, \"end\": 281}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7483195662498474, \"start\": 270, \"end\": 281}}, {\"mention_name\": {\"text\": \"NDVI\", \"confidence\": 0.3080144226551056, \"start\": 2020, \"end\": 2024, \"clean_text\": \"NDVI\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2020, \"end\": 2024, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2020, \"end\": 2024, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2020, \"end\": 2024}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2020, \"end\": 2024, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.44492605328559875, \"start\": 428, \"end\": 432}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3080144226551056, \"start\": 428, \"end\": 432}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7483195662498474, \"start\": 428, \"end\": 432}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7483195662498474, \"start\": 428, \"end\": 432}}, {\"mention_name\": {\"text\": \"historical data in Malawi\", \"confidence\": 0.37496286630630493, \"start\": 2148, \"end\": 2173, \"clean_text\": \"historical data in Malawi\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2148, \"end\": 2173, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2148, \"end\": 2173, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2148, \"end\": 2173}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2148, \"end\": 2173, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.8021473288536072, \"start\": 556, \"end\": 581}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.37496286630630493, \"start\": 556, \"end\": 581}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4621736407279968, \"start\": 556, \"end\": 581}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4621736407279968, \"start\": 556, \"end\": 581}}, {\"mention_name\": {\"text\": \"stochastic catalog\", \"confidence\": 0.41101181507110596, \"start\": 2502, \"end\": 2520, \"clean_text\": \"stochastic catalog\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2502, \"end\": 2520, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2502, \"end\": 2520, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2502, \"end\": 2520}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2502, \"end\": 2520, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.7657689452171326, \"start\": 46, \"end\": 64}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.41101181507110596, \"start\": 46, \"end\": 64}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9137471318244934, \"start\": 46, \"end\": 64}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9137471318244934, \"start\": 46, \"end\": 64}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"# 3.3. Welfare data\\n\\nThe main welfare measures examined are constructed from household consumption surveys. To provide a comparable measure of welfare across countries, the aggregate that is used for national poverty measures and that forms the basis for global poverty measurement is used. Consumption aggregates are converted into 2011 purchasing power parities (PPPs) for this analysis.<sup>11</sup> We selected nine countries representing geographic variation in SSA for the analysis of the relationship between droughts and monetary welfare (consumption). They include Ethiopia in Eastern Africa; Mauritania, Niger, and Nigeria in Western Africa; and Lesotho, Malawi, Mozambique, Zambia, and Zimbabwe in Southern Africa. For five of these countries data are available for multiple years. The countries, years, and outcome measures used in analysis are listed in Table 3.3. To allow for ease of matching rainfall patterns to consumption in the first analysis, the selected countries are all from parts of the continent where there is a clear main season of production. Selection was also designed to include countries that had multiple survey rounds that could be pooled for analysis and that would provide sufficient numbers for analysis across livelihood zones.\\n\\nThe empirical analysis is carried out by livelihood zone, which entails pooling these countries and estimating like areas of production within them. The zones used are those defined in Dixon et al. (2019). They allow for the required level of analysis. The authors provide gridded data for the livelihood zones estimated for 2015, which is within our analysis period. We further aggregated the zones estimated for 2015 into five zones. Table 3.4 indicates how the countries included match to the livelihood zones used in the analysis. Table 3.5 reports descriptive statistics for the variables used in the analysis by country and survey year.\\n\\n**Table 3.3. Household surveys and welfare measures included in the analysis**\\n\\n|**Country (years)**|**Welfare measure**|\\n|---|---|\\n|Ethiopia (2010/11, 2014/15)|Total expenditure per adult equivalent<br>Food expenditure per adult equivalent<br>Nonfood expenditureper adult equivalent|\\n|Lesotho(2017/18)|Total expenditureper capita|\\n|Malawi (2010/11, 2016/17,<br>2019/20)|Total expenditure per capita<br>Total food expenditure per capita<br>Total nonfood expenditure per capita<br>Household dietary diversity: based on 12 food categories<br>Household food consumption score: based on food types consumed and<br>frequencyof consumption in a seven-day period(maximum 120)|\\n|Mauritania(2014,2019/20)|Total expenditureper capita|\\n|Mozambique (2014)|Total expenditure per capita<br>Total food expenditure per capita<br>Total nonfood expenditureper capita|\\n|Niger (2018)|Total expenditure per capita<br>Total food expenditure per capita<br>Total nonfood expenditureper capita|\\n|Nigeria(2011,2013,2016)|Total expenditureper capita|\\n|Zambia(2015)|Total expenditureper capita|\\n|Zimbabwe(2017)|Total expenditureper capita|\\n\\n_Source: World Bank._\\n\\n> 11 In most countries, the aggregates take spatial variation in prices into account and where this is done, this is retained. In addition, if the consumption aggregate used for national poverty measurement uses per adult equivalent instead of per capita, this is also retained. This makes the aggregate different from the ones reported in the World Bank’s Poverty and Inequality Platform.\\n\\n12\", \"datasets\": [{\"mention_name\": {\"text\": \"household consumption surveys\", \"confidence\": 0.8566541075706482, \"start\": 77, \"end\": 106, \"clean_text\": \"household consumption surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 77, \"end\": 106, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 77, \"end\": 106, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 77, \"end\": 106}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 77, \"end\": 106, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9571431279182434, \"start\": 77, \"end\": 106}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8566541075706482, \"start\": 77, \"end\": 106}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.999182403087616, \"start\": 77, \"end\": 106}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.999182403087616, \"start\": 77, \"end\": 106}}, {\"mention_name\": {\"text\": \"Total food expenditure per capita\", \"confidence\": 0.30972686409950256, \"start\": 2809, \"end\": 2842, \"clean_text\": \"Total food expenditure per capita\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2809, \"end\": 2842, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2809, \"end\": 2842, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2809, \"end\": 2842}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2809, \"end\": 2842, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.5980542898178101, \"start\": 4, \"end\": 37}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.30972686409950256, \"start\": 4, \"end\": 37}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6663607954978943, \"start\": 4, \"end\": 37}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6663607954978943, \"start\": 4, \"end\": 37}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"**Table 3.4. Matching of household surveys to livelihood zones**\\n\\n||Ethiopia|Lesotho|Malawi|Mauritania|Mozambique|Niger|Nigeria|Zambia|Zimbabwe|Total|\\n|---|---|---|---|---|---|---|---|---|---|---|\\n|Maize mixed|2,913|377|26,172|0|2,610|0|0|5,580|12,614|50,266|\\n|Highland|10,651|2,444|0|0|227|0|29|0|1,695|15,046|\\n|Roots and<br>lowland|0|0|0|0|0|0|5,705|0|0|5,705|\\n|Agropastoral|0|0|0|0|0|2,315|2,278|0|0|4,593|\\n|Pastoral|0|0|0|11,463|0|2,178|0|0|0|13,641|\\n\\n_Source: World Bank. Note:_ Livelihood zones are generated from Dixon et al. (2019). Agropastoral and pastoral regions in Ethiopia are not included in the analysis. Pastoral includes pastoral and arid pastoral oasis farming systems. Highland includes highland perennial and highland mixed farming systems. Roots and lowland combines root and tuber crop, cereal root crop, and humid lowland tree crop farming systems.\\n\\n**Table 3.5. Descriptive statistics for variables by** **<u>year and country</u>**\\n\\n|**Year**|**Country**|**Average per**<br>**capita**<br>**consumption**<sup>**a**</sup><br>**(2011 US$ PPP)**|**Share of**<br>**households**<br>**with land**<br>**ownership**<sup>**b**</sup>|**Share of**<br>**households**<br>**with**<br>**educated**<br>**head**<sup>**b**</sup>|**Share of**<br>**households**<br>**with**<br>**female**<br>**head**|**Share of**<br>**households**<br>**interviewed**<br>**during the lean**<br>**season**|**Share of**<br>**households**<br>**in locations**<br>**considered**<br>**remote**|\\n|---|---|---|---|---|---|---|---|\\n|2011|Ethiopia|950|0.98|0.37|0.15|0.34|0.68|\\n|2016|Ethiopia|1,227|0.96|0.38|0.17|0.35|0.69|\\n|2017|Lesotho|1,281|0.79|0.16|0.39|0.48|0.44|\\n|2010|Malawi|590|0.94|0.45|0.20|0.42|0.21|\\n|2016|Malawi|543|0.90|0.47|0.27|0.58|0.19|\\n|2019|Malawi|498|0.90|0.50|0.29|0.42|0.22|\\n|2014|Mauritania|1,782|0.44|0.36|0.27|0.96|0.54|\\n|2019|Mauritania|1,755|0.42|0.36|0.33|0.00|0.55|\\n|2014|Mozambique|626|0.96|0.49|0.22|0.12|0.67|\\n|2018|Niger|869|0.88|0.06|0.13|0.42|0.16|\\n|2011|Nigeria|738|0.82|0.46|0.08|0.55|0.19|\\n|2013|Nigeria|873|0.84|0.43|0.08|0.54|0.19|\\n|2016|Nigeria|859|0.84|0.46|0.06|0.54|0.19|\\n|2015|Zambia|486|0.89|0.65|0.18|0.67|0.72|\\n|2017|Zimbabwe|950|0.68|0.47|0.34|0.40|0.69|\\n\\n_Source: World Bank. Notes_ : a. Or per adult equivalent consumption, depending on the country. b. See appendix B, table B.2 for country-specific definitions used.\\n\\n13\", \"datasets\": [{\"mention_name\": {\"text\": \"household surveys\", \"confidence\": 0.7913339138031006, \"start\": 25, \"end\": 42, \"clean_text\": \"household surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 25, \"end\": 42, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 25, \"end\": 42, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 25, \"end\": 42}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 25, \"end\": 42, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.7954288721084595, \"start\": 25, \"end\": 42}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7913339138031006, \"start\": 25, \"end\": 42}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5017734169960022, \"start\": 25, \"end\": 42}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5017734169960022, \"start\": 25, \"end\": 42}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"_Note:_ Squares represent the 5 km by 5 km grids for which the hazard measures are available; the (starred) centroid of the square represents the household or cluster location; shaded squares represent the grids used to spatially aggregate the hazard measure for a 10 km radius (darker shade) and a 20 km radius (lighter shade) around the household’s location.\\n\\nThe results presented in this paper are based on matching a household’s geographic location with a spatially aggregated hazard measure within a 20 km radius. In the absence of any GPS data at the household or cluster level, the hazard measures are aggregated to the geographic or closest administrative level that is coded in the household survey. We chose the level of spatial aggregation so as to provide a roughly consistent area of spatial aggregation for surveys where matching was done using GPS data and administrative data.\\n\\nInitial analysis suggested that this level of spatial aggregation was appropriate. We compared the sensitivity of the relationship between welfare and hazard measures aggregated using different radii in the case of Malawi, a country that has GPS coordinates for surveyed households and multiple years of data. The statistical significance of results increased for radii of up to 20 km and then started to decrease again with larger radii, suggesting that a radius of 20 km is able to reflect the local shock to a household’s crop yields and is also large enough to capture some localized impacts on prices and wages in the local market area. However, as shown in section 4.2, for the full data set a larger geographic area it is not clear that 20 km works better in every setting, and there is very little difference between the results across aggregating radii.\\n\\n# 3.5. Using standardized anomalies for analysis\\n\\nThe hazard measures generated are in levels as well as standardized anomalies (z-scores).<sup>13</sup> The latter assume that differences in the level of the hazard from the historic average (i.e., level changes) matter not in an absolute sense but in proportion to an area’s usual variation. Welfare measures are only partly affected by the relation between hazards and crop yields and are the outcome of various ex ante and ex post risk-coping strategies developed by households over the years for the purpose of reducing income fluctuations and/or protecting their level of welfare (e.g., consumption). Hence the extent to which the experienced hazard deviates from what households expect (i.e., the historic mean value of the hazard) appears to be a better specification than the actual level of the hazard. Thus it is anticipated that the specification of shocks as standardized anomalies (in standard deviation units or z- scores) is likely to yield more sensible welfare impact estimates than those obtained using the level of the hazard (i.e., rainfall in millimeters). In the case of droughts, a positive anomaly means the hazard experienced was less severe than the historical mean, and a negative anomaly means it was more severe.\\n\\nWhen analysis of the specification of the hazard level as a level includes enough years of welfare data, controlling for fixed effects at some spatial level (e.g., district level) transforms the hazard and all other control variables as deviations from the mean (Dell, Jones, and Olken 2014). In cases such as ours, where the sample of welfare data contains a limited number of observations over time, this is not the case, and it is preferable to use a full time series of hazard data prior to regression analysis to generate a standardized anomaly.<sup>14</sup> 13 Some hazard measures such as the WRSI cannot be expressed as a standardized anomaly because by construction they are top coded (for example, the top value of WRSI is 100 percent). However, WRSI can be expressed as an anomaly (i.e., the ratio of current value with the mean or median).\\n\\n14 Consider, for example, a cross-sectional household data set with three survey rounds and an identifier for the same districts (or clusters, though that is less likely) in each round. Using fixed effects for districts (clusters) transforms all the variables as deviations from the sample district (cluster) mean over all the household observations across the years in the same district (cluster) within the sample. With three survey rounds over the 20-year period analyzed, the district-specific sample mean of any given hazard measured is likely to differ significantly from the district mean estimated over a 20-year period. As a consequence, estimates of the welfare impacts of shocks are likely to differ significantly depending on whether the analysis uses the specification of the 15\", \"datasets\": [{\"mention_name\": {\"text\": \"household survey\", \"confidence\": 0.4379176199436188, \"start\": 692, \"end\": 708, \"clean_text\": \"household survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 692, \"end\": 708, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 692, \"end\": 708, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 692, \"end\": 708}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 692, \"end\": 708, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9017645716667175, \"start\": 692, \"end\": 708}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4379176199436188, \"start\": 692, \"end\": 708}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5426437854766846, \"start\": 692, \"end\": 708}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5426437854766846, \"start\": 692, \"end\": 708}}, {\"mention_name\": {\"text\": \"GPS data\", \"confidence\": 0.8621429800987244, \"start\": 860, \"end\": 868, \"clean_text\": \"GPS data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 860, \"end\": 868, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 860, \"end\": 868, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 860, \"end\": 868}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 860, \"end\": 868, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.7842404842376709, \"start\": 860, \"end\": 868}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8621429800987244, \"start\": 860, \"end\": 868}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.49793344736099243, \"start\": 860, \"end\": 868}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.49793344736099243, \"start\": 860, \"end\": 868}}, {\"mention_name\": {\"text\": \"administrative data\", \"confidence\": 0.33234065771102905, \"start\": 873, \"end\": 892, \"clean_text\": \"administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 873, \"end\": 892, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 873, \"end\": 892, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 873, \"end\": 892}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 873, \"end\": 892, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.7842404842376709, \"start\": 873, \"end\": 892}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.33234065771102905, \"start\": 873, \"end\": 892}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.49793344736099243, \"start\": 873, \"end\": 892}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.49793344736099243, \"start\": 873, \"end\": 892}}, {\"mention_name\": {\"text\": \"welfare data\", \"confidence\": 0.9742267727851868, \"start\": 3388, \"end\": 3400, \"clean_text\": \"welfare data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3388, \"end\": 3400, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3388, \"end\": 3400, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3388, \"end\": 3400}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3388, \"end\": 3400, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6788932681083679, \"start\": 990, \"end\": 1002}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9742267727851868, \"start\": 990, \"end\": 1002}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5157141089439392, \"start\": 990, \"end\": 1002}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5157141089439392, \"start\": 990, \"end\": 1002}}, {\"mention_name\": {\"text\": \"cross-sectional household data set\", \"confidence\": 0.5735705494880676, \"start\": 3933, \"end\": 3967, \"clean_text\": \"cross-sectional household data set\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3933, \"end\": 3967, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3933, \"end\": 3967, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3933, \"end\": 3967}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3933, \"end\": 3967, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9470410943031311, \"start\": 754, \"end\": 788}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5735705494880676, \"start\": 754, \"end\": 788}, \"usage_context\": {\"text\": \"background\", \"confidence\": 0.37100398540496826, \"start\": 754, \"end\": 788}, \"is_used\": {\"text\": \"False\", \"confidence\": 0.37100398540496826, \"start\": 754, \"end\": 788}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"Another advantage offered by the standardization of the deviation from the local mean is that it offers the opportunity to gauge the magnitude of the shock experienced across a variety of geographic contexts. We use z- scores for the hazard anomalies, which means that in addition to being differenced from the mean, the hazard variable is scaled by the standard deviation. This approach assumes that changes in the hazard matter not in an absolute sense, but in proportion to an area’s usual variation. This assumption is useful in taking some production and livelihood differences into account across space. A 3 mm decrease in rain does not have the same impact in wet and dry regions. Considering variations relative to an area’s usual variation is more important when there is no control for this in the hazard measure being used, e.g., total rainfall, soil moisture. It is arguably less important when using an outcome measure that already takes crop choice into account, such as NDVI or WRSI.\\n\\nTable 3.6 presents the means and standard deviations for the hazard variables used in the analysis. The means indicate lower than average years for many of the hazards across the livelihood zones. The standard deviations highlight substantial variation around these means. It is this variation that the analysis exploits.\\n\\n**Table 3.6. Mean and standard deviation of hazards used in the analysis**\\n\\n|**Hazard variable**|**Maize**|**Highland**|**Roots**|**Pastoral**|**Agropastoral**|\\n|---|---|---|---|---|---|\\n|||**Growin**|**g season**|||\\n|Soil moisture|-0.312|-0.322|-0.293|0.098|-0.359|\\n||(0.929)|(1.126)|(0.769)|(1.086)|(1.000)|\\n|NDVI|-0.284|-0.349|-0.472|0.334|0.058|\\n||(0.949)|(1.038)|(0.923)|(1.046)|(0.955)|\\n|SPEI|0.111|-0.345|-0.600|0.367|0.067|\\n||(0.721)|(0.820)|(0.850)|(0.834)|(1.183)|\\n|Rainfall|0.035|-0.378|-0.291|0.417|-0.002|\\n||(1.135)|(1.170)|(0.909)|(0.980)|(1.183)|\\n|WRSI|99.050|98.860|99.339|101.488|100.866|\\n||(9.785)|(9.789)|(2.118)|(11.026)|(10.109)|\\n|**Rai**|**ny season(us**|**ed inpastoral**|**and agropast**|**oral specificatio**|**ns)**|\\n|Soil moisture||||0.693|0.610|\\n|||||(0.497)|(0.173)|\\n|NDVI||||0.396|0.161|\\n|||||(1.038)|(0.852)|\\n|Rainfall||||0.205|0.334|\\n|||||(0.942)|(0.826)|\\n\\n# _Source:_ World Bank.\\n\\n_Note:_ NDVI = Normalized Difference Vegetation Index; SPEI = Standardized Precipitation Evapotranspiration Index; WRSI = Water Requirement Satisfaction Index. Countries and surveys included are: Ethiopia (2011, 2016), Lesotho (2017), Malawi (2010, 2016, 2019), Mauritania (2014, 2109/20), Mozambique (2014), Niger (2018), Nigeria (2011, 2013, 2016), Zambia (2015), and Zimbabwe (2017).\\n\\nhazard as a level along with the fixed effects method or it uses the specification of the hazard as a standardized anomaly from the local normal (over the 20-year period). In fact, in the extreme case in which only one crosssectional round of survey data is available, then the specification of the hazard as a standardized anomaly is the only sensible option, since in this instance, use of the level of hazard with the fixed effects for districts (for example) would simply express the hazard as a deviation from the district mean of the hazard values experienced in the same survey round by other households in the survey.\\n\\n16\", \"datasets\": [{\"mention_name\": {\"text\": \"Water Requirement Satisfaction Index\", \"confidence\": 0.31623339653015137, \"start\": 2359, \"end\": 2395, \"clean_text\": \"Water Requirement Satisfaction Index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2359, \"end\": 2395, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2359, \"end\": 2395, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2359, \"end\": 2395}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2359, \"end\": 2395, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9144878387451172, \"start\": 223, \"end\": 259}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.31623339653015137, \"start\": 223, \"end\": 259}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6115525960922241, \"start\": 223, \"end\": 259}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6115525960922241, \"start\": 223, \"end\": 259}}, {\"mention_name\": {\"text\": \"survey data\", \"confidence\": 0.4483599364757538, \"start\": 2868, \"end\": 2879, \"clean_text\": \"survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2868, \"end\": 2879, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2868, \"end\": 2879, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2868, \"end\": 2879}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2868, \"end\": 2879, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9693098664283752, \"start\": 231, \"end\": 242}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4483599364757538, \"start\": 231, \"end\": 242}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6976752281188965, \"start\": 231, \"end\": 242}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6976752281188965, \"start\": 231, \"end\": 242}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"# 4. Results\\n\\n## 4.1. Main results\\n\\nTables 4.1 to 4.3 show results for continuous measures of hazard anomalies for the three cropping livelihood zones of maize, highland, and roots and lowland. Each table shows results for linear, quadratic, and cubic measures of soil moisture, vegetation (NDVI), evapotranspiration (SPEI), total seasonal rainfall, and WRSI; notes to each table indicate which countries and years are included. The impact of the hazard is estimated and presented separately for households with and without land to reflect the difference in the degree to which households are likely to be impacted by the hazard.<sup>15</sup> There are three important conclusions from these results. First, not all measures perform equally well at capturing the occurrence of the shock, to the extent that welfare is positively and significantly correlated with good weather conditions and negatively affected by weather shocks for households with land. Of the four measures, soil moisture and NDVI appear to give more precise and consistent results across specifications, with soil moisture offering the most consistent results across livelihood zones. SPEI, total seasonal rainfall and WRSI provide much less consistent results. Total seasonal rainfall often provides counterintuitive results, as does SPEI in some specifications. WRSI is often not significant.\\n\\nSecond, although nonlinearities are known to be present in the relationship between weather and agricultural production, we do not observe a consistent nonlinear pattern across regression results. This could simply reflect the fact that we are looking at the relationship between weather and consumption rather than agricultural production where non-linear effects are well established (e.g., Lobell, Schlenker and Costa-Roberts 2011). It could also reflect the fact that non-linearities are already part of the index construction in some cases. For example, WRSI weights losses in rainfall differently based on when they occur and NDVI, by capturing vegetation should, already capture the non-linear impact of weather conditions on yields. Another reason for not observing a consistent non-linear pattern could be that we have few household survey observations at extreme values of the anomaly which is where observations would be needed to pick up non-linear effects.\\n\\nIn the maize livelihood zone some nonlinearity is present: there is a stronger relationship between hazard anomalies and welfare in the middle of the distribution (where anomalies and therefore welfare changes are smaller) and a smaller impact of anomalies on welfare at the tails of the distribution. A steeper relationship is also observed in the quadratic and cubic specification for the roots and lowland livelihood zone, but no tapering off for negative values (Figures 4.1 and 4.2). Despite being significant in this case, the squared and cubic terms do not indicate much nonlinearity in the relationship. Nonlinearities are much more important when using the total rainfall measure; this makes sense because this measure does not reflect any nonlinear relationship between rainfall and crop production in how it is constructed. In these cases, the nonlinearity indicates a larger impact on welfare as the anomaly becomes bigger and more negative, and a smaller impact on welfare as the anomaly becomes more positive, even turning negative for very large values of total rainfall that may reflect waterlogging and flooding.\\n\\nThird, for soil moisture negative impacts are felt for both landed and landless households with no consistent pattern of who is most affected. For the highlands and NDVI measures in roots and tubers livelihood area, there is a strong and significant impact on landed households, but little impact on landless households. Overall, though given the relatively similar impact in many specifications it suggests that using weather variation in a 20km radius includes the impact of the weather on immediate labor markets.\\n\\n15 The term “hazard” as used here includes the full distribution of the weather anomaly used—that is, positive and negative events, not just events that cause welfare losses according to the definition of hazard.\\n\\n17\", \"datasets\": [{\"mention_name\": {\"text\": \"household survey observations\", \"confidence\": 0.9180063605308533, \"start\": 2198, \"end\": 2227, \"clean_text\": \"household survey observations\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2198, \"end\": 2227, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2198, \"end\": 2227, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2198, \"end\": 2227}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2198, \"end\": 2227, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9888637065887451, \"start\": 486, \"end\": 515}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9180063605308533, \"start\": 486, \"end\": 515}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5028709173202515, \"start\": 486, \"end\": 515}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5028709173202515, \"start\": 486, \"end\": 515}}, {\"mention_name\": {\"text\": \"NDVI measures in roots and tubers livelihood area\", \"confidence\": 0.3723175525665283, \"start\": 3633, \"end\": 3682, \"clean_text\": \"NDVI measures in roots and tubers livelihood area\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3633, \"end\": 3682, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3633, \"end\": 3682, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3633, \"end\": 3682}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3633, \"end\": 3682, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9131934642791748, \"start\": 193, \"end\": 242}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3723175525665283, \"start\": 193, \"end\": 242}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8522637486457825, \"start\": 193, \"end\": 242}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8522637486457825, \"start\": 193, \"end\": 242}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"**Table 4.1. Results for maize livelihood zone**\\n\\n|||**Soil moistu**|**re**||**NDVI**|||**SPEI**||**Total**|**seasonal rain**|**fall**||**WRSI**||\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|(9)|(10)|(11)|(12)|(13)|(14)|(15)|\\n|**Landed h**|**ouseholds**|||||||||||||||\\n|Hazard|**0.055*****|**0.059*****|**0.119*****|**0.024****|**0.034*****|**0.071*****|0.019|0.019|**0.062*****|-0.019|-0.017|0.012|0.001|0.012|-0.015|\\n||**[3.852]**|**[4.097]**|**[4.792]**|**[2.122]**|**[2.755]**|**[4.120]**|[1.616]|[1.547]|**[2.843]**|[-1.570]|[-1.407]|[0.701]|[0.685]|[1.169]|[-0.433]|\\n|Hazard<sup>2</sup>||**0.022***|0.011||**0.020*****|-0.001||0.003|-0.004||**-0.022*****|**-0.015****||-0.000|0.000|\\n|||**[1.831]**|[0.839]||**[2.807]**|[-0.082]||[0.243]|[-0.250]||**[-3.021]**|**[-1.964]**||[-1.162]|[0.653]|\\n|Hazard<sup>3</sup>|||**-0.028*****|||**-0.021*****|||**-0.021*****|||-0.011**|||-0.000|\\n||||**[-2.954]**|||**[-3.814]**|||**[-2.600]**|||[-2.205]|||[-0.863]|\\n|**Landless**|**households**|||||||||||||||\\n|Hazard|**0.043****|**0.038***|0.032|0.013|**0.030***|**0.077****|**-0.028****|**-0.041****|0.071|**-0.045*****|**-0.042*****|**-0.075****|**-0.003*****|0.000|-0.016|\\n||**[2.239]**|**[1.844]**|[0.875]|[0.801]|**[1.696]**|**[2.441]**|**[-2.006]**|**[-2.249]**|[1.550]|**[-3.161]**|**[-3.128]**|**[-2.526]**|**[-3.064]**|[0.056]|[-0.422]|\\n|Hazard<sup>2</sup>||0.002|0.004||**0.029****|0.005||-0.022|**-0.039****||**-0.033****|**-0.039****||-0.000|0.000|\\n|||[0.101]|[0.181]||**[2.189]**|[0.347]||[-1.305]|**[-2.213]**||**[-2.114]**|**[-2.571]**||[-0.444]|[0.374]|\\n|Hazard<sup>3</sup>|||0.007|||**-0.023****|||**-0.041*****|||0.012|||-0.000|\\n||||[0.399]|||**[-2.352]**|||**[-2.917]**|||[1.215]|||[-0.406]|\\n|Obs.|50,266|50,266|50,266|50,242|50,242|50,242|50,266|50,266|50,266|50,266|50,266|50,266|48,506|48,506|48,506|\\n|R-sq|0.362|0.363|0.364|0.358|0.358|0.359|0.361|0.361|0.361|0.361|0.362|0.362|0.370|0.371|0.371|\\n\\n_Source:_ World Bank.\\n\\n_Note:_ NDVI = Normalized Difference Vegetation Index; SPEI = Standardized Precipitation Evapotranspiration Index; WRSI = Water Requirement Satisfaction Index. Results are from a weighted regression such that each household is assigned its survey-specific population weight. Countries and years included are Ethiopia (2011, 2016), Lesotho (2017), Malawi (2010, 2016, 2019), Mozambique (2014), Zambia (2015), and Zimbabwe (2017). Dependent variable is the log of per capita (or adult equivalent) consumption. Other variables included and not shown are education of household head, gender of household head, whether the household location is remote, agroecological zone indicators, whether the household was surveyed during the lean season, survey year, country, country x survey year, and administrative fixed effects for district (Ethiopia. Lesotho, Malawi) and province (Mozambique, Zambia, Zimbabwe). Robust t-statistics are shown in parenthesis. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent. When variable is significant the coefficient and standard error are in bold font.\\n\\n18\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"**Table 4.2. Results for highland livelihood zone**\\n\\n|||**Soil moistur**|**e**||**NDVI**|||**SPEI**||**Tot**|**al seasonal ra**|**infall**||**WRSI**||\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|(9)|(10)|(11)|(12)|(13)|(14)|(15)|\\n|**Landed h**|**ouseholds**|||||||||||||||\\n|Hazard|**0.045*****|**0.044*****|**0.044***|**0.020***|0.017|0.025|**0.041****|0.001|**-0.049***|**0.030****|0.017|0.031|**0.003***|0.017|0.065|\\n||**[4.495]**|**[4.350]**|**[1.653]**|**[1.649]**|[1.154]|[1.382]|**[2.426]**|[0.060]|**[-1.649]**|**[2.575]**|[1.172]|[1.307]|**[1.782]**|[1.332]|[0.902]|\\n|Hazard<sup>2</sup>||-0.004|-0.004||-0.003|-0.013||**-0.055*****|**-0.041****||**-0.017***|**-0.021***||-0.000|-0.001|\\n|||[-0.427]|[-0.333]||[-0.338]|[-0.917]||**[-3.940]**|**[-2.496]**||**[-1.895]**|**[-1.813]**||[-1.147]|[-0.783]|\\n|Hazard<sup>3</sup>|||-0.000|||-0.005|||**0.024****|||-0.005|||0.000|\\n||||[-0.007]|||[-0.853]|||**[2.153]**|||[-0.694]|||[0.699]|\\n|**Landless**|**households**|||||||||||||||\\n|Hazard|-0.005|-0.006|-0.030|-0.002|-0.027|-0.026|0.020|-0.011|0.031|0.017|0.016|0.042|0.002|0.021|0.111|\\n||[-0.214]|[-0.231]|[-0.533]|[-0.063]|[-0.565]|[-0.556]|[0.765]|[-0.294]|[0.600]|[0.747]|[0.622]|[0.876]|[1.046]|[1.368]|[1.577]|\\n|Hazard<sup>2</sup>||-0.002|-0.000||-0.023|-0.024||-0.031|**-0.111****||0.001|-0.009||-0.000|-0.001|\\n|||[-0.100]|[-0.002]||[-0.852]|[-0.496]||[-1.080]|**[-2.233]**||[0.062]|[-0.358]||[-1.303]|[-1.425]|\\n|Hazard<sup>3</sup>|||0.010|||-0.001|||**-0.051***|||-0.009|||0.000|\\n||||[0.523]|||[-0.045]|||**[-1.746]**|||[-0.625]|||[1.308]|\\n|Obs.|15,046|15,046|15,046|15,046|15,046|15,046|15,046|15,046|15,046|15,046|15,046|15,046|12,159|12,159|12,159|\\n|R-sq|0.265|0.265|0.265|0.260|0.260|0.261|0.261|0.266|0.268|0.262|0.263|0.263|0.258|0.259|0.259|\\n\\n_Source:_ World Bank.\\n\\n_Note:_ NDVI = Normalized Difference Vegetation Index; SPEI = Standardized Precipitation Evapotranspiration Index; WRSI = Water Requirement Satisfaction Index. Results are from a weighted regression such that each household is assigned its survey-specific population weight. Countries and years included are Ethiopia (2011, 2016), Lesotho (2017), Mozambique (2014), and Zimbabwe (2017). Dependent variable is the log of per capita (or adult equivalent) consumption. Other variables included and not shown are education of household head, gender of household head, whether the household location is remote, whether the household was surveyed during the lean season, survey year, country, country x survey year, and administrative fixed effects for district (Ethiopia. Lesotho) and province (Mozambique, Zimbabwe). Robust t-statistics are shown in parenthesis. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent. When variable is significant the coefficient and standard error are in bold font.\\n\\n19\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"**Table 4.3. Results for roots and lowland livelihood zone**\\n\\n|||**Soil moistur**|**e**||**NDVI**|||**SPEI**||**Tot**|**al seasonal rai**|**nfall**||**WRSI**||\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|(9)|(10)|(11)|(12)|(13)|(14)|(15)|\\n|**Landed h**|**ouseholds**|||||||||||||||\\n|Hazard|**0.152*****|**0.092*****|**0.136*****|**0.069*****|**0.092*****|**0.091*****|**0.082*****|**0.063***|0.027|**0.069*****|**0.036***|0.022|0.019|**-1.337*****|-3.865|\\n||**[8.123]**|**[3.668]**|**[4.314]**|**[4.269]**|**[4.214]**|**[3.038]**|**[2.634]**|**[1.785]**|[0.527]|**[3.406]**|**[1.766]**|[0.744]|[1.360]|**[-2.927]**|[-0.237]|\\n|Hazard<sup>2</sup>||**-0.07*****|**-0.149*****||0.022|0.023||-0.017|0.013||**-0.041*****|**-0.036*****||**0.007*****|0.033|\\n|||**[-2.801]**|**[-4.316]**||[1.496]|[1.419]||[-0.715]|[0.379]||**[-3.894]**|**[-3.504]**||**[2.967]**|[0.196]|\\n|Hazard<sup>3</sup>|||**-0.051*****|||0.001|||0.023|||0.005|||-0.000|\\n||||**[-2.668]**|||[0.092]|||[0.999]|||[0.839]|||[-0.154]|\\n|**Landless**|**households**|||||||||||||||\\n|Hazard|**0.166*****|**0.154*****|**0.165*****|0.036|0.013|-0.001|0.027|0.112|0.022|0.021|0.015|-0.010|**0.050****|**-2.006****|11.437|\\n||**[3.698]**|**[3.391]**|**[2.970]**|[0.971]|[0.236]|[-0.016]|[0.442]|[1.254]|[0.209]|[0.543]|[0.429]|[-0.148]|**[2.281]**|**[-2.266]**|[0.441]|\\n|Hazard<sup>2</sup>||-0.032|-0.037||-0.018|0.005||0.069|**0.167***||-0.008|-0.010||**0.011****|-0.126|\\n|||[-0.579]|[-0.571]||[-0.486]|[0.097]||[1.264]|**[1.677]**||[-0.323]|[-0.370]||**[2.335]**|[-0.476]|\\n|Hazard<sup>3</sup>|||-0.007|||0.014|||0.077|||0.010|||0.000|\\n||||[-0.157]|||[0.466]|||[1.373]|||[0.600]|||[0.514]|\\n|Obs.|5,705|5,705|5,705|5,705|5,705|5,705|5,705|5,705|5,705|5,705|5,705|5,705|2,206|2,206|2,206|\\n|R-sq|0.175|0.179|0.182|0.157|0.158|0.158|0.154|0.155|0.157|0.156|0.160|0.160|0.107|0.115|0.115|\\n\\n_Source:_ World Bank.\\n\\n_Note:_ NDVI = Normalized Difference Vegetation Index; SPEI = Standardized Precipitation Evapotranspiration Index; WRSI = Water Requirement Satisfaction Index. Results are from a weighted regression such that each household is assigned its survey-specific population weight. One country, Nigeria, is included for years 2011, 2013, and 2016. Dependent variable is the log of per capita (or adult equivalent) consumption. Other variables included and not shown are education of household head, gender of household head, whether the household location is remote, whether the household was surveyed during the lean season, survey year, country, country x survey year, and zone fixed effects. Robust t-statistics are shown in parenthesis. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent. When variable is significant the coefficient and standard error are in bold font.\\n\\n20\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"<!-- Start of picture text --> Landed households, maize<br>° | |<br>§e '| !3<br>F48 |\\\\\\\\ |\\\\\\\\<br>—_ 4<br>2<br>so |<br>Bo: a |<br>i}<br>I!<br>-2 0 2<br>Soil moisture anomaly<br>— Linear -——— Quadratic ——— Cubic<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> Landed households, highland<br>= 1\\\\ '\\\\<br>c | '<br>a0ES |<br>FAc I! 1i}<br>8. | |<br>£<br>S 1 |'<br>Sw3g :1 :1<br>Sr | |<br>z<br>8. \\\\<br>2° 1 '<br>-2 0 2<br>Soil moisture anomaly<br>——— Linear ———— Quadratic ——— Cubic<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> Landed households, roots & lowland<br>re) |''\\\\<br>|<br>c ' '<br>2 '<br>ee<br>Fa |<br>8 | \\\\<br>bdry }<br>2 ° i |<br>o ! \\\\<br>S |i} i|<br>= \\\\ \\\\<br>3°<br>& 1 '<br>re) | \\\\\\\\<br>7<br>-2 0 2<br>Soil moisture anomaly<br>——— Linear ———— Quadratic ——— Cubic<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> Landed households, maize<br>sf | |<br>5 ! !<br>es=2 !;H !<br>8 \\\\<br>2<br>o |<br>3: ;<br>= ' |<br>' \\\\<br>a 1 \\\\<br>'<br>-2 NDVI anomaly0 2<br>——— Linear ———— Quadratic ———— Cubic<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> Landed households, highland<br>82 ;<br>| |<br>8 | '<br>E> |<br>Zo<br>8 | 1<br>e© \\\\ \\\\;<br>$=)= | '<br>S., | \\\\<br>se<br>§ 1 !<br>2 | |<br>e<br>= | |<br>:<br>-2 Novi anomaly0 2<br>——— Linear ———— Quadratic ——— Cubic<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> Landed households, roots & lowland<br>+ | '<br>c 1 1<br>2 ' '<br>a f i}<br>E | \\\\<br>Bn | |<br>88 }\\\\ |'<br>£ | |<br>2 |<br>©ad i} i}<br>Bo<br>=<br>8 | |<br>a \\\\ 1\\\\<br>x \\\\ \\\\<br>'<br>-2 0 2<br>NDVI anomaly<br>——— Linear ———— Quadratic ———— Cubic<br><!-- End of picture text -->\", \"datasets\": [{\"mention_name\": {\"text\": \"Soil moisture anomaly\", \"confidence\": 0.31943926215171814, \"start\": 149, \"end\": 170, \"clean_text\": \"Soil moisture anomaly\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 149, \"end\": 170, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 149, \"end\": 170, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 149, \"end\": 170}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 149, \"end\": 170, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.38648682832717896, \"start\": 149, \"end\": 170}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.30676209926605225, \"start\": 149, \"end\": 170}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.48517826199531555, \"start\": 149, \"end\": 170}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.48517826199531555, \"start\": 149, \"end\": 170}}, {\"mention_name\": {\"text\": \"NDVI anomaly\", \"confidence\": 0.3898758590221405, \"start\": 1573, \"end\": 1585, \"clean_text\": \"NDVI anomaly\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1573, \"end\": 1585, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1573, \"end\": 1585, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1573, \"end\": 1585}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1573, \"end\": 1585, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.5208790302276611, \"start\": 344, \"end\": 356}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3898758590221405, \"start\": 344, \"end\": 356}, \"usage_context\": {\"text\": \"background\", \"confidence\": 0.4975389242172241, \"start\": 344, \"end\": 356}, \"is_used\": {\"text\": \"False\", \"confidence\": 0.4975389242172241, \"start\": 344, \"end\": 356}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"<!-- Start of picture text --> Landed households, maize<br>|<br>§e '\\\\ \\\\\\\\<br>5 | |<br>8 'i |<br>& |<br>3°<br>5 !<br>I! \\\\<br>8°= 1| 1\\\\<br>cd | '<br>‘<br>-2 0 2<br>SPEI anomaly<br>—— Linear ———— Quadratic ————— Cubic<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> Landed households, highland<br>nN \\\\' \\\\!<br>§<br>Bo<br>E ! |<br>8 | i<br>c \\\\ '<br>=“o,;a< || ]'f<br>z<br>5g' v '| '|<br>a '|!|<br>o |?<br>‘<br>-2 0 2<br>SPEI anomaly<br>——— Linear -———— Quadratic ———— Cubic<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> Landed households, roots & lowland<br>\\\\<br>\\\\ |<br>|<br>Sy \\\\ \\\\|<br>a” I I<br>—> \\\\ H<br>ac | !<br>8% |\\\\ I\\\\<br>£3 : :<br>c<br>go \\\\<br>Boyo a ' '<br>'<br>-2 0 2<br>SPEI anomaly<br>—— Linear ———— Quadratic ————— Cubic<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> Landed households, maize<br>' '<br>2 }<br>2 H i<br>8 1 |<br>=. | '<br>2<br>: |<br>Co }<br>& \\\\ \\\\<br>'<br>-2 0 2<br>Total rainfall anomaly<br>—— Linear -———— Quadratic — Cubic<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> Landed households, highland<br>“4' \\\\'<br>So] |<br>e | |<br>=] ! !<br>So | \\\\<br>= \\\\<br>=©Sewso |H' 'i!<br>g |<br>8a = If I'<br>a 1 1<br>2, {7 !<br>-2 0 2<br>Total rainfall anomaly<br>——— Linear -———— Quadratic —— Cubic<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> Landed households, roots & lowland<br>\\\\ \\\\<br>c<br>2 | |<br>2 |<br>E \\\\ I<br>ge<br>8 |! '<br>£©2 |\\\\ \\\\\\\\<br>o | |<br>BN\\\\\\\\<br>a5 I |<br>’<br>-2 0 2<br>Total rainfall anomaly<br>—— Linear -———— Quadratic — Cubic<br><!-- End of picture text -->\", \"datasets\": [{\"mention_name\": {\"text\": \"SPEI anomaly\", \"confidence\": 0.3082347810268402, \"start\": 708, \"end\": 720, \"clean_text\": \"SPEI anomaly\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 708, \"end\": 720, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 708, \"end\": 720, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 708, \"end\": 720}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 708, \"end\": 720, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.45880255103111267, \"start\": 100, \"end\": 112}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3082347810268402, \"start\": 100, \"end\": 112}, \"usage_context\": {\"text\": \"background\", \"confidence\": 0.3499206304550171, \"start\": 100, \"end\": 112}, \"is_used\": {\"text\": \"False\", \"confidence\": 0.3499206304550171, \"start\": 100, \"end\": 112}}, {\"mention_name\": {\"text\": \"Total rainfall anomaly\", \"confidence\": 0.30235880613327026, \"start\": 939, \"end\": 961, \"clean_text\": \"Total rainfall anomaly\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 939, \"end\": 961, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 939, \"end\": 961, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 939, \"end\": 961}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 939, \"end\": 961, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.46488797664642334, \"start\": 42, \"end\": 64}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.30235880613327026, \"start\": 42, \"end\": 64}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5205233097076416, \"start\": 42, \"end\": 64}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5205233097076416, \"start\": 42, \"end\": 64}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"Tables 4.4 and 4.5 show results for pastoral and agropastoral areas respectively. The choice of hazard and functional form is influenced by in-depth work undertaken on Mauritania (Hill and Vinha 2022), one of the pastoral countries in the sample. This work highlights the importance of using Tropical Livestock Unit (TLU) ownership as a way of distinguishing between household types. It also highlights the importance of including households living in areas classified as urban, as many of these households were found to have significant TLU ownership. Also, given that especially many of the pastorals are in locations without growing season per se, the hazards are constructed over the whole rainy season. The results for pastoral areas show that poor weather impacts consumption much more strongly for households with fewer animals. Those with more animals are able to smooth their consumption in the face of poor weather. The results for agropastoralists are harder to interpret. The same pattern of worse impacts for households with fewer animals holds for agropastoralists, but weather shocks appear positive for landless households. Only results for NDVI are shown, but no measure provided sensible results. Further work, and probably additional data, are needed to better model and understand the impact of weather shocks on agropastoralist welfare in Sub-Saharan Africa, including considering cumulative impacts of rainfall losses across seasons and temperature impacts on cattle.\\n\\n**Table 4.4. Results for pastoral zone in Western Africa**\\n\\n|||**NDVI**<br>||**S**<br>|**oil moisture**<br>||**Tota**<br>|**l seasonal rai**<br>|**nfall**<br>|\\n|---|---|---|---|---|---|---|---|---|---|\\n||**(1)**|**(2)**|**(3)**|**(4)**|**(5)**|**(6)**|**(7)**|**(8)**|**(9)**|\\n|**No TLU**||||||||||\\n|Hazard|**0.069****|**0.065*****|**0.130*****|**0.120*****|0.050|0.014|0.014|0.019|0.111|\\n||**[2.463]**|**[2.966]**|**[3.510]**|**[2.809]**|[0.929]|[0.157]|[0.518]|[0.667]|[1.632]|\\n|Hazard<sup>2</sup>||0.004|**0.056***||0.079|0.022||-0.008|0.025|\\n|||[0.126]|**[1.832]**||[1.304]|[0.313]||[-0.227]|[0.662]|\\n|Hazard<sup>3</sup>|||**-0.037*****|||0.059|||**-0.049***|\\n||||**[-2.782]**|||[0.681]|||**[-1.659]**|\\n|**Few(less**|**than 2.1) TL**|**U**||||||||\\n|Hazard|**0.087*****|**0.065*****|**0.120*****|**0.110*****|**0.072****|**0.060**|**0.073****|**0.078****|**0.156****|\\n||**[2.899]**|**[3.656]**|**[4.102]**|**[2.841]**|**[2.378]**|**[1.233]**|**[2.432]**|**[2.579]**|**[2.552]**|\\n|Hazard<sup>2</sup>||0.024|**0.062*****||0.047|0.030||-0.013|0.005|\\n|||[1.049]|**[2.891]**||[1.185]|[0.506]||[-0.658]|[0.235]|\\n|Hazard<sup>3</sup>|||**-0.028*****|||0.016|||**-0.040***|\\n||||**[-3.379]**|||[0.375]|||**[-1.914]**|\\n|**Many (m**|**ore than 2.1)**|**TLU**||||||||\\n|Hazard|0.020|0.022|-0.011|0.016|**0.087****|0.078|0.001|0.009|0.028|\\n||[1.008]|[1.074]|[-0.382]|[0.433]|**[2.355]**|[1.592]|[0.047]|[0.330]|[0.520]|\\n|Hazard<sup>2</sup>||-0.009|-0.019||**-0.076***|-0.090||-0.027|-0.021|\\n|||[-0.954]|[-1.352]||**[-1.895]**|[-1.538]||[-1.522]|[-1.153]|\\n|Hazard<sup>3</sup>|||**0.013***|||0.013|||-0.005|\\n||||**[1.867]**|||[0.264]|||[-0.284]|\\n|Obs.|13,641|13,641|13,641|13,641|13,641|13,641|13,641|13,641|13,641|\\n|R-sq|0.417|0.418|0.422|0.415|0.416|0.416|0.415|0.415|0.416|\\n\\n_Source:_ World Bank.\\n\\n_Note:_ NDVI = Normalized Difference Vegetation Index; TLU = Tropical Livestock Unit. Results are from a weighted regression such that each household is assigned its survey-specific population weight. Countries and years included are Mauritania (2014, 2109/20) and Niger (2018). Dependent variable is the log of per capita (or adult equivalent) consumption. Other variables included and not shown are education of household head, gender of household head, whether the household location is remote, whether the household was surveyed during the lean season, survey year, country, country x survey year, and administrative fixed effects. Robust t-statistics are shown in parenthesis. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent.\\n\\n23\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"**Table 4.5. Results for agropastoral zone in Western Africa**\\n\\n|||**NDVI**||\\n|---|---|---|---|\\n||**(1)**|**(2)**|**(3)**|\\n|**Landed ho**|**useholds**|||\\n|**No TLU**||||\\n|Hazard|-0.046|-0.047|-0.031|\\n||[-0.817]|[-0.843]|[-0.329]|\\n|Hazard<sup>2</sup>||-0.002|-0.001|\\n|||[-0.068]|[-0.025]|\\n|Hazard<sup>3</sup>|||-0.009|\\n||||[-0.224]|\\n|**Few(less t**|**han 2.1) TLU**|||\\n|Hazard|**0.052***|**0.052***|0.064|\\n||**[1.658]**|**[1.693]**|[1.240]|\\n|Hazard<sup>2</sup>||0.022|0.022|\\n|||[1.034]|[0.970]|\\n|Hazard<sup>3</sup>|||-0.006|\\n||||[-0.292]|\\n|**Many (mor**|**e than 2.1) TLU**|||\\n|Hazard|0.039|0.039|-0.010|\\n||[0.908]|[0.906]|[-0.192]|\\n|Hazard<sup>2</sup>||-0.010|-0.008|\\n|||[-0.235]|[-0.215]|\\n|Hazard<sup>3</sup>|||0.032|\\n||||[1.044]|\\n|**Landless ho**|**useholds**|||\\n|Hazard|**-0.212*****|**-0.214*****|0.112|\\n||**[-3.393]**|**[-3.553]**|[0.453]|\\n|Hazard<sup>2</sup>||0.018|0.023|\\n|||[0.265]|[0.352]|\\n|Hazard<sup>3</sup>|||-0.186|\\n||||[-1.391]|\\n|Obs.|4,593|4,593|4,593|\\n|R-sq|0.132|0.133|0.137|\\n\\n_Source:_ World Bank.\\n\\n_Note:_ NDVI = Normalized Difference Vegetation Index. Results are from a weighted regression such that each household is assigned its survey-specific population weight. Countries and years included are Niger (2018) and Nigeria (2011, 2013, 2016). Dependent variable is the log of per capita (or adult equivalent) consumption. Other variables included and not shown are education of household head, gender of household head, whether the household location is remote, whether the household was surveyed during the lean season, survey year, country, country x survey year, and administrative fixed effects. Robust t-statistics are shown in parenthesis. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent.\\n\\n# 4.2. Exploring different areas over which to aggregate the hazard data\\n\\nInitial analysis suggested that a 20 km level of spatial aggregation of hazards was appropriate. This radius is able to reflect the local shock to a household’s crop yields and may be large enough to capture some localized impacts on prices and wages in the local market area. We tested this level of aggregation again at the end of the analysis for all countries and found that the level of spatial aggregation made very little difference. In some specifications taking a 24\", \"datasets\": [{\"mention_name\": {\"text\": \"Obs.\", \"confidence\": 0.32398778200149536, \"start\": 944, \"end\": 948, \"clean_text\": \"Obs.\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 944, \"end\": 948, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 944, \"end\": 948, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 944, \"end\": 948}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 944, \"end\": 948, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.6581107974052429, \"start\": 944, \"end\": 948}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.32398778200149536, \"start\": 944, \"end\": 948}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6513825058937073, \"start\": 944, \"end\": 948}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6513825058937073, \"start\": 944, \"end\": 948}}, {\"mention_name\": {\"text\": \"hazard data\", \"confidence\": 0.3456640839576721, \"start\": 1799, \"end\": 1810, \"clean_text\": \"hazard data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1799, \"end\": 1810, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1799, \"end\": 1810, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1799, \"end\": 1810}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1799, \"end\": 1810, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6883273720741272, \"start\": 258, \"end\": 269}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3456640839576721, \"start\": 258, \"end\": 269}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4896898865699768, \"start\": 258, \"end\": 269}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4896898865699768, \"start\": 258, \"end\": 269}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"larger level of spatial aggregation increases the estimated impact of the shock—although not significantly—without affecting statistical significance (table 4.6). In future work it would be interesting to explore much larger areas of aggregation if this is possible (i.e., if enough variation in the hazard is left and GPS data are available) to see if this affects the results.\\n\\n**Table 4.6. Results for different areas of spatial aggregation of hazard data**\\n\\n||**Maize(GPS**<br>|**coordinate c**<br>|**ountries only)**<br>|**2**<br>|**0 km radius**<br>||<br>|**50 km radius**<br>||\\n|---|---|---|---|---|---|---|---|---|---|\\n||**10km**|**20km**|**50km**|**Maize**|**Highland**|**Roots**|**Maize**|**Highland**|**Roots**|\\n|**Landed h**|**ouseholds**|||||||||\\n|Hazard|0.066***|0.075***|0.086***|0.055***|0.045***|0.152***|0.056***|0.045***|0.154***|\\n||[4.252]|[4.608]|[4.765]|[3.852]|[4.495]|[8.123]|[3.722]|[4.495]|[8.487]|\\n|**Landless**|**households**|||||||||\\n|Hazard|0.109***|0.112***|0.123***|0.043**|-0.005|0.166***|0.045**|-0.004|0.159***|\\n||[4.165]|[4.212]|[4.451]|[2.239]|[-0.214]|[3.698]|[2.276]|[-0.176]|[4.017]|\\n|Obs.|34,824|34,739|35,212|50,266|15,046|5,705|50,739|15,060|5,856|\\n|R-sq|0.138|0.137|0.133|0.362|0.265|0.175|0.357|0.260|0.177|\\n\\n_Source:_ World Bank.\\n\\n_Note:_ Results are from a weighted regression such that each household is assigned its survey-specific population weight. Dependent variable is the log of per capita (or adult equivalent) consumption. The hazard variable used is the soil moisture index. Other variables included and not shown are education of household head, gender of household head, whether the household location is remote, agroecological zone indicators, whether the household was surveyed during the lean season, survey year, country, country x survey year, and administrative fixed effects for district. Robust t-statistics are shown in parenthesis. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent.\\n\\n# 4.3. Exploring food and nonfood consumption impacts separately\\n\\nFour of the nine countries (Ethiopia, Malawi, Mozambique, and Niger) had separate aggregates of food and nonfood consumption in the data sets available for analysis. We conducted the same analysis separately on food and nonfood consumption for these countries to assess whether the main impact was on food consumption or whether households protected food consumption by reducing nonfood consumption; the latter result is found, for example, in Kochhar and Knippenberg (2023) using a similar method. Tables 4.7 and 4.8 present results for the maize and highland livelihood zones respectively. Results are shown using just the soil moisture variable. For both livelihood zones, the impacts of the hazard variables on food consumption are larger than impacts on nonfood consumption. Nonfood consumption is impacted only among landed households. Further work with data from additional countries would be informative in understanding whether this pattern holds.\\n\\n25\", \"datasets\": [{\"mention_name\": {\"text\": \"GPS data\", \"confidence\": 0.8886116743087769, \"start\": 319, \"end\": 327, \"clean_text\": \"GPS data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 319, \"end\": 327, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 319, \"end\": 327, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 319, \"end\": 327}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 319, \"end\": 327, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.9919881820678711, \"start\": 319, \"end\": 327}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8886116743087769, \"start\": 319, \"end\": 327}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5908600091934204, \"start\": 319, \"end\": 327}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5908600091934204, \"start\": 319, \"end\": 327}}, {\"mention_name\": {\"text\": \"data from additional countries\", \"confidence\": 0.6127178072929382, \"start\": 2897, \"end\": 2927, \"clean_text\": \"data from additional countries\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2897, \"end\": 2927, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2897, \"end\": 2927, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2897, \"end\": 2927}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2897, \"end\": 2927, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.4902617633342743, \"start\": 282, \"end\": 312}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6127178072929382, \"start\": 282, \"end\": 312}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.48162904381752014, \"start\": 282, \"end\": 312}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.48162904381752014, \"start\": 282, \"end\": 312}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"**Table 4.7. Food and nonfood consumption, maize livelihood zone (soil moisture)**\\n\\n||Total consu|mption, all c<br>zone|ountries in|Total cons<br>disagg|umption, co<br>regated info|untries with<br>rmation|Fo|od consumpti|on|Non|food consum|ption|\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n|Landed h|ouseholds||||||||||||\\n|Hazard|**0.055*****|**0.059*****|**0.119*****|**0.058*****|**0.060*****|**0.131*****|**0.081*****|**0.082*****|**0.158*****|**0.037***|**0.041****|**0.101*****|\\n||**[3.852]**|**[4.097]**|**[4.792]**|**[3.731]**|**[3.900]**|**[4.681]**|**[4.754]**|**[4.680]**|**[5.581]**|**[1.903]**|**[2.168]**|**[2.751]**|\\n|Hazard<sup>2</sup>||**0.022***|0.011||**0.024***|0.012||0.011|-0.002||**0.030***|0.019|\\n|||**[1.831]**|[0.839]||**[1.753]**|[0.809]||[0.729]|[-0.152]||**[1.758]**|[1.009]|\\n|Hazard<sup>3</sup>|||**-0.028*****|||**-0.032*****|||**-0.035*****|||**-0.027***|\\n||||**[-2.954]**|||**[-3.041]**|||**[-3.740]**|||**[-1.902]**|\\n|Landless|households||||||||||||\\n|Hazard|**0.043****|**0.038***|0.032|**0.060*****|**0.053***|0.063|**0.093*****|**0.100*****|**0.112****|0.028|0.009|0.029|\\n||**[2.239]**|**[1.844]**|[0.875]|**[2.593]**|**[1.910]**|[1.161]|**[3.942]**|**[3.610]**|**[2.121]**|[0.950]|[0.255]|[0.407]|\\n|Hazard<sup>2</sup>||0.002|0.004||0.005|0.002||0.022|0.018||-0.013|-0.019|\\n|||[0.101]|[0.181]||[0.195]|[0.071]||[0.739]|[0.546]||[-0.384]|[-0.515]|\\n|Hazard<sup>3</sup>|||0.007|||-0.002|||-0.003|||-0.008|\\n||||[0.399]|||[-0.067]|||[-0.125]|||[-0.269]|\\n|Obs.|50,266|50,266|50,266|31,671|31,671|31,671|31,671|31,671|31,671|31,671|31,671|31,671|\\n|R<sup>2</sup>|0.362|0.363|0.364|0.363|0.363|0.365|0.281|0.281|0.284|0.387|0.388|0.389|\\n\\n_Source:_ World Bank.\\n\\n_Note:_ Results are from a weighted regression such that each household is assigned its survey-specific population weight. Countries and years included are Ethiopia (2011, 2016), Malawi (2010, 2016, 2019), and Mozambique (2014). Dependent variable is the log of per capita (or adult equivalent) consumption. Other variables included and not shown are education of household head, gender of household head, whether the household location is remote, agroecological zone indicators, whether the household was surveyed during the lean season, survey year, country, country x survey year, and administrative fixed effects for district (Ethiopia. Lesotho, Malawi) and province (Mozambique, Zambia, Zimbabwe). Robust t-statistics are shown in parenthesis. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent.\\n\\n26\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [27]}}, {\"page\": 28, \"chunk\": 28, \"input_text\": \"**Table 4.8. Food and nonfood consumption, highland livelihood zone (soil moisture)**\\n\\n||Total consu|mption, all co<br>zone|untries in|Total consu<br>disaggr|mption, coun<br>egated inform|tries with<br>ation|Foo|d consumptio|n|Non-|food consum|ption|\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n|Landed h|ouseholds||||||||||||\\n|Hazard|**0.045*****|**0.044*****|**0.044***|**0.044*****|**0.043*****|0.044|**0.054*****|**0.052*****|0.037|**0.042*****|**0.043*****|**0.060***|\\n||**[4.495]**|**[4.350]**|**[1.653]**|**[4.402]**|**[4.227]**|[1.600]|**[4.547]**|**[4.168]**|[1.195]|**[3.365]**|**[3.413]**|**[1.720]**|\\n|Hazard<sup>2</sup>||-0.004|-0.004||-0.005|-0.005||-0.010|-0.005||0.004|-0.001|\\n|||[-0.427]|[-0.333]||[-0.515]|[-0.422]||[-1.038]|[-0.420]||[0.395]|[-0.086]|\\n|Hazard<sup>3</sup>|||-0.000|||-0.000|||0.006|||-0.007|\\n||||[-0.007]|||[-0.040]|||[0.520]|||[-0.558]|\\n|Landless|households||||||||||||\\n|Hazard|-0.005|-0.006|-0.030|-0.012|-0.012|-0.053|0.013|0.012|-0.080|-0.026|-0.025|-0.016|\\n||[-0.214]|[-0.231]|[-0.533]|[-0.441]|[-0.450]|[-0.740]|[0.454]|[0.408]|[-1.218]|[-0.718]|[-0.685]|[-0.153]|\\n|Hazard<sup>2</sup>||-0.002|-0.000||-0.013|-0.009||-0.014|-0.002||-0.008|-0.012|\\n|||[-0.100]|[-0.002]||[-0.578]|[-0.384]||[-0.530]|[-0.091]||[-0.275]|[-0.365]|\\n|Hazard<sup>3</sup>|||0.010|||0.015|||0.035|||-0.004|\\n||||[0.523]|||[0.696]|||[1.624]|||[-0.117]|\\n|Obs.|15,046|15,046|15,046|10,877|10,877|10,877|10,877|10,877|10,877|10,868|10,868|10,868|\\n|R<sup>2</sup>|0.265|0.265|0.265|0.268|0.268|0.268|0.333|0.334|0.334|0.228|0.228|0.228|\\n\\n_Source:_ World Bank.\\n\\n_Note:_ Results are from a weighted regression such that each household is assigned its survey-specific population weight. Countries and years included are Ethiopia (2011, 2016) and Mozambique (2014). Dependent variable is the log of per capita (or adult equivalent) consumption. Other variables included and not shown are education of household head, gender of household head, whether the household location is remote, whether the household was surveyed during the lean season, survey year, country, country x survey year, and administrative fixed effects for district (Ethiopia, Lesotho) and province (Mozambique, Zimbabwe). Robust t-statistics are shown in parenthesis. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent.\\n\\n27\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [28]}}, {\"page\": 29, \"chunk\": 29, \"input_text\": \"=B +BX R (R Zz =W (R ((R ~~DD|~~\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [29]}}, {\"page\": 30, \"chunk\": 30, \"input_text\": \"|Maize(GS)|32%|29%|23%|36%|\\n|---|---|---|---|---|\\n|Highland(GS)|23%|27%|33%|29%|\\n|Roots and lowland(GS)|9%|7%|42%|23%|\\n|Agropastoral(RS)|35%|28%||28%|\\n|Pastoral(RS)|40%|36%||27%|\\n\\n_Source:_ World Bank.\\n\\n_Note:_ NDVI = Normalized Difference Vegetation Index; SPEI = Standardized Precipitation Evapotranspiration Index; GS = growing season; RS = rainy season.\\n\\nTo assess whether the extreme values of the hazards are within or outside the range of hazards experienced by households within the districts and years used, we assessed the percentage of households in a given livelihood zone that experienced a hazard value more than two standard deviations lower than the mean or more than three standard deviations lower than the mean. The results for soil moisture for maize, highland and roots and lowland livelihood zones and for NDVI for agropastorals and pastorals are shown in table 5.2. Assessing the common support between the range of possible hazards and those used in estimating coefficients 2 and 3 can help address concerns regarding the plausibility of the welfare impact predictions that use the fixed effects estimates. In sum, while moderate deviations in rainfall are well reflected in the data used to estimate 2 and 3 , more extreme deviations are less well reflected in the ββ̂ ββ̂ data. ββ̂ ββ̂\\n\\n**Table 5.2. Share of households with an extreme hazard outcome for soil moisture**\\n\\n|**Livelihood zone**|**Hazard z-score < -3**|**Hazard z-score < -2**|**Negative hazard z-score**|\\n|---|---|---|---|\\n|Soil moisture (GS)||||\\n|Maize|0|0.01|0.63|\\n|Highland|0|0.05|0.64|\\n|Roots and lowland|0|0.03|0.63|\\n|NDVI (RS)||||\\n|Agropastoral|0|0.00|0.40|\\n|Pastoral|0|0.01|0.38|\\n\\n_Source:_ World Bank. _Note:_ GS = growing season; RS = rainy season. We generated the alternate draws of the hazard distribution, k , a in two ways. The first used the historical weather distribution to select k for all households. This provided a first assessment of what an EP curve would look like but is based on a limited set of historical data points—in the case of soil moisture, only 13 years of data. The second way RR used the data set of 10,000 possible weather simulations for Malawi developed by the risk modeling firm AIR, which RR was based on the historical pattern of rainfall in Malawi. This catalog is available only for Malawi but suggests the degree of precision that modeling the weather offers as compared to using historical weather data. The generated k are used with parameter estimates from tables 4.1 to 4.5 to simulate what consumption would be for each household under these weather conditions. Once consumption has been simulated for a given draw of the weather, variables that describe the simulated consumption distribution can be generated. We focus particularly on: (i) RR the poverty headcount rate—the share of households whose simulated consumption falls below the extreme poverty line of $1.90—and (ii) the total poverty gap—the difference between consumption and the $1.90 poverty line added up for all households whose consumption falls below the poverty line.\\n\\nThe increase in the poverty rate and total poverty gap for eight of the nine countries included in this analysis are shown in figures 5.1 and 5.2. Niger was not included given the low percentage of the population in livelihood zones for which there are robust results. Each country comprises more than one livelihood zone, so each country draws on more than one set of parameter estimates. For the simulation results, the linear relationship is used with the soil moisture hazard for countries with growing season. For Mauritania we use the rainy season NDVI. The linear model was chosen 29\", \"datasets\": [{\"mention_name\": {\"text\": \"Standardized Precipitation Evapotranspiration Index\", \"confidence\": 0.35062500834465027, \"start\": 265, \"end\": 316, \"clean_text\": \"Standardized Precipitation Evapotranspiration Index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 265, \"end\": 316, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 265, \"end\": 316, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 265, \"end\": 316}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 265, \"end\": 316, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.8542601466178894, \"start\": 265, \"end\": 316}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.35062500834465027, \"start\": 265, \"end\": 316}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.591593861579895, \"start\": 265, \"end\": 316}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.591593861579895, \"start\": 265, \"end\": 316}}, {\"mention_name\": {\"text\": \"ββ̂ ββ̂ data\", \"confidence\": 0.513024091720581, \"start\": 1290, \"end\": 1302, \"clean_text\": \"ββ̂ ββ̂ data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1290, \"end\": 1302, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1290, \"end\": 1302, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1290, \"end\": 1302}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1290, \"end\": 1302, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9407384395599365, \"start\": 887, \"end\": 899}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.513024091720581, \"start\": 887, \"end\": 899}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.46331968903541565, \"start\": 887, \"end\": 899}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.46331968903541565, \"start\": 887, \"end\": 899}}, {\"mention_name\": {\"text\": \"historical weather data\", \"confidence\": 0.57245272397995, \"start\": 2418, \"end\": 2441, \"clean_text\": \"historical weather data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2418, \"end\": 2441, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2418, \"end\": 2441, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2418, \"end\": 2441}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2418, \"end\": 2441, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.8179948329925537, \"start\": 775, \"end\": 798}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.57245272397995, \"start\": 775, \"end\": 798}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5487986207008362, \"start\": 775, \"end\": 798}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5487986207008362, \"start\": 775, \"end\": 798}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [30]}}, {\"page\": 31, \"chunk\": 31, \"input_text\": \"<!-- Start of picture text --> oO N<br>e<br>22<br>£ e £ °<br>On Oo a<br>a ° a ‘2 -<br>€ € o (Good rainfall Average rainfall Poor rainfall<br>D es D<br>cL <<br>2ee<br>ooo :<br>fe] e fe]<br>o ood rainfall Average rainfall Poor rainfall o<br>neo e nD<br>£cc e . £cncn<br>Q o' e<br>oOD@oODD@oOD@oODoODD<br>a77 ad e a ; o...%<br>Peed oe<br>Tt e<br>22<br>£ £ ow ® £££ ‘ ‘<br>On>> ° o>><br>°°<br>ae-e- 2<br>a= iS * e<br>oo2..<br>2.. 2 :<br>So {Good rainfall Averagg rainfall Poor rain rain g& eo?<br>[e) oO e .}°<br>= e = e ° ge<br>aoo gre te Bee 804oo Good rainfall Average:: rainfall Poor rainfall<br>aD aD<br>£Y £ Paid<br>o S e.e ®<br>2oo e2o2oo eo ©<br>a a<br>a<br>y ‘|e<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> oO N<br>e<br>22<br>£ e £ °<br>On Oo a<br>a ° a ‘2 -<br>€ € o (Good rainfall Average rainfall Poor rainfall<br>D es D<br>cL <<br>2ee<br>ooo :<br>fe] e fe]<br>o ood rainfall Average rainfall Poor rainfall o<br>neo e nD<br>£cc e . £cncn<br>Q o' e<br>oOD@oODD@oOD@oODoODD<br>e a ; o...%<br>a77 ad<br>Peed oe<br>Tt e<br>22<br>£ £ ow ® £££ ‘ ‘<br>On>> ° o>><br>°°<br>ae-e- 2<br>a= iS * e<br>oo2.. 2 :<br>So {Good rainfall Averagg rainfall Poor rain rain g& eo?<br>[e) oO e .}°<br>= e = e ° ge<br>gre te Bee Good rainfall Average:: rainfall Poor rainfall<br>aoo 804oo<br>aD aD<br>£Y £ Paid<br>o S e.e ®<br>2oo e2o2oo eo ©<br>a a<br>a<br>y ‘|e<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [31]}}, {\"page\": 32, \"chunk\": 32, \"input_text\": \"<!-- Start of picture text --> 3 °<br>ale e<br>2 2 a<br>iy iy .<br>P= _@ 2 e e<br>on e bd g<br>fo} fo}<br>a. a - ~—e 7<br>£ ¢ © Good rainfall Average rainfall Poor rainfall<br>@ e e Ss e e<br>S&S o {Good rainfall Average rainfall Poor rainfall g e<br>==Oo=OoOo e : » “e@ .=oaOooaOoOo<br>8©© e 8o”o” e<br>22fe=oi&oOo2)fe=oi&oOo2)=oi&oOo2)oi&oOo2)&oOo2)oOo2)o2)e e ° aDG5feG5fe5fefe o O ° °<br>aaa<br>e<br>+77<br><!-- End of picture text --> <!-- Start of picture text --> 3 °<br>ale e<br>2 2 a<br>iy iy .<br>P= _@ 2 e e<br>on e bd g<br>fo} fo}<br>a. a - ~—e 7<br>£ ¢ © Good rainfall Average rainfall Poor rainfall<br>@ e e Ss e e<br>S&S o {Good rainfall Average rainfall Poor rainfall g e<br>==Oo=OoOo e : » “e@ .=oaOooaOoOo<br>e e<br>8©© 8o”o”<br>2) ° °<br>22fe=oi&oOo2)fe=oi&oOo2)=oi&oOo2)oi&oOo2)&oOo2)oOo2)o2)e aDG5feG5fe5fefe o O °<br>aaa<br>e<br>+77<br><!-- End of picture text --> <!-- Start of picture text --> N *<br>as °<br>gr<br>fo}a<br>o<br>cS= © |Good rainfall Averagé° rainfall Poor rain rain<br>r= °i<br>3a C)<br>©Sie ° .<br>aQ bs<br>oO<br>o<br>a s<br>y<br><!-- End of picture text -->\\n\\n<!-- Start of picture text -->\\n* T+ eo<br>° s ;<br>g °<br>.fo}ae<br>@ © {Good rainfall Average rainfall Poor rain<br>Poor rain rain cS& eo °<br>oa£o ° ;<br>=9a °<br>©& :<br>=ay ‘<br>Oo<br>oD<br>a<br>s<br>©<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> ag—~2<br>a” e<br>= as<br>—N bad<br>ia<br>w<br>=| e<br>= e<br>£8 ae<br>a<br>mo ad ®<br>£[o) ¢. :<br>3 i<br>a8 cs)<br>£7 le\\\"®<br>®<br>D<br>g<br>eeO & -Good rainfall7 Average rainfall: Poor rainfallF<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> a2=~<br>a<br>Eepara e<br>a e<br>oO ee.<br>in ‘<br>= :<br>& ee: ¢<br>£°e<br>a<br>S e<br>5 (0) ake<br>3<br>a e<br>£e . °@<br>a<br>D e<br>g<br>fw© = -Good rainfall, Average rainfall‘ Poor rainfallF<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [32]}}, {\"page\": 33, \"chunk\": 33, \"input_text\": \"<!-- Start of picture text --> as e a7<br>as . a<br>oa a °<br>-== 6 -aa ; :<br>qSN=SN== a~~~~ N<br>a s ea<br>a 1)<br>5 3 5 “ee<br>[ay [ay<br>E e = oe 8. ee<br>£o & e<br>a<br>D>oO<br>>oO oe >Do<br>5 = ° e $ oe @<br>2 ‘ ® il 2 ee<br>=aoTM°3—aoTM°3—oTM°3—°3—3—— 7 oe e<br>Da== Dao==<br>Be |e FE<br>6 § Good rainfall Average rainfall Poor rainfall © % -+_ Good Good rainfall Average rainfall Poor rainfall<br>as—~2—~2 e asonon e<br>= e 7<br>-<br>5 as 8asas . :<br>LS —H<br>7)a3a33 er).. eA7)57)55<br>SI ws SI e<br>Lo —ee Bo od<br>===<br>&££ .<br>aasas aoeoe e<br>a2>!>! ad oisPaPa<br>=«=«==<br>ooo<br>3 3 °<br>£8 £S °<br>o<br>D<br>S e 2SSS<br>se a : F £8 7 : F<br>O § -Good rainfall Average rainfall Poor rainfall © © Good rainfall Average rainfall Poor rainfall<br>as ° ag<br>oaacacac aoaoaoa<br>= ° =P<br>oSa—a——a——— e oameeameemeeameemeemee .FFF e<br>88 B® :<br>5 5 ew’<br>2[ay[ay[ay oa[ay[ay[ay -<br>=e=e==e===<br>&&&&<br>£o ; ‘£°e£°e£°e<br>a ; a e e<br>oO e © ;<br>D * teste) e<br>> e Naas) wr”<br>==<br>g$8$8$8 r g888888 .<br>—e—e——e———:<br>ooeoeoe<br>D ay<br>Cc s Cc e<br>Eo Eo<br>GO © +Good rainfall Average rainfall Poor rainfall 6 © +Good rainfall Average rainfall Poor rainfall<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> as e a7<br>as . a<br>oa a °<br>-== 6 -aa ; :<br>qSN=SN== a~~~~ N<br>a s ea<br>a 1)<br>5 3 5 “ee<br>[ay [ay<br>E e = oe 8. ee<br>£o & e<br>a<br>D>oO oe Do>Do<br>5 = ° e $ oe @<br>2 ‘ ® il 2 ee<br>7 e<br>=aoTM°3—aoTM°3—oTM°3—°3—3—— oe<br>Da== o=Dao==<br>Be |e FE<br>6 § Good rainfall Average rainfall Poor rainfall © % -+_ Good Good rainfall Average rainfall Poor rainfall<br>as—~2—~2 e asonon e<br>= e 7<br>-<br>5 as 8asas . :<br>LS —H<br>7)a3a33 er).. eA7)57)55<br>SI ws SI e<br>Lo —ee Bo od<br>===<br>&££ .<br>aasas aoeoe e<br>a2>!>! ad oisPaPa<br>=«=«==<br>ooo<br>3 3 °<br>£8 £S °<br>o<br>D<br>S e 2SSS<br>se a : F £8 7 : F<br>O § -Good rainfall Average rainfall Poor rainfall © © Good rainfall Average rainfall Poor rainfall<br>as ° ag<br>oaacacac aoaoaoa<br>= ° =P<br>.FFF e<br>oSa—a——a——— e oameeameemeeameemeemee<br>88 B® :<br>5 5 ew’<br>2[ay[ay[ay oa[ay[ay[ay -<br>=e=e==e===<br>&&&&<br>£o ; ‘£°e£°e£°e<br>a ; a e e<br>oO e © ;<br>D * teste) e<br>> e Naas) wr”<br>==<br>g$8$8$8 r g888888 .<br>—e—e——e———:<br>ooeoeoe<br>D ay<br>Cc s Cc e<br>Eo Eo<br>GO © +Good rainfall Average rainfall Poor rainfall 6 © +Good rainfall Average rainfall Poor rainfall<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> as ° ag<br>oaacacac aoaoaoa<br>= ° =P<br>.FFF e<br>oSa—a——a——— e oameeameemeeameemeemee<br>88 B® :<br>5 5 ew’<br>2[ay[ay[ay oa[ay[ay[ay -<br>=e=e==e===<br>&&&&<br>£o ; ‘£°e£°e£°e<br>a ; a e e<br>oO e © ;<br>D * teste) e<br>> e Naas) wr”<br>==<br>g$8$8$8 r g888888 .<br>—e—e——e———:<br>ooeoeoe<br>D ay<br>Cc s Cc e<br>Eo Eo<br>GO © +Good rainfall Average rainfall Poor rainfall 6 © +Good rainfall Average rainfall Poor rainfall<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [33]}}, {\"page\": 34, \"chunk\": 34, \"input_text\": \"<!-- Start of picture text --> 12<br>oo4<br>_—<br>2o2<br>2<br>o<br>a’ra2<br>So=' /<br>100 90 80 70 60 50 40 30 20 10 0<br>Loss exeedance (%)<br><!-- End of picture text --> <!-- Start of picture text --> =Ss<br>ao<br>aa<br>=Ss<br>gs<br>aLQ<br>oO<br>5<br>Go<<br>=<br>=<br>&<a<br>58<br>>\\\"<br>©>3<br>a,Soco= if<br>‘40090 80 70 60 50 40 30 20 10 0<br>Loss exeedance (%)<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [34]}}, {\"page\": 35, \"chunk\": 35, \"input_text\": \"revealed the limits of using standardized data sets in seeking to understand the heterogeneity of drought impacts across household types.\\n\\nWe find that consumption is reduced by 10-20 percent in the worst weather observations in our surveys (corresponding to about a 1 in 10-year drought event). We use the estimated vulnerability functions and the full hazard distribution in each household location to calculate the cost drought poses to welfare outcomes. The results indicate that risk is large. Poverty is 1–12 percent higher under the worst weather conditions relative to the best conditions observed in the last 13 years. This amounts to an increase in the total national poverty gap that ranges from US$4 million to US$2.4 billion (2011 PPP) across countries. Estimates are more precise for countries with a large share of production in maize, highland and pastoral areas. Potential biases in estimating impacts mean this is likely an underestimate of the true cost.\\n\\nThe cost of weather shocks on welfare provides an indication of the size of the increased need that would need to be met by adaptive social protection systems or access to financial services. The cost of weather shocks on consumption could be reduced by investments in adaptation policies such as soil management and tree cover or by other adaptive farming techniques. Investments in development that enable households to switch into less rainfall-dependent economic activities, either when droughts occur or as part of broader structural transformation, would also reduce the welfare cost of drought. Further work is needed to estimate how different investments would reduce the welfare cost of drought. This type of work is needed to help prioritize investments.\\n\\nStatistical analyses that rely only on observed historical data are generally limited by their short observation periods. In the context of drought risk, patterns and severity of future drought events might differ from what has been observed historically. Stochastic data sets provide a synthetic view of all events that could occur in a given year, together with their likelihood of occurring. These data sets typically allow for a more reliable analysis of drought risk, especially for events that occur less frequently. Use of a stochastic data set for Malawi showed how such an analysis can be carried out, but further work is needed to produce and use more reliable analyses of spatial and temporal correlations of drought hazard. A further extension of this work would incorporate climate trends into the analysis of welfare impacts in order to estimate the costs of longer-run welfare impacts (Baquie and Foucault 2023).\\n\\n# References\\n\\nABI (Association of British Insurers). 2011. _Industry Good Practice for Catastrophe Modelling_ . London: ABI. - <u>https://www.abi.org.uk/globalassets/sitecore/files/documents/publications/public/migrated/solvency ii/industry-good-practice-for-catastrophe-modelling.pdf.</u> Angrist, J. D., and J. S. Pischke. 2009. _Mostly Harmless Econometrics_ . Princeton, NJ: Princeton University Press. Anttila-Hughes, J., and M. Sharma. 2015. “Linking Risk Models to Microeconomic Indicators.” Policy Research Working Paper 7359, World Bank, Washington, DC. <u><mark>http://hdl.handle.net/10986/22235.</mark></u> <mark>Artuc. E., G. Porto, B. Rijkers. 2023. Crops, Conflict, and Climate Change. World Bank Mimeo.</mark>\\n\\n- Auffhammer, Maximilian. 2018. “Quantifying Economic Damages from Climate Change.” Journal of Economic Perspectives 32 (4): 33–52. https://doi.org/10.1257/jep.32.4.33.\\n\\n- Baez, Javier E., Varun Kshirsagar, and Emmanuel Skoufias. 2019. “Adaptive Safety Nets for Rural Africa : DroughtSensitive Targeting with Sparse Data.” Policy Research Working Paper Series, Policy Research Working Paper Series, , December. https://ideas.repec.org//p/wbk/wbrwps/9071.html.\\n\\n- Baquie, S., and G. Foucault. 2023. “Background Note on Bringing Climate Change into Vulnerability Analysis.” Unpublished background paper. World Bank.\\n\\n- Baquie, Sandra, and Habtamu Neda Fuje. 2020. “Vulnerability to Poverty Following Extreme Weather Events in Malawi.” Policy Research Working Paper Series, Policy Research Working Paper Series, , October. https://ideas.repec.org//p/wbk/wbrwps/9435.html.\\n\\n- Bates, Paul D., James Savage, Oliver Wing, Niall Quinn, Christopher Sampson, Jeffrey Neal, and Andrew Smith. 2023. “A Climate-Conditioned Catastrophe Risk Model for UK Flooding.” _European Geosciences Union_ 23 (2): 891–908. <u>https://nhess.copernicus.org/articles/23/891/2023/.</u>\\n\\n34\", \"datasets\": [{\"mention_name\": {\"text\": \"standardized data sets\", \"confidence\": 0.5298550128936768, \"start\": 29, \"end\": 51, \"clean_text\": \"standardized data sets\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 29, \"end\": 51, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 29, \"end\": 51, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 29, \"end\": 51}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 29, \"end\": 51, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.49767595529556274, \"start\": 29, \"end\": 51}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5298550128936768, \"start\": 29, \"end\": 51}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7697003483772278, \"start\": 29, \"end\": 51}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7697003483772278, \"start\": 29, \"end\": 51}}, {\"mention_name\": {\"text\": \"observed historical data\", \"confidence\": 0.4345763027667999, \"start\": 1780, \"end\": 1804, \"clean_text\": \"observed historical data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1780, \"end\": 1804, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1780, \"end\": 1804, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1780, \"end\": 1804}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1780, \"end\": 1804, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.7101039886474609, \"start\": 950, \"end\": 974}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4345763027667999, \"start\": 950, \"end\": 974}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8301770091056824, \"start\": 950, \"end\": 974}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8301770091056824, \"start\": 950, \"end\": 974}}, {\"mention_name\": {\"text\": \"Stochastic data sets\", \"confidence\": 0.7303600311279297, \"start\": 1997, \"end\": 2017, \"clean_text\": \"Stochastic data sets\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1997, \"end\": 2017, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1997, \"end\": 2017, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1997, \"end\": 2017}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1997, \"end\": 2017, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.5742718577384949, \"start\": 1167, \"end\": 1187}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7303600311279297, \"start\": 1167, \"end\": 1187}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5429643988609314, \"start\": 1167, \"end\": 1187}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5429643988609314, \"start\": 1167, \"end\": 1187}}, {\"mention_name\": {\"text\": \"stochastic data set for Malawi\", \"confidence\": 0.7547754645347595, \"start\": 2273, \"end\": 2303, \"clean_text\": \"stochastic data set for Malawi\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2273, \"end\": 2303, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2273, \"end\": 2303, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2273, \"end\": 2303}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2273, \"end\": 2303, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.43702659010887146, \"start\": 535, \"end\": 565}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.7547754645347595, \"start\": 535, \"end\": 565}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.6273422837257385, \"start\": 535, \"end\": 565}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6273422837257385, \"start\": 535, \"end\": 565}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [35]}}, {\"page\": 36, \"chunk\": 36, \"input_text\": \"Blanchard, Antoine, and Luis Sousa. 2021. “Probabilistic Hazard Analysis for Drought in Sub-Saharan Africa: Stochastic Precipitation, Soil Moisture, and Vegetation Index Datasets for Malawi.” AIR Worldwide.\\n\\nClarke, Daniel J and Stefan Dercon. 2016. _Dull Disasters? How planning ahead will make a difference._ Oxford University Press.\\n\\nDell, M., B. Jones, and B. Olken. 2014. “What Do We Learn from the Weather? The New Climate–Economy Literature.” _Journal of Economic Literature_ 52 (3): 740–98. <u>http://dx.doi.org/10.1257/jel.52.3.740.</u>\\n\\n- Deveraux, S. 2007. \\\"The impact of droughts and floods on food security and policy options to alleviate negative effects,\\\" Agricultural Economics 37(s1): 47-58, December.\\n\\n- Lott, and George Mburathi, eds. 2019. _Farming Systems and Food Security in Africa: Priorities for Science and Policy under Global Change._ Routledge.\\n\\n- Fielder, T., A. J. Pitman, K. Mackenzie, N. Wood, C. Jakob, and S. E. Perkins-Kirkpatrick. 2021. “Business Risk and the Emergence of Climate Analytics.” _Nature Climate Change_ 11: 87–94.\\n\\n<u>https://www.nature.com/articles/s41558-020-00984-6.</u>\\n\\n- Hallegatte, Stephane, Mook Bangalore, Laura Bonzanigo, Marianne Fay, Tamaro Kane, Ulf Narloch, Julie Rozenberg, David Treguer, and Adrien Vogt-Schilb. 2016. Shock Waves: Managing the Impacts of Climate Change on Poverty. Washington, DC: World Bank. https://doi.org/10.1596/978-1-4648-0673-5.\\n\\n- Hill, R., and C. Porter. 2017. “Vulnerability to Drought and Food Price Shocks: Evidence from Ethiopia.” _World Development_ 96: 65–77.\\n\\n- Hill, R., and K. Vinha. 2022. “Mauritania and Vulnerability to Drought.” Unpublished paper, World Bank.\\n\\n- Hill, Ruth, and Carolina Mejia-Mantilla. 2017. With a Little Help: Shocks, Agricultural Income, and Welfare in Uganda. Policy Research Working Papers. The World Bank. https://doi.org/10.1596/1813-9450-7935.\\n\\n- James, G., D. Witten, T. Hastie, and R. Tibshirani. 2013. _An Introduction to Statistical Learning with Applications in R_ . Springer.\\n\\nKing, G., and L. Zeng. 2006. “The Dangers of Extreme Counterfactuals.” _Political Analysis_ 14: 131–59.\\n\\n- Kochhar, Nishtha, and Erwin Knippenberg. 2023. “Droughts and Welfare in Afghanistan.” Policy Research Working Paper 10272, World Bank, Washington, DC.\\n\\n- Lobell, David B., Wolfram Schlenker, and Justin Costa-Roberts. 2011. “Climate Trends and Global Crop Production Since 1980.” Science 333 (6042): 616–20. https://doi.org/10.1126/science.1204531.\\n\\n- Mummolo, J., and E. Peterson. 2018. “Improving the Interpretation of Fixed Effects Regression Results.” _Political Science Research and Methods_ 6 (4): 829–35. doi:10.1017/psrm.2017.44.\\n\\n- NGFS (Network for Greening the Financial System). 2022. “Physical Climate Risk Assessment: Practical Lessons for the Development of Climate Scenarios with Extreme Weather Events from Emerging Markets and Developing Economies.” <u>https://www.ngfs.net/sites/default/files/media/2022/09/02/ngfs_physical_climate_risk_assessment.pdf.</u>\\n\\n- Porter, Catherine, and Emily White (2016) “Potential for Application of a Probabilistic Catastrophe Risk Modelling Framework to Poverty Outcomes: General Form Vulnerability Functions Relating Household Poverty Outcomes to Hazard Intensity in Ethiopia.” Policy Research Working Paper 7717, World Bank, Washington DC. <u>http://documents.worldbank.org/curated/en/854961468184437420/pdf/WPS7717.pdf.</u>\\n\\n- Purdy, Meghan. 2021. “Cat Models: Convenient, Familiar—But Flawed for Projecting Physical Climate Risk.” _Medium_ , June 14, 2021. <u>https://medium.com/jupiterintel/cat-models-f0fc7e890bef.</u>\\n\\n- Sen, A. 1981. Poverty and Famines: An Essay on Entitlement and Deprivation. Oxford University Press.\\n\\n- Skoufias, E., et al. 2023. “Climatic Variability and Child Undernutrition in Sub-Saharan Africa Redux.” World Bank Mimeo.\\n\\n- Verisk. 2013. “Modeling Fundamentals: What Is AAL?” March 25, 2013. <u>https://www.air</u>\\n\\n<u>worldwide.com/publications/air-currents/2013/Modeling-Fundamentals--What-Is-AAL-/.</u>\\n\\n- Wineman, Ayala, Nicole M. Mason, Justus Ochieng, and Lilian Kirimi. 2017. “Weather Extremes and Household Welfare in Rural Kenya.” Food Security 9 (2): 281–300. https://doi.org/10.1007/s12571-016-0645-z.\\n\\n35\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [36]}}, {\"page\": 37, \"chunk\": 37, \"input_text\": \"# Appendix A. Exposure Data\\n\\nThe exposure data layers are static inputs for the drought/welfare modeling exercise that are fixed in analysis, unlike hazard components. These exposure data layers contain information relating to various spatial characteristics as set out in table A.1.\\n\\n**Table A.1. Information on exposure data**\\n\\n|**Data set**|**Description**|**Source and shape files**|**Citation**|\\n|---|---|---|---|\\n|Uniform<br>Resolution Grid<br>(URG)|A country x, y grid (w|here extents are buffered to 2° from countr|y boundary).|\\n|World Bank –<br>country<br>administrative<br>boundaries|World Bank–approved administrative<br>boundaries (Admin0) (and polygons) including<br>international boundaries, disputed areas,<br>coastlines,and lakes.|https://datacatalog.worldbank.org/datas<br>et/world-bank-official-boundaries|World Bank Basemap Collection,<br>February 2020.|\\n|GADM – Database<br>of Global<br>Administrative<br>Areas (Admin0–<br>Admin6)|GADM is a database of the location of the<br>world's administrative areas (boundaries).<br>Administrative areas in this database include<br>countries, counties, departments, etc. and cover<br>every country in the world. For each area some<br>attributes are provided, including the name and<br>in some cases variant names.|https://biogeo.ucdavis.edu/data/gadm3.<br>6/gadm36_shp.zip|Global Administrative Areas. 2012.<br>GADM database of Global<br>Administrative Areas, version 2.0.|\\n|GAUL – Global<br>Administrative<br>Unit Layers<br>(Gaul0–Gaul1)|Information is provided at two spatial levels:<br>national (Gaul0) and subnational (Gaul1). The<br>reference spatial layers are derived from the<br>GAUL data set, implemented by the UN Food<br>and Agriculture Organization (FAO) within the<br>CountrySTAT and Agricultural Market<br>Information System(AMIS) projects.|https://data.europa.eu/euodp/en/data/<br>dataset/jrc-10112-<br>10004/resource/609a472e-7e26-4b63-<br>8003-17298cb45e2a<br>https://mars.jrc.ec.europa.eu/asap/files/<br>gaul1_asap.zip|Europa.EU<br>http://data.europa.eu/eli/dec/2011/8<br>33/oj.|\\n|Crop coverage –<br>Copernicus Global<br>Land Cover Layers|Mapped crop coverage data layer (2019<br>vintage).|https://land.copernicus.eu/global/produ<br>cts/lc|Buchhorn, M., B. Smets, L. Bertels, B.<br>De Roo, M. Lesiv, N.-E. Tsendbazar, M.<br>Herold, and S. Fritz. Copernicus Global<br>Land Service: Land Cover 100m:<br>Collection 3: epoch 2019: Globe 2020.|\\n|Population – GHS-<br>POP R2019A|This spatial raster data set depicts the<br>distribution of population, expressed as the<br>number ofpeopleper cell.|https://ghsl.jrc.ec.europa.eu/ghs_pop20<br>19.php|GHS population grid multi-temporal<br>(1975–1990–2000–2015).|\\n\\n36\", \"datasets\": [{\"mention_name\": {\"text\": \"GAUL data set\", \"confidence\": 0.7814273238182068, \"start\": 1646, \"end\": 1659, \"clean_text\": \"GAUL data set\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1646, \"end\": 1659, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1646, \"end\": 1659, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1646, \"end\": 1659}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1646, \"end\": 1659, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.8312357068061829, \"start\": 1646, \"end\": 1659}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7814273238182068, \"start\": 1646, \"end\": 1659}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8155488967895508, \"start\": 1646, \"end\": 1659}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8155488967895508, \"start\": 1646, \"end\": 1659}}, {\"mention_name\": {\"text\": \"Mapped crop coverage data layer\", \"confidence\": 0.5414444804191589, \"start\": 2116, \"end\": 2147, \"clean_text\": \"Mapped crop coverage data layer\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2116, \"end\": 2147, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2116, \"end\": 2147, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2116, \"end\": 2147}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2116, \"end\": 2147, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.964525580406189, \"start\": 2116, \"end\": 2147}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5414444804191589, \"start\": 2116, \"end\": 2147}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.3785569369792938, \"start\": 2116, \"end\": 2147}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.3785569369792938, \"start\": 2116, \"end\": 2147}}, {\"mention_name\": {\"text\": \"POP R2019A\", \"confidence\": 0.4847664535045624, \"start\": 2439, \"end\": 2449, \"clean_text\": \"POP R2019A\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2439, \"end\": 2449, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2439, \"end\": 2449, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2439, \"end\": 2449}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2439, \"end\": 2449, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.9480947852134705, \"start\": 5, \"end\": 15}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4847664535045624, \"start\": 5, \"end\": 15}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.3591984510421753, \"start\": 5, \"end\": 15}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.3591984510421753, \"start\": 5, \"end\": 15}}, {\"mention_name\": {\"text\": \"spatial raster data set\", \"confidence\": 0.33542492985725403, \"start\": 2455, \"end\": 2478, \"clean_text\": \"spatial raster data set\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2455, \"end\": 2478, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2455, \"end\": 2478, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2455, \"end\": 2478}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2455, \"end\": 2478, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.9480947852134705, \"start\": 21, \"end\": 44}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.33542492985725403, \"start\": 21, \"end\": 44}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.3591984510421753, \"start\": 21, \"end\": 44}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.3591984510421753, \"start\": 21, \"end\": 44}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [37]}}, {\"page\": 38, \"chunk\": 38, \"input_text\": \"|Travel Time to<br>Markets –<br>International Food<br>Policy Research<br>Institute (IFPRI)|The data set includes five data layers<br>representing travel time to the nearest market<br>of five sizes (population of 20,000, 50,000,<br>100,000, 250,000, and 500,000) respectively, on<br>five arc-minute grids (~9 km) across SSA (2016-<br>01-29).|https://dataverse.harvard.edu/file.xhtml<br>?persistentId=doi:10.7910/DVN/YKDWJD<br>/UFOGZS&version=2.2|HarvestChoice, International Food<br>Policy Research Institute (IFPRI). 2016.<br>“Travel Time to Markets in Africa<br>South of the Sahara.”<br>https://doi.org/10.7910/DVN/YKDWJD<br>,Harvard Dataverse,V2.|\\n|---|---|---|---|\\n|Global Agro-<br>Ecological Zones<br>(GAEZ)|GAEZ generates large databases of (i) natural<br>resource endowments relevant for agricultural<br>uses; (ii) spatially detailed results of individual<br>land utilization type assessments in terms of<br>suitability and attainable yields; (iii) spatially<br>detailed results of estimate/actual yields of main<br>food and fiber commodities for all rain-fed and<br>irrigated cultivated areas; and (iv) spatially<br>detailed yield and production gaps, also for main<br>food and fiber commodities.|http://pure.iiasa.ac.at/id/eprint/13290|Fischer, G., F. O. Nachtergaele, S<br>Prieler, E. Teixeira, G. Toth, H. van<br>Velthuizen, L. Verelst, and D. Wiberg.<br>2012._Global Agro-Ecological Zones_<br>_(GAEZ v 3.0)—Model Documentation_.<br>Laxenburg, Austria: IIASA<br>(International Institute for Applied<br>Systems Analysis) and Rome, Italy:<br>FAO.|\\n|Livelihood zones|Map of farming systems in Africa in 2015. The<br>farming systems analysis integrates an extensive<br>range of spatial data, administrative statistics,<br>assessment reports, and expert knowledge in<br>order to update the African component of the<br>2001 FAO/World Bank farming systems analysis.<br>Data sources include GAEZ FAO/IIASA, FAOSTAT,<br>and HarvestChoice.|Personal copy of data from Christopher<br>M. Auricht(https://auricht.com/)|Dixon, John, Dennis P. Garrity, Jean-<br>Marc Boffa, Timothy O. Williams,<br>Tilahun Amede, Christopher Auricht,<br>Rosemary Lott, and George Mburathi,<br>eds. 2019._Farming Systems and Food_<br>_Security in Africa: Priorities for Science_<br>_and Policy Under Global Change._<br>Routledge.(See figure 2.2b.)|\\n\\n_Source:_ World Bank compilation.\\n\\n# Appendix B. Hazard Data\\n\\n**Table B.1. Information on hazard data**\\n\\n||**ERA5**|**CHIRPS**|**SPEI**|**WRSI**|**SWI10**|**NDVI**|\\n|---|---|---|---|---|---|---|\\n|**Hazard**<br>**measure**|Precipitation<br>(modeled)|Precipitation|Evapotranspiration|Availability of water for<br>crops|Soil moisture|Vegetation<br>(greenness)|\\n|Full name|European Centre for<br>Medium-Range<br>Weather Forecasts<br>(ECMWF) Reanalysis<br>5th Generation|Climate Hazards Group<br>InfraRed Precipitation<br>with Station data|Standardized<br>Precipitation-<br>Evapotranspiration<br>Index|Water Requirement<br>Satisfaction Index|Average of Soil Water<br>Index over a 10-day<br>period|Normalized<br>Difference Vegetation<br>Index|\\n\\n37\", \"datasets\": [{\"mention_name\": {\"text\": \"World Bank compilation\", \"confidence\": 0.3604324460029602, \"start\": 2336, \"end\": 2358, \"clean_text\": \"World Bank compilation\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2336, \"end\": 2358, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2336, \"end\": 2358, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2336, \"end\": 2358}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2336, \"end\": 2358, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.43973439931869507, \"start\": 192, \"end\": 214}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3604324460029602, \"start\": 192, \"end\": 214}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7372497320175171, \"start\": 192, \"end\": 214}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7372497320175171, \"start\": 192, \"end\": 214}}, {\"mention_name\": {\"text\": \"Hazard Data\", \"confidence\": 0.36165937781333923, \"start\": 2375, \"end\": 2386, \"clean_text\": \"Hazard Data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2375, \"end\": 2386, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2375, \"end\": 2386, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2375, \"end\": 2386}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2375, \"end\": 2386, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6865462064743042, \"start\": 231, \"end\": 242}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.36165937781333923, \"start\": 231, \"end\": 242}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5687376856803894, \"start\": 231, \"end\": 242}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5687376856803894, \"start\": 231, \"end\": 242}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [38]}}, {\"page\": 39, \"chunk\": 39, \"input_text\": \"|Description|ERA5 climate<br>reanalysis gives a<br>numerical<br>description of the<br>recent climate,<br>produced by<br>combining models<br>with observations.<br>Total precipitation is<br>the accumulated<br>liquid and frozen<br>water, comprising<br>rain and snow, that<br>falls to the Earth’s<br>surface. It is the sum<br>of large-scale<br>precipitation and<br>convective<br>precipitation.|The CHIRPS data<br>archive is a quasi-<br>global (50S–50N)<br>gridded 0.05°<br>resolution times series<br>covering precipitation<br>from 1981 to near-real<br>time.|The SPEI is a multi-<br>scalar drought index<br>based on climatic data.<br>It is a ratio of actual to<br>potential<br>evapotranspiration.|This spatially explicit<br>index is an indicator of<br>crop performance<br>based on the<br>availability of water to<br>the crop during a<br>growing season.|The Soil Water Index<br>quantifies the<br>moisture condition at<br>various depths in the<br>soil. It is mainly driven<br>by precipitation via<br>the process of<br>infiltration. Various t-<br>values are available: 1,<br>5, 10, 15, 20, 40, 60,<br>100. The T- value<br>simulates the<br>infiltration time into<br>deeper soil layers and<br>hence indicates<br>qualitatively the depth<br>of the SWI estimate.<br>T40 was selected as<br>being generally<br>representative, based<br>on expert advice.|The Terra Moderate<br>Resolution Imaging<br>Spectroradiometer<br>(MODIS) Vegetation<br>Indices Monthly<br>(MOD13C2) Version<br>6.1 product provides<br>a vegetation index<br>value at a per pixel<br>basis. The NDVI is<br>referred to as the<br>continuity index to<br>the existing National<br>Oceanic and<br>Atmospheric<br>Administration<br>Advanced Very High<br>Resolution<br>Radiometer (NOAA-<br>AVHRR)-derived<br>NDVI.|\\n|---|---|---|---|---|---|---|\\n|Temporal<br>coverage|1979 to present|1981 to near-real time<br>precipitation time<br>series|1901–2018|2000–2021|2007–2021|2000–2021|\\n|Spatial coverage|Global|Global|Global|Regional|Global|Global|\\n|Resolution|Reanalysis: 0.25°<br>(~28 km)|0.05° (~5.5 km)|0.5° (~5.5 km)|0.1° (~11 km)|0.1° (~11 km)|0.5° (~5.5 km)|\\n|Output<br>information|24-hour aggregation<br>since the beginning<br>of the forecast|mm/per unit of time|Z-values, monthly|WRSI per raster cell;<br>current WRSI data give<br>range of 0–100 percent<br>of total|An index describing<br>the relation between<br>surface soil moisture<br>and profile soil<br>moisture as a function<br>of time(percent)|Vegetation density,<br>monthly|\\n|Data set source|https://cds.climate.c<br>opernicus.eu/cdsapp<br>#!/dataset/reanalysi|https://data.chc.ucsb.e<br>du/products/CHIRPS-<br>2.0/africa_dekad/tifs/|https://spei.csic.es/|https://edcftp.cr.usgs.g<br>ov/project/fews/dekad<br>al/|https://land.copernicu<br>s.eu/global/products/s<br>wi|https://lpdaac.usgs.g<br>ov/products/mod13c<br>2v061/|\\n\\n38\", \"datasets\": [{\"mention_name\": {\"text\": \"CHIRPS data\", \"confidence\": 0.6300767064094543, \"start\": 394, \"end\": 405, \"clean_text\": \"CHIRPS data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 394, \"end\": 405, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 394, \"end\": 405, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 394, \"end\": 405}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 394, \"end\": 405, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.7712457180023193, \"start\": 394, \"end\": 405}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6300767064094543, \"start\": 394, \"end\": 405}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4344020187854767, \"start\": 394, \"end\": 405}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4344020187854767, \"start\": 394, \"end\": 405}}, {\"mention_name\": {\"text\": \"WRSI data\", \"confidence\": 0.3085503876209259, \"start\": 2257, \"end\": 2266, \"clean_text\": \"WRSI data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2257, \"end\": 2266, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2257, \"end\": 2266, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2257, \"end\": 2266}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2257, \"end\": 2266, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.2637444734573364, \"start\": 2257, \"end\": 2266}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3085503876209259, \"start\": 2257, \"end\": 2266}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.45674222707748413, \"start\": 2257, \"end\": 2266}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.45674222707748413, \"start\": 2257, \"end\": 2266}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [39]}}, {\"page\": 40, \"chunk\": 40, \"input_text\": \"||s-era5-single-<br>levels?tab=overview|||https://earlywarning.us<br>gs.gov/fews/product/1<br>28|||\\n|---|---|---|---|---|---|---|\\n|Data set<br>reference/citati<br>on|Hersbach, H., B. Bell,<br>P. Berrisford, G.<br>Biavati, A. Horányi, J.<br>Muñoz Sabater, J.<br>Nicolas, C. Peubey,<br>R. Radu, I. Rozum, D.<br>Schepers, A.<br>Simmons, C. Soci, D.<br>Dee, and J.-N.<br>Thépaut. 2018.<br>“ERA5 Hourly Data<br>on Single Levels<br>from 1979 to<br>Present.” Copernicus<br>Climate Change<br>Service (C3S) Climate<br>Data Store (CDS).|Funk, C.C., P. J.<br>Peterson, M. F.<br>Landsfeld, D. H.<br>Pedreros, J. P. Verdin,<br>J. D. Rowland, B. E.<br>Romero, G. J. Husak, J.<br>C. Michaelsen, and A.<br>P. Verdin. 2014. “A<br>Quasi-Global<br>Precipitation Time<br>Series for Drought<br>monitoring.” US<br>Geological Survey Data<br>Series 832, ftp://chg-<br>ftpout.geog.ucsb.edu/<br>pub/org/chg/products/<br>CHIRPS-<br>2.0/docs/USGS-<br>DS832.CHIRPS.pdf.|Sergio M. Vicente-<br>Serrano, Instituto<br>Pirenaico de Ecología,<br>Zaragoza, Spain.<br>Santiago Beguería,<br>Estación Experimental<br>de Aula Dei, Zaragoza,<br>Spain.|Verdin, J., and R. Klaver.<br>2002. “Grid-cell-based<br>Crop Water Accounting<br>for the Famine Early<br>Warning System.”<br>_Hydrological Processes_<br>16 (8): 1617–30. DOI:<br>10.1002/hyp.1025.|Copernicus Service<br>information 2018.|Didan, K. 2021.<br>MODIS/Terra<br>Vegetation Indices<br>Monthly L3 Global<br>0.05Deg CMG V061<br>[data set]. NASA<br>EOSDIS Land<br>Processes DAAC.<br>Accessed 2021-06-24<br>from<br>https://doi.org/10.50<br>67/MODIS/MOD13C2.<br>061.|\\n|Additional<br>information/<br>notes||||Range of WRSI values:<br>< 50: Failure<br>50–60: Poor<br>60–80: Mediocre<br>80–95: Average<br>95–99: Good<br>99–100: Very Good<br>0: N/A<br>The following values<br>are flags (and not<br>specific WRSI values in<br>the normal range 0–<br>100):<br>253: No start (late)<br>254: Yet to start|||\\n\\n39\", \"datasets\": [{\"mention_name\": {\"text\": \"ERA5 Hourly Data\", \"confidence\": 0.3860248923301697, \"start\": 386, \"end\": 402, \"clean_text\": \"ERA5 Hourly Data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 386, \"end\": 402, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 386, \"end\": 402, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 386, \"end\": 402}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 386, \"end\": 402, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.6716985702514648, \"start\": 386, \"end\": 402}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3860248923301697, \"start\": 386, \"end\": 402}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.45796963572502136, \"start\": 386, \"end\": 402}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.45796963572502136, \"start\": 386, \"end\": 402}}, {\"mention_name\": {\"text\": \"Geological Survey Data\", \"confidence\": 0.4474940001964569, \"start\": 792, \"end\": 814, \"clean_text\": \"Geological Survey Data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 792, \"end\": 814, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 792, \"end\": 814, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 792, \"end\": 814}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 792, \"end\": 814, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9679162502288818, \"start\": 792, \"end\": 814}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4474940001964569, \"start\": 792, \"end\": 814}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4464907646179199, \"start\": 792, \"end\": 814}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4464907646179199, \"start\": 792, \"end\": 814}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [40]}}, {\"page\": 41, \"chunk\": 41, \"input_text\": \"_Source:_ World Bank compilation.\\n\\nTable B.2. Definitions used for household descriptive variables from country surveys\\n\\n|**Country**|**Land ownership (0 if none, 1 if any)**|**Education**|\\n|---|---|---|\\n|Ethiopia|Ownership of land plots|Literacy|\\n|Lesotho|Ownership or cultivation of land|Level: primary complete|\\n|Malawi|Land area cultivated|Can read and write|\\n|Mauritania|Hectares owned|Literacy|\\n|Mozambique|Number of plots|Can read and write|\\n|Niger|Land area cultivated|Level: any|\\n|Nigeria|Land area cultivated|Level: primary complete|\\n|Zambia|Land area cultivated|Level: primary complete|\\n|Zimbabwe|Ownership of any arable land|Level: primary complete|\\n\\n# _Source:_ World Bank.\\n\\n_Note:_ Country-specific definitions for land ownership and education are based on survey information.\\n\\n40\", \"datasets\": [{\"mention_name\": {\"text\": \"country surveys\", \"confidence\": 0.7074818015098572, \"start\": 104, \"end\": 119, \"clean_text\": \"country surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 104, \"end\": 119, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 104, \"end\": 119, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 104, \"end\": 119}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 104, \"end\": 119, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8421221971511841, \"start\": 104, \"end\": 119}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7074818015098572, \"start\": 104, \"end\": 119}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9095562696456909, \"start\": 104, \"end\": 119}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9095562696456909, \"start\": 104, \"end\": 119}}, {\"mention_name\": {\"text\": \"survey information\", \"confidence\": 0.4067617356777191, \"start\": 771, \"end\": 789, \"clean_text\": \"survey information\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 771, \"end\": 789, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 771, \"end\": 789, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 771, \"end\": 789}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 771, \"end\": 789, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9831657409667969, \"start\": 771, \"end\": 789}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4067617356777191, \"start\": 771, \"end\": 789}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.6645610332489014, \"start\": 771, \"end\": 789}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6645610332489014, \"start\": 771, \"end\": 789}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [41]}}, {\"page\": 42, \"chunk\": 42, \"input_text\": \"# Appendix C. Hazard Simulations\\n\\nThe Centre for Disaster Protection and the World Bank asked Verisk’s AIR Worldwide (AIR) to develop a series of stochastic catalogs of relevant climate variables in Malawi. These would support the ongoing effort to develop a standardized approach to ex ante estimation of the welfare impacts of moderate and severe droughts in SSA, complementing the ongoing empirical analysis. The catalogs would allow a forward-looking assessment of possible future events that goes beyond the reliance on the history of past events (i.e., would allow a more robust probabilistic view of future events).\\n\\nAIR’s methodology, data sources (table C.1), data validation, and results are set out in Blanchard and Sousa (2021) and summarized below, with extracts largely taken directly from the paper.\\n\\n## **Developing the Stochastic Catalogs for Malawi**\\n\\nUnder the scope of this project, AIR developed stochastic catalogs of three drought-related variables for Malawi— precipitation, soil moisture, and vegetation index.\\n\\nThe 10,000-year stochastic catalog produced for each variable includes 10,000 samples of annual time histories for the variable of interest, which all reflect conditions that are likely under the current (and near-current) climate.\\n\\n**Table C.1. Data sets used to generate the stochastic catalogs**\\n\\n|**Variable**|**Description**|**Data set**|**Resolution,**<br>**frequency,**<br>**period covered**|**Source**|\\n|---|---|---|---|---|\\n|Atmospheric<br>moisture|Atmospheric<br>reanalysis of the<br>global climate|ECMWF Reanalysis<br>v5 (ERA5)|0.25°, three-<br>hourly, 1979–<br>2018|ECMWF|\\n|Precipitation|Atmospheric<br>precipitation<br>estimates from<br>rain gauge and<br>satellite<br>observations|CHIRPS 2.0|0.05°, dekadal<br>(10-day), 2000–<br>2020|CHIRPS|\\n|Soil moisture|Soil Water Index<br>(SWI) at a depth of<br>40 cm|SWI10-40|0.1°, dekadal,<br>2007–present|Copernicus Global<br>Land Service|\\n|Vegetation index|Normalized<br>Difference<br>Vegetation Index|MOD13C2 Version 6|0.05°, monthly,<br>2000–2020|MODIS|\\n\\n_Source:_ Blanchard and Sousa 2021.\\n\\n_Note:_ CHIRPS = Climate Hazards Group InfraRed Precipitation with Station; ECMWF = <mark>European Centre for Medium-Range Weather Forecasts;</mark> MODIS = Terra Moderate Resolution Imaging Spectroradiometer.\\n\\nThe approach is based on random sampling of stochastic perturbations of historical data to generate stochastic realizations of precipitation values over Malawi. The stochastic catalogs of NVDI and SWI are then derived from the stochastic precipitation catalog.\\n\\nMore specifically, the 10,000-year stochastic precipitation catalog has a spatial resolution of 0.05° and dekadal temporal resolution. It was developed using random sampling of stochastic perturbations of atmospheric moisture data (ERA5 [( <mark>European Centre for Medium-Range Weather Forecasts)</mark> Reanalysis v5]) between 1979 and 2018, at a native resolution of 0.25° and three-hour intervals (as an input to the process). While CHIRPS data could not be used in place of ERA5 at the modeling stage, they were used to augment a climatological adjustment to reflect “real” rainfall readings and climatology compared to ERA5.\\n\\n41\", \"datasets\": [{\"mention_name\": {\"text\": \"stochastic catalogs of relevant climate variables in Malawi\", \"confidence\": 0.31823426485061646, \"start\": 146, \"end\": 205, \"clean_text\": \"stochastic catalogs of relevant climate variables in Malawi\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 146, \"end\": 205, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 146, \"end\": 205, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 146, \"end\": 205}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 146, \"end\": 205, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.5017535090446472, \"start\": 146, \"end\": 205}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.31823426485061646, \"start\": 146, \"end\": 205}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.652547299861908, \"start\": 146, \"end\": 205}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.652547299861908, \"start\": 146, \"end\": 205}}, {\"mention_name\": {\"text\": \"ECMWF Reanalysis\", \"confidence\": 0.35802483558654785, \"start\": 1546, \"end\": 1562, \"clean_text\": \"ECMWF Reanalysis\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1546, \"end\": 1562, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1546, \"end\": 1562, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1546, \"end\": 1562}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1546, \"end\": 1562, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.47253069281578064, \"start\": 742, \"end\": 758}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.35802483558654785, \"start\": 742, \"end\": 758}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6285204291343689, \"start\": 742, \"end\": 758}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6285204291343689, \"start\": 742, \"end\": 758}}, {\"mention_name\": {\"text\": \"Soil Water Index\", \"confidence\": 0.35946983098983765, \"start\": 1807, \"end\": 1823, \"clean_text\": \"Soil Water Index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1807, \"end\": 1823, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1807, \"end\": 1823, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1807, \"end\": 1823}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1807, \"end\": 1823, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9240495562553406, \"start\": 1003, \"end\": 1019}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.35946983098983765, \"start\": 1003, \"end\": 1019}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7056294083595276, \"start\": 1003, \"end\": 1019}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7056294083595276, \"start\": 1003, \"end\": 1019}}, {\"mention_name\": {\"text\": \"Vegetation index\", \"confidence\": 0.3197989761829376, \"start\": 1932, \"end\": 1948, \"clean_text\": \"Vegetation index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1932, \"end\": 1948, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1932, \"end\": 1948, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1932, \"end\": 1948}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1932, \"end\": 1948, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.8867694139480591, \"start\": 53, \"end\": 69}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3197989761829376, \"start\": 53, \"end\": 69}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6550785303115845, \"start\": 53, \"end\": 69}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6550785303115845, \"start\": 53, \"end\": 69}}, {\"mention_name\": {\"text\": \"stochastic precipitation catalog\", \"confidence\": 0.5147249698638916, \"start\": 2594, \"end\": 2626, \"clean_text\": \"stochastic precipitation catalog\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2594, \"end\": 2626, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2594, \"end\": 2626, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2594, \"end\": 2626}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2594, \"end\": 2626, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.6579141616821289, \"start\": 715, \"end\": 747}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5147249698638916, \"start\": 715, \"end\": 747}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4623093008995056, \"start\": 715, \"end\": 747}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4623093008995056, \"start\": 715, \"end\": 747}}, {\"mention_name\": {\"text\": \"CHIRPS data\", \"confidence\": 0.3939986526966095, \"start\": 2996, \"end\": 3007, \"clean_text\": \"CHIRPS data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2996, \"end\": 3007, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2996, \"end\": 3007, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2996, \"end\": 3007}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2996, \"end\": 3007, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.36591169238090515, \"start\": 340, \"end\": 351}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3939986526966095, \"start\": 340, \"end\": 351}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.9392896294593811, \"start\": 340, \"end\": 351}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9392896294593811, \"start\": 340, \"end\": 351}}, {\"mention_name\": {\"text\": \"ERA5\", \"confidence\": 0.7267515063285828, \"start\": 3038, \"end\": 3042, \"clean_text\": \"ERA5\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3038, \"end\": 3042, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3038, \"end\": 3042, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3038, \"end\": 3042}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3038, \"end\": 3042, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.36591169238090515, \"start\": 382, \"end\": 386}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7267515063285828, \"start\": 382, \"end\": 386}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.9392896294593811, \"start\": 382, \"end\": 386}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9392896294593811, \"start\": 382, \"end\": 386}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [42]}}, {\"page\": 43, \"chunk\": 43, \"input_text\": \"The result is 250 independent stochastic realizations of what a 40-year precipitation time series might be. The pseudo-precipitation values generated based on ERA5 (prior to the adjustment for CHIRPS) were subsequently used to generate SWI and NDVI catalogs using empirical correlations known to exist between precipitation and these two variables to develop statistical relationships. Each stochastic realization of NDVI and SWI was thus obtained from (and consistent with) the stochastic realizations of precipitation.\\n\\n# **Validation**\\n\\nStochastic catalogs are designed to provide a view of hazard that goes beyond the data available in the historical record. In developing such catalogs, care must be exercised to avoid compromising the ability to realistically represent the physical processes governing the natural phenomena being modeled.\\n\\nSeveral diagnostic figures were generated using the rainfall catalog to show that the catalog created statistically robust stochastic realizations. Specifically, these realizations (i) preserved the intrinsic properties in terms of spatial and temporal correlations; (ii) were consistent with the empirical distributions of precipitation at each individual location; and (iii) produced a realistic view of hazard, consistent with historical observations while allowing for the creation of new extremes.\\n\\nValidation diagnostics included several comparisons:\\n\\n- The synthetic rainfall time series was compared with historical data (CHIRPS satellite rainfall data and gauge rainfall data), across different spatial aggregations; this included comparing the distribution of monthly rainfall averages for each month in the stochastic catalog (figure C.1) with the CHIRPS data set. The result shows that the stochastic catalog captures well both historical trends and an extension of the tails of the monthly distributions, suggesting the presence of previously unseen extremes and consistent with physical processes at play.\\n\\n- The cumulative distribution functions (CDFs) for historical and simulated hazards were compared to assess the relationship between frequency of an event’s occurrence and its severity.\\n\\n- The spatial distributions of rainfall averaged over various time scales were compared.\\n\\n42\", \"datasets\": [{\"mention_name\": {\"text\": \"ERA5\", \"confidence\": 0.6442366242408752, \"start\": 159, \"end\": 163, \"clean_text\": \"ERA5\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 159, \"end\": 163, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 159, \"end\": 163, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 159, \"end\": 163}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 159, \"end\": 163, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.3821199834346771, \"start\": 159, \"end\": 163}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6442366242408752, \"start\": 159, \"end\": 163}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5901792049407959, \"start\": 159, \"end\": 163}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5901792049407959, \"start\": 159, \"end\": 163}}, {\"mention_name\": {\"text\": \"rainfall catalog\", \"confidence\": 0.7838321924209595, \"start\": 899, \"end\": 915, \"clean_text\": \"rainfall catalog\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 899, \"end\": 915, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 899, \"end\": 915, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 899, \"end\": 915}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 899, \"end\": 915, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.5171722769737244, \"start\": 899, \"end\": 915}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7838321924209595, \"start\": 899, \"end\": 915}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.82927405834198, \"start\": 899, \"end\": 915}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.82927405834198, \"start\": 899, \"end\": 915}}, {\"mention_name\": {\"text\": \"CHIRPS satellite rainfall data\", \"confidence\": 0.8040305376052856, \"start\": 1477, \"end\": 1507, \"clean_text\": \"CHIRPS satellite rainfall data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1477, \"end\": 1507, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1477, \"end\": 1507, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1477, \"end\": 1507}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1477, \"end\": 1507, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.4808731973171234, \"start\": 514, \"end\": 544}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8040305376052856, \"start\": 514, \"end\": 544}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5718586444854736, \"start\": 514, \"end\": 544}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5718586444854736, \"start\": 514, \"end\": 544}}, {\"mention_name\": {\"text\": \"gauge rainfall data\", \"confidence\": 0.4734508693218231, \"start\": 1512, \"end\": 1531, \"clean_text\": \"gauge rainfall data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1512, \"end\": 1531, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1512, \"end\": 1531, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1512, \"end\": 1531}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1512, \"end\": 1531, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.4808731973171234, \"start\": 549, \"end\": 568}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4734508693218231, \"start\": 549, \"end\": 568}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5718586444854736, \"start\": 549, \"end\": 568}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5718586444854736, \"start\": 549, \"end\": 568}}, {\"mention_name\": {\"text\": \"CHIRPS data set\", \"confidence\": 0.6775749921798706, \"start\": 1706, \"end\": 1721, \"clean_text\": \"CHIRPS data set\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1706, \"end\": 1721, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1706, \"end\": 1721, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1706, \"end\": 1721}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1706, \"end\": 1721, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.46230819821357727, \"start\": 743, \"end\": 758}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6775749921798706, \"start\": 743, \"end\": 758}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7463152408599854, \"start\": 743, \"end\": 758}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7463152408599854, \"start\": 743, \"end\": 758}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [43]}}, {\"page\": 44, \"chunk\": 44, \"input_text\": \"<!-- Start of picture text --> leé Country les Coastal<br>1.0 Historical<br>_oO8 _ 3 Catalogue<br>E ie led f le3<br>= 0.6 yl +] Sad iu | 4<br>UW 2 u is al \\\\<br>fra 1 _ |e . -- | I<br>2 od . w= a - =. z<br>\\\"02i] | H !o} as I.il fe ©o i || Fl }o) as 7i<br>0.0 LT re |’ iu 0 i<br>J FMAM J | & 5 OND ) FMAM |] J AS ONO<br>Month Manth<br>Le5 Northern 1e5 Southern<br>2,0 =<br>— | 3 E Le<br>—e? i 778 t e* al ed<br>Yiode .. 2 - | 27) me el<br>a a l oa “. 1 =.<br>P05 a | cy Me a |<br>r J oJ AS i ry i J J A Ss il<br>0,0 ce ee ee 0 | ee) ere |<br>| FMAM J] | &A §S ON OD ) FMAM | J AS ONO<br>Month Month<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [44]}}, {\"page\": 45, \"chunk\": 45, \"input_text\": \"<!-- Start of picture text --> 5 1e5 Country led Coastal<br>8 Historical<br>Catalogue<br>eZ i | e ae || ;<br>:<br>;<br>|] FMAM J] | AS OND |] FMAMJ | AS ON D<br>Month Month<br>1e5 Northern 1e5 Southern<br>1.25 2.5<br>100, 7 SU 207 |b s<br>A o,75 | ws15 I =.<br>F 9.50 iy > | 210 im :<br>0,25 + oat rm, OS Thao d!<br>_ __<br>0,00 0.0<br>) FMAM | J AS OND ) FMAMJ J AS OND<br>Manth Month<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [45]}}, {\"page\": 46, \"chunk\": 46, \"input_text\": \"<!-- Start of picture text --> le? Country<br>10<br>_o9 :<br>E<br>Eas *<br>c.<br>= 07<br>ao<br>2 06<br>fe Catalogue<br>O,5 — Template<br>o.4 — Historical<br>l 2 5 lo 20 #40<br><!-- End of picture text --> <!-- Start of picture text --> led Country<br>a Catalogue<br>—— Historical<br>68<br>_a<br>a<br>3<br>2 fs<br>6<br>1 2 5 lod 20 40<br><!-- End of picture text -->\", \"datasets\": [{\"mention_name\": {\"text\": \"fe Catalogue\", \"confidence\": 0.5631834864616394, \"start\": 103, \"end\": 115, \"clean_text\": \"fe Catalogue\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 103, \"end\": 115, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 103, \"end\": 115, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 103, \"end\": 115}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 103, \"end\": 115, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.9693155288696289, \"start\": 103, \"end\": 115}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5631834864616394, \"start\": 103, \"end\": 115}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.550848662853241, \"start\": 103, \"end\": 115}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.550848662853241, \"start\": 103, \"end\": 115}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [46]}}, {\"page\": 47, \"chunk\": 47, \"input_text\": \"<!-- Start of picture text --> Catalogue — Historical<br>a 4<br>a)<br>= 3<br>2<br>o<br>ay<br>hl<br>-2<br>o . 1 % bs ts & 4 %<br>Dn a Si Si i: i a<br>time<br>2<br>f<br>- N<br>&1<br>5a9<br>ie<br>-2<br>al Pad Pas aad eal aa aw a\\\" aw<br>time<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099325301292478621/pdf/IDU1ae0eac0e145d214c6218002156b672eb8155.pdf\", \"pages\": [47]}}]}"}}},{"rowIdx":89,"cells":{"doc_id":{"kind":"string","value":"34247735"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34247735\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"authors\": {\"0\": {\"author\": \"Islam,Asif Mohammed\"}, \"1\": {\"author\": \"Gatti,Roberta V.\"}}, \"count\": \"World\", \"count_key\": \"517191\", \"docna\": {\"0\": {\"docna\": \"Dysfunctional Family Management :\\n Family-Managed Businesses and the Quality of Management Practices\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"Office of the Chief Economist (MNACE)\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34247735\"}, \"repnb\": \"WPS10684\", \"docdt\": \"2024-01-29T00:00:00Z\", \"datestored\": \"2024-01-29T04:26:07Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"Better managed firms perform better.\\n Existing evidence has shown that family-managed firms have\\n poorer management practices. Several reasons have been\\n proposed. Limiting to family members reduces the talent pool\\n of potential managers. Family management creates\\n disincentives for other talented workers given that the\\n environment is not meritocratic. Family managers themselves\\n may be less motivated given that they may not have to\\n compete for the position. This study scales up the evidence\\n by exploring the relationship between family managers and\\n management practices for about 9,000 medium and large firms\\n across 41 developing and advanced economies. The study\\n contributes to the literature by investigating several\\n internal and external operating factors that attenuate or\\n accentuate the relationship between family management and\\n the quality of management practices. The engagement of\\n governments in terms of corruption and political connections\\n is found to be influential.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10684; PEOPLE\", \"display_title\": \"Dysfunctional Family Management :\\n Family-Managed Businesses and the Quality of Management Practices\", \"disclosure_date\": \"2024-01-29T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099626001292432099/text/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.txt\", \"listing_relative_url\": \"/research/2024/01/34247735/dysfunctional-family-management-family-managed-businesses-quality-management-practices\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099626001292432099/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b\", \"new_url\": \"2024/01/34247735/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b\", \"ext_pub_date\": \"2024-01-29T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Private Sector Development m1326060 1921,Investment Climate m1326293 405,Firm Innovation, Productivity and Growth\\n m1327229 377\", \"docm_id\": \"IDU-6da65b2c-ef89-4b1a-9d5e-ac4acf945a1b\", \"chronical_docm_id\": \"IDU-6da65b2c-ef89-4b1a-9d5e-ac4acf945a1b\", \"publishtoextweb_dt\": \"2024-01-29T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10684\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"guid\": \"099626001292432099\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"36\", \"author_keys\": {\"0\": {\"author_key\": \"950662\"}, \"1\": {\"author_key\": \"321051\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099626001292432099/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10684\\n\\n# Dysfunctional Family Management\\n\\nFamily-Managed Businesses and the Quality of Management Practices _Asif M. Islam Roberta Gatti_ Middle East and North Africa Region Office of the Chief Economist January 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10684\\n\\n# **Abstract**\\n\\nBetter managed firms perform better. Existing evidence has shown that family-managed firms have poorer management practices. Several reasons have been proposed. Limiting to family members reduces the talent pool of potential managers. Family management creates disincentives for other talented workers given that the environment is not meritocratic. Family managers themselves may be less motivated given that they may not have to compete for the position. This study scales up the evidence by exploring the relationship between family managers and management practices for about 9,000 medium and large firms across 41 developing and advanced economies. The study contributes to the literature by investigating several internal and external operating factors that attenuate or accentuate the relationship between family management and the quality of management practices. The engagement of governments in terms of corruption and political connections is found to be influential.\\n\\nThis paper is a product of the Office of the Chief Economist, Middle East and North Africa Region. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at aislam@worldbank.org.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# **Dysfunctional Family Management: Family-Managed Businesses and the Quality of Management Practices**\\n\\nAsif M. Islam and Roberta Gatti<sup>1</sup> Keywords: Management Practices, Middle East and North Africa, Government Ownership JEL codes: L20, L22, L23, O12, M10, G32\\n\\n> 1 Roberta Gatti is the Chief Economist of the Middle East and North Africa region at the World Bank. Asif M. Islam is a Senior Economist in the Chief Economist Office of the Middle East and North Africa at the World Bank. The opinions expressed in this paper do not represent the views of the World Bank Group, its Board of Directors, or the Governments they represent. All errors and omissions are the authors’ responsibility.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **Dysfunctional Family Management: Family-Managed Businesses and the Quality of Management Practices**\\n\\n## **I. Introduction**\\n\\nFirms are a crucial engine of economic growth. How these businesses are managed will determine how they perform. The seminal work by Bloom and Van Reenen (2007) initiated a large literature that documented the positive effects of good management practices. Firms that are better managed are more productive, have higher operating profits, are more outward oriented, and invest more in research and development (Bruhn et al., 2018; Bloom et al., 2019; see Scur et al., 2021 for a summary). The productivity of skilled labor may also increase (Gosnell et al., 2020). The firms are also more likely to train their workers (Islam and Gatti, 2023). Experimental research has validated these findings with rigorous identification strategies (Bloom et al., 2013a; Bloom et al., 2020).\\n\\nFamily managed firms are expected to have poorer management practices (see Tsoutsoura, 2021 for a review of the recent literature). There are number of reasons for this.<sup>2</sup> First, choosing managers only within the family restricts the talent pool of potential managers, increasing the chances that family-managed firms will miss out on better quality managers (Bennedsen et al., 2007; Bloom and Van Reenen, 2007). Second, family members that are expected to become managers of the family firm may have disincentives to work hard given that they will attain the position regardless of effort. Thus, the lack of competition creates disincentives to excel. Villalonga and Amit (2006) find that in descendant-CEO firms - where the CEO is from the family but is not the founder - the conflict between family and nonfamily shareholder firms is more costly than the owner-manager conflict in nonfamily firms. Third, other managers or workers may view this form of nepotism unfavorably, thereby discouraging their effort, and creating incentives for them to work elsewhere. There is also the possibility of family conflicts playing out, that could demotivate employees. There could be reasons to expect the opposite – family management could resolve the classic agency problem between owners and managers (Jensen and Meckling, 1976; Fama, 1980). Owners may find it easier to monitor family members who are managers, and thus the family CEOs are unlikely to deviate from the principal’s objectives. However, the existing empirical evidence has largely pointed to family-managed businesses having poorer management practices. Bloom et al (2015) show that family-run\\n\\n> 2 https://hbr.org/2011/03/family-firms-need-professional.\\n\\n2\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"firms do tend to have poorer management practices for a sample of 10,000 public and private sector manufacturing firms across 35 economies. Lemos and Scur (2019) also find a similar relationship in a sample of 2,710 firms across 18 economies and implement an instrumental variable identification strategy for 912 firms.\\n\\nWe build on this literature by using unique data from the 2019 EBRD-WIB-World Bank Enterprise Surveys (WBES). This data has a special module on family-managed firms and political connections on top of the comprehensive information on internal and external operating environment of businesses found in the regular World Bank Enterprise Surveys. These surveys were rolled out across a similar timeframe (20182019) for developing and advanced economies across Europe, the Middle East and North Africa, and Mongolia. The module defines our sample and enables this study to make a unique contribution by examining the relationship between family managers and management practices for a set of economies, many of which have not been explored before. This study additionally contributes to the literature by leveraging the surveys to uncover firm characteristics and business environment factors that may strengthen or weaken the effects of family management on management practices.\\n\\nFor around 9,000 medium and large firms across manufacturing and service sectors in 41 countries, we largely confirm the findings of Bloom et al (2015) that family management is negatively correlated with the quality of management practices. This is an important finding given the differences of our sample – we explore both manufacturing and service firms, while Bloom et al (2015) only study manufacturing firms. Our sample also consists of firms in Eastern Europe and the Middle East and North Africa that are largely omitted in the Bloom (2015) sample.\\n\\nThe extensive nature of the WBES allows us to enrich the analysis in several ways. First, the surveys collected information on the share of managers that are family, providing some nuance on whether any family management or all family management is detrimental to management practices. We find evidence towards the latter. Second, the surveys allow for the exploration of whether certain factors either exacerbate or alleviate the negative relationship between family managers and the quality of management practices. These include a range of internal and external operating factors. Third, the surveys allow us to explore specifically what types of management practices are lacking in family-run businesses. Finally, the surveys provide information that describes the nature of family run businesses.\\n\\nThis study uncovers the following key findings. First, family management is negatively correlated with management practices, more so if all managers in the firm are from the same family. This is true in the 3\", \"datasets\": [{\"mention_name\": {\"text\": \"EBRD-WIB-World Bank Enterprise Surveys\", \"confidence\": 0.6262863278388977, \"start\": 384, \"end\": 422, \"clean_text\": \"EBRD-WIB-World Bank Enterprise Surveys\"}, \"acronym\": {\"text\": \"WBES\", \"confidence\": 0.6243708729743958, \"start\": 424, \"end\": 428, \"clean_text\": \"WBES\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 384, \"end\": 422, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 384, \"end\": 422}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 384, \"end\": 422, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9974775910377502, \"start\": 384, \"end\": 422}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6262863278388977, \"start\": 384, \"end\": 422}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5225202441215515, \"start\": 384, \"end\": 422}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5225202441215515, \"start\": 384, \"end\": 422}}, {\"mention_name\": {\"text\": \"World Bank Enterprise Surveys\", \"confidence\": 0.6103050708770752, \"start\": 634, \"end\": 663, \"clean_text\": \"World Bank Enterprise Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 634, \"end\": 663, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 634, \"end\": 663, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 634, \"end\": 663}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 634, \"end\": 663, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9904559254646301, \"start\": 634, \"end\": 663}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6103050708770752, \"start\": 634, \"end\": 663}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.6816174387931824, \"start\": 634, \"end\": 663}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6816174387931824, \"start\": 634, \"end\": 663}}, {\"mention_name\": {\"text\": \"surveys\", \"confidence\": 0.3000394105911255, \"start\": 1137, \"end\": 1144, \"clean_text\": \"surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1137, \"end\": 1144, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1137, \"end\": 1144, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1137, \"end\": 1144}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1137, \"end\": 1144, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9642007350921631, \"start\": 261, \"end\": 268}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3000394105911255, \"start\": 261, \"end\": 268}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.6433517932891846, \"start\": 261, \"end\": 268}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6433517932891846, \"start\": 261, \"end\": 268}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"sample regardless of income classification or regional classification. Second, bribery and political connections tend to exacerbate the negative effects of family management on the quality of management practices. This is likely because bribery tends to disincentivize good management practices (Athanasouli and Goujard, 2015), compounding the negative effects of family managers on the quality of management practices. The weaker the competitive forces, the less pressure family managers face to improve performance and adopt good management practices. Political connections accentuate the negative relationship between family management and management practices, possibly highlighting the attenuation of competitive forces to increase management practices due to political connections. Similarly, the findings show that lack of perceived competitors -serving as a proxy of competition – also accentuates the negative effect of family management on the quality of management practices.\\n\\nOne can also expect that if the pool of talented managers is restricted for all firms, then the negative effects of family management on management practices may be ameliorated as all firms draw from a limited talent pool. We proxy for this using the variable that captures whether firms perceive labor regulations to be a major or severe obstacle to operations. The more stringent the labor regulations, the greater the difficulty in hiring and firing, imposing constraints on the potential of hiring talented managers. Thus, we find that for firms that find labor regulations burdensome, the relationship between family management and quality of management practices is alleviated. There is also some evidence that innovative firms are less likely to be hampered by family managers. It is also worth noting that family ownership, firm size, exporter status, foreign ownership, manufacturing sector, and female managers do not play any significant role (neither exacerbate nor attenuate) in the relationship between family management and management practices. Finally, when all managers are from the same family, these firms perform poorly in the management practice dimensions of the number of performance indicators monitored, knowledge of production or service provision targets, and not surprisingly, the basis of bonuses.\\n\\nIn summary, this study contributes to the literature in several ways. It complements analysis in the literature by investigating the relationship between family management and quality of management practices for a sample that has not been explored by previous studies. It then extends the analysis by showing how certain key factors, including corruption, competition, and political connections, accentuate, while stringent regulations and innovation ameliorate this relationship. It also provides some nuance – showing that a mix of family and non-family managers may not be as detrimental to management practices as having all management be family members. The rest of the paper is structured as follows. Section II describes the data. Section III explains the empirical strategy. Section IV presents the results, and Section V concludes.\\n\\n4\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"# **II. Data**\\n\\nThe main source of firm-level data is the EBRD-EIB-World Bank’s Enterprise Surveys (WBES) conducted around 2018-2019. The choice of this particular set of WBES is because they contain questions on family management that are not available for the regular Enterprise Surveys carried out by the World Bank. This set of surveys were a joint venture between the European Bank for Reconstruction and Development (EBRD), the European Investment Bank (EIB) and the World Bank (see EIB-EBRD-WB 2022). Data is available for 41 economies encompassing developing and advanced economies in Europe, the Middle East and North Africa (MENA) region, and Mongolia. The surveys have a section on management practices that was only implemented for medium and large firms (those with at least 20 permanent, full-time employees), where management practices were more likely to matter. The surveys also contain the standard World Bank Enterprise Surveys modules, collecting information on a representative sample of formal (registered) private firms operating in manufacturing or service sectors. The WBES data are fully comparable across countries and are collected via face-to-face interviews with business owners or top managers by using a global methodology. The data have been widely used by several studies to explore the private sector in developing economies (Paunov, 2016; Besley and Mueller, 2018; Chauvet and Ehrhar, 2018; Hjort and Poulsen, 2019). One advantage of this round of surveys is that they consist of a set of economies surveyed around a similar timeframe, employing a consistent survey instrument and methodology. These surveys were largely completed before the COVID-19 pandemic outbreak.\\n\\nThe key outcome variable is the quality of management practices, following the methodology implemented by Bloom et al. (2013b). This consists of eight components: (i) Problem resolution, (ii) Number of performance indicators measured, (iii) Level of ease or difficulty to achieve production or service provision targets, (iv) Knowledge of production or service provision targets, (v) Basis of manager bonuses, (vi) Length of focus of production targets, (vii) Promotion of non-managers, and (viii) Dismissal of underperforming managers. The scoring for each component is provided in table A4. The management practices module is only implemented for medium and large firms. Figure 1 provides the management practices scores for the economies in the sample.\\n\\nThe main variable of interest is the presence of family members in management. The specific question asked in the survey instrument is “ _What percentage of the key management positions of this firm are occupied by members of this family?_ ” We use two variables in the analysis. First, we use this exact variable to capture the intensity of family management. Second, we capture firms where all management is family.\\n\\n5\", \"datasets\": [{\"mention_name\": {\"text\": \"WBES data\", \"confidence\": 0.4157491624355316, \"start\": 1094, \"end\": 1103, \"clean_text\": \"WBES data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.5523301362991333, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1094, \"end\": 1103, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1094, \"end\": 1103}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1094, \"end\": 1103, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.49277979135513306, \"start\": 294, \"end\": 303}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4157491624355316, \"start\": 294, \"end\": 303}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8864100575447083, \"start\": 294, \"end\": 303}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8864100575447083, \"start\": 294, \"end\": 303}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"This is achieved through a binary variable that attaints a value of 1 if the response to the question is 100 percent, and 0 otherwise. Using both variables allows the analysis to make a distinction between firms that are entirely managed by family, and those that are partially managed by family. The average share of family managers is about 40.7 percent. Around 32 percent of medium and large businesses in the sample are entirely managed by family (Table 1). Figure 2 provides the country averages for the percentage of firms that are entirely family managed. The share of family management for each country is presented in figure 3.\\n\\nData for the control variables are also obtained from the Enterprise Surveys. These encompass standard firm-level characteristics and the operating environment including firm size, age, outward orientation, quality certification, access to finance, manager experience in the sector, and perceptions of labor regulations as a constraint. Whether the firm is owned by the same family, and the presence of political connections are also included as control variables and are not collected in the standard ES but were part of the extended questions in this round of the surveys. The specific information on family ownership is obtained from the survey question: “ _What percentage of the firm is owned by the same family? (If more than one family, refer to the one with largest ownership_ ).” The information on political connections is obtained from the following survey question: “ _Has the owner, CEO, top manager, or any of the board members of this firm ever been elected or appointed to a political position in this country?_ ” The rationale for the control variables is detailed in the empirical strategy section. Summary statistics are provided in table 1.\\n\\nTable A1 provides some description of the nature of family-managed firms. Entirely family-managed firms have lower management scores than other firms. About 81 percent of fully family-managed firms are of medium size, compared to 72 percent for other firms. Around 12 percent of fully family-managed firms are large, compared to 24 percent of other firms. Entirely family-managed firms are also less likely to be foreign-owned (5.53 vs. 13.89 percent), and more likely to have female managers (17.57 vs 13.50 percent). Furthermore, fully family-managed firms have a larger share of family owners (96 vs 32 percent) and have more managerial experience (24 vs 20 years). They are however less likely to be politically connected (5 vs 9 percent). These differences are statistically significant at least at the 5 percent level. There are no statistically significant differences between entirely family-managed firms and the rest with regards to exporting status, access to finance (in terms of loans or bank accounts), firm age, quality certification, perceptions of labor regulations as a major or severe constraint, innovation, bribery, whether the firm started formally or whether they have one or no competitor.\\n\\n6\", \"datasets\": [{\"mention_name\": {\"text\": \"Enterprise Surveys\", \"confidence\": 0.5490940809249878, \"start\": 696, \"end\": 714, \"clean_text\": \"Enterprise Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 696, \"end\": 714, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 696, \"end\": 714, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 696, \"end\": 714}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 696, \"end\": 714, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9477112293243408, \"start\": 696, \"end\": 714}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5490940809249878, \"start\": 696, \"end\": 714}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.7132918834686279, \"start\": 696, \"end\": 714}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7132918834686279, \"start\": 696, \"end\": 714}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"A set of extended questions in the survey explore how managers of large firms use their time. We leverage this information to explore whether family managers behave differently from non-family managers. Manager time allocation is captured in terms of how frequently they interact with other decision makers in the organization (COO, CAO, board members), suppliers, and employees. Frequency of engagement is recorded in five buckets: (i) Never, (ii) Once a week, (iii) Between 2 and 4 times a week, (iv) Daily, and (v) More than once a day. Table A2 shows that that there is no statistically significant difference between family managers and non-family managers across the five buckets along the dimensions of other decision makers, suppliers, and employees. This suggests that the difference in the quality of management practices is not due to differential levels of engagement across the organizations between family managers and various actors. Or at least, family managers of large firms appear to allocate their time similarly to those of non-family managers. Note that this sample is not exactly comparable with the previous results as it only pertains to large firms in the sample of analysis. However, there is a negative relationship between family management and management practices for this sample of large firms.<sup>3</sup>\\n\\n# **III. Empirical Strategy**\\n\\nThe following equation is estimated for the pooled cross-section sample using Ordinary Least Squares (OLS). M M 0 1F F 2F F F 3S S 4 5M M ii 6P ii 7M ii 1 ii (1)ii ii = ββ + ββ MMFF + ββ FF + ββ + ββ AAFFSS + ββ FFMMFFSSM ii ii zz ii cc ii + ββ FFP FFP + ββ FFMM + ββ ZZ + δδ DD + εε Where M M is the average management practices score. The family management variable (F F ) is either (i) the share of key managers that are family members or (ii) a binary variable equal to 1 if all key MMFF managers are family, and zero otherwise. To control for as many confounding factors as possible, several firm-level variables are accounted for. These include the share of family owners (F F F ), firm size as measured by the number of full-time employees (S S ), firm age ( ), manager experience in the same FF sector (M M ), political connections in terms of whether the owner, CEO, top manager, or any of AAFFSS the board members of the firm have ever been elected or appointed to a political position in their country FFMMFFSSM (P ). We also account for whether the sector of activity is in the manufacturing sector (M ). FFP FFP FFMM\\n\\n> 3 The findings are available upon request.\\n\\n7\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"Other control variables ( include whether the firm is an exporter (defined as firms with 10% or more of sales directly exported), is foreign owned (defined as firms with 10% or more private foreign ownership), ZZ) the presence of a checking or savings account, quality certification, bribery, competition, and whether the firm perceives labor regulations to be a major or severe constraint to operations. Country fixed effects ( ) are included to account for time invariant country-specific omitted variables. is the standard error term cc DD with the usual desirable properties. Survey weights are used, and the standard errors are clustered at the ii εε location-sector-size strata level.\\n\\nThe identification strategy is to exploit cross-sectional variation in family management and the quality of management practices to establish the relationship between the two. The assignment of family management to firms may not be random, thereby raising endogeneity concerns. While one cannot completely rule out the possibility of simultaneity bias, it is unlikely that the presence of family management is determined by the quality of management practices. A greater concern is omitted variable bias that may be correlated with family management and management practices. To address this, we account for as many possible control variables as possible. An important determinant of the quality of management practices is the degree of competition the firm is exposed to. These can come in the form of foreign direct investment and trade that increase exposure to competition and thereby the quality of management practices (Bloom et al., 2016). These are proxied by exporter status and foreign ownership in the estimations. We account for family ownership as it is likely to influence the quality of management practices as the ownership structure will determine the performance incentives in the firm (Tsoutsoura, 2021; Bloom et al., 2015). Family ownership could resolve principal-agent problems if the manager is a family member (Villalonga and Amit, 2006; Burkart and Panunzi, 2006). Family-owned firms may also facilitate labor contracts as ownership tends to be concentrated and thus more credible and less likely to change due to the low probability of hostile takeovers (Shleifer and Summers, 1988; Mueler and Philippon, 2011; Sraer and Thesmar, 2007; Ellul et al., 2018). On the other hand, family-owned firms may extract private benefits at the cost of other shareholders (Shleifer and Vishny, 1986). Family-owned firms may also be less susceptible to hostile takeovers, and thus have fewer incentives to maximize value (Fama, 1980). This may be reflected in the adoption of poor management practices.\\n\\nThe presence of political connections and corruption can disincentivize firm performance by lessening competitive forces (Rijkers et al., 2017; Athanasouli and Goujard, 2015). This may in turn distort incentives to adopt good management practices. Firms may invest in quality certification to increase the returns to innovation (Paunov, 2016). Thus, quality certification may also encourage certain managerial behaviors in 8\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"the firm, and thereby encourage good management practices. Rigid labor laws can create obstacles to the adoption of management practices with regards to people management (Bloom et al., 2019).\\n\\n# **IV. Results**\\n\\nTable 2 provides the main estimation results. The coefficient of the share of family management is negatively correlated with the quality of management practices (column 1), statistically significant at the 5 percent level. In column 2, we replicate the same estimation using full family management instead of the share of family management. The coefficient is similar, but the statistical significance increases to the 1 percent level. This implies that the relationship may be stronger for firms where all management is family versus firms that have partial family management. The other covariates behave largely as expected. Exporter status and foreign ownership are positively correlated with good management practices. Larger firms have better management practices, while older firms do not. Access to finance, in terms of having a loan, and ISO certification are also positively correlated with good management practices. All these findings are statistically significant at least at the 5 percent level. Perception of labor market stringency is negatively correlated with good management practices – the more a firm perceives stringent labor regulations as an obstacle, the lower the management practices score. However, this relationship is not statistically significant. The coefficient of political connections is also negative, but not statistically significant.\\n\\nThe sample of analysis can be largely split into three regional groups: (i) developing MENA, (ii) Eastern Europe and Central Asia, and (ii) Western Europe that only includes four countries – Italy, Portugal, Greece, and Cyprus. In table 3, we explore whether the relationship between family management and the quality of management practices stands for each of these subgroups. There is a negative correlation between the share of family management and the quality of management practices for each of these regional subgroups, although the coefficient is only statistically significant for developing MENA and Eastern Europe and Central Asia, but not for Western Europe. However, for full family management - all key management positions are held by family - the coefficient is negative and statistically significant at least at the 5 percent level for all three regional subgroups (columns 4-6, table 3). The findings suggest that for the four Western European economies, only full family management is detrimental for the quality of management practices. The coefficient of political connections is negative and only statistically significant for the Western Europe subgroup of four economies, possibly suggesting that political connections might be more detrimental for management practices for Western Economies than in the rest of the world.\\n\\n9\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"In table 4 we explore the relationship by income groups. Using World Bank income classifications, economies are classified into (i) low- and middle-income economies and (ii) high income economies. We find that the share of management practices is negatively correlated with the quality of management practices, although this is statistically insignificant for high income economies, and barely statistically significant for low- and middle-income economies (10 percent level). However, the coefficient for fully family-managed firms is negative and statistically significant at least at the 5 percent level across the income groups. Consistent with the other findings, results are stronger when we look at all management being family versus partial family management.\\n\\nThe nature by which internal operating environments and external business environments interact with family management may accentuate or debilitate the correlation with management practices. We explore interactions with the government in particular. We look at two types of interactions in terms of (i) political connections and (ii) corruption (bribes). Results are presented in table 5. The coefficient of the interaction between all family management and political connections is negative, but statistically insignificant. The coefficient of the interaction between partial family management and political connections is negative and statistically significant at the 10 percent level. Thus, political connections, by further eroding exposure to competitive forces, may make family management even weaker in terms of management practices. Similar findings are obtained for _bribes_ . _Bribes_ is a binary variable that attains a value of 1 if firms are expected to make informal payments to get things done, and zero otherwise. The coefficient of the interaction between bribes and family management is negative and statistically significant at the 5 percent level, regardless of whether the firm has full or partial family management. Thus, both forms of government engagement – political connections and bribes – seem to exacerbate the negative relationship between family management and the quality of management practices.\\n\\nIn table 6 we explore the interaction between family management and domestic competition, and also perceptions of stringency of labor regulations. Domestic competition is measured by a binary variable that takes a value of 1 if a firm claims that it has one competitor or no competition in the main domestic market. The coefficient of the interaction between family management and domestic competition is negative and statistically significant at the 5 percent level, regardless of whether full or partial family management is considered (columns 1 and 2 of table 6). Therefore, when family-managed firms face less competition in terms of fewer competitors, they have fewer reasons to innovate and adopt better management practices. Similarly, we explore whether perceptions of labor regulations being a severe or major obstacle matter for the relationship between family management and the quality of management practices. Given that one of 10\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"the reasons family management leads to poorer management practices as there is a smaller talent pool to draw from, stringent labor regulations may mean that even firms without family managers face a limited talent pool. That is what we see in columns 5 and 6 of table 6. The coefficient of the interaction between family management and perceptions of labor regulations being a major or severe obstacle to operations is positive, and statistically at the 1 percent level, regardless of full or partial family management. Thus, the negative relationship between family management and quality of management practices is weakened as outsider managed firms also face restrictions in terms of managers that could be hired. We also find that if businesses are innovative – in terms of introducing processes – then this ameliorates the negative relationship between family management and management practices. The implication could be that if a business is in an innovative sector (or innovative by nature), the high degree of competition in such sectors exerts competitive forces that curtail the disincentive for competition brought about by family management.\\n\\nWhile we find significant differences between family managed and non-family managed firms for the factors discussed above, it is worth noting that there are several dimensions by which there are no statistically significant differences. In table A3 we show that the coefficient of the interactions between family management and family ownership, firm size, exporter status, foreign ownership, manufacturing sector, and female managers are statistically insignificant. We do find that firm age may further worsen the negative relationship between family management and quality of management practices, but this is barely statistically significant at the 10 percent level. Table A3 presents findings for full family management, but the results are the same for partial family management with one exception - the interaction between firm age and partial family management is statistically insignificant. Thus, external factors rather an internal firm characteristics play an important role in shaping the relationship between family management and management practices.\\n\\nIn table 7 we explore whether family management is negatively correlated with specific management quality sub-scores. Entirely family-managed firms are negatively correlated with all the management practices sub-scores, but the coefficient is only statistically significant for the sub-scores of (i) Number of performance indicators monitored, (ii) Knowledge of production or service provision targets, and (iii) Basis of bonuses. Thus, firms that are entirely family-managed, perform poorly with regards to monitoring of indicators, sharing of information on targets, and the basis of bonuses. The findings for partially familymanaged firms are similar, but not statistically significant for the basis of bonuses. However, partial family management is negatively correlated with the promotion of non-managers – statistically significant at the 10 percent level. These findings suggest that family managers do not monitor specific indicators and targets. This also affects bonuses which ideally should be based on performance targets.\\n\\n11\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"# **V. Conclusion**\\n\\nThis study established a negative relationship between family management and the quality of management practices, leveraging a unique dataset using a special module spanning 41 developing and advanced economies. The results confirm what has been found in the literature, although the sample in this study is of a different nature, including many firms in regions that have not been covered elsewhere. The study extends the literature by examining how external factors could attenuate or exacerbate the negative relationship between management practices. Corruption, competition, and political connections accentuate, while stringent regulations and innovation ameliorate the relationship between family management and the quality of management practices. Furthermore, when all management is from the same family, these firms perform poorly on the management practices dimensions of the number of performance indicators monitored, knowledge of production or service provision targets, and not surprisingly, the basis of promotion. Finally, a mix of family and non-family managers may not be as detrimental to management practices as having all management be family members.\\n\\nWhat is also established is the lack of statistical significance in certain relationships. Family ownership, firm size, exporter status, foreign ownership, manufacturing sector, and female managers do not play any significant role (neither exacerbate nor attenuate) in the relationship between family management and management practices. Furthermore, for larger firms, family managers do not exhibit any statistically significant difference compared with non-family managers with regards to how they allocate (or appear to allocate) their time to interact with other decision makers, suppliers, and employees as measured by frequency of meetings. This information is leveraged from a special set of questions on time use for large firms in the sample.\\n\\nWhile the study makes significant strides in the literature, it is also important to note the limitations. Endogeneity concerns are difficult to alleviate given the cross-sectional nature of the data. Furthermore, the survey data do not contain information on whether the family owners are the founders, or descendants of the founders, which has been found to be important in the literature (Villalonga and Amit, 2006). Regardless, this study makes an important contribution by highlighting the role that the external operating environment can have in the relationship between family management and management practices. The key implication is that the business environment matters, and for some cases it may be easier to address some of these concerns than to change the management or ownership structure of family firms.\\n\\n12\", \"datasets\": [{\"mention_name\": {\"text\": \"survey data\", \"confidence\": 0.9291301369667053, \"start\": 2166, \"end\": 2177, \"clean_text\": \"survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2166, \"end\": 2177, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2166, \"end\": 2177, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2166, \"end\": 2177}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2166, \"end\": 2177, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9812955260276794, \"start\": 304, \"end\": 315}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9291301369667053, \"start\": 304, \"end\": 315}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.45709753036499023, \"start\": 304, \"end\": 315}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.45709753036499023, \"start\": 304, \"end\": 315}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"# **References**\\n\\nAthanasouli, Daphne and Antoine Goujard (2015) “Corruption and Management Practices: Firm Level Evidence.” _Journal of Comparative Economics_ 43: 1014-1034.\\n\\nBennedsen Morten, Kasper Meisner Nielsen, Francisco Perez-Gonzalez, and Daniel Wolfenzon (2007), “Inside the Family Firm: The Role of Families in Succession Decisions and Performance.” _The Quarterly Journal of Economics_ 122(2): 647–691.\\n\\nBesley, Timothy and Hannes Mueller (2018) “Predation, Protection, and Productivity: A Firm-Level Perspective.” _American Economic Journal: Macroeconomics_ 10(2): 184-221.\\n\\nBloom, Nicholas and John Van Reenen (2007) “Measuring and Explaining Management Practices Across Firms and Countries.” _The Quarterly Journal of Economics_ 122(4):1351–1408.\\n\\nBloom, Nicholas, Benn Eifert, Aprajit Mahajan, David McKenzie, and John Roberts (2013a). “Does Management Matter? Evidence from India.” _The Quarterly Journal of Economics_ 128(1):1–51.\\n\\nBloom, Nicholas, Raffaella Sadun, and John Van Reenen (2015). “Do Private Equity Owned Firms have Better Management Practices?” _American Economics Review: Papers & Proceedings_ 105(5):442-446.\\n\\nBloom, Nicholas, Mirko Draca and John Van Reenen (2016). “Trade Induced Technical Change? The Impact of Chinese Imports on Innovation, IT and Productivity.” _The Review of Economic Studies_ 83 (1):87–117 Bloom, Nicholas, Erik Brynjolfsson, Lucia Foster, Ron Jarmin, Megha Patnaik, Itay Saporta-Eksten, and John Van Reenen (2019). “What Drives Differences in Management Practices?” _American Economic Review_ 109(5):1648–83.\\n\\nBloom, Nicholas, Aprajit Mahajan, David McKenzie, and John Roberts (2020). “Do management Interventions Last? Evidence from India.” _American Economic Journal: Applied Economics_ 12(2):198– 219.\\n\\nBruhn, Miriam, Dean Karlan, and Antoinette Schoar (2018). “The Impact of Consulting Services on Small and Medium Enterprises: Evidence from a Randomized Trial in Mexico.” _Journal of Political Economy_ 126(2): 635-687.\\n\\nBurkart, Mike and Fausto Panunzi (2006). “Agency conflicts, ownership concentration, and legal shareholder protection.” _Journal of Financial Intermediation_ 15(1): 1-31.\\n\\nChauvet, Lisa and Helene Ehrhart (2018) “Aid and Growth: Evidence from Firm-Level Data.” _Journal of Development Economics_ 135: 461-477.\\n\\nEBRD-EIB-WB (2022). Unlocking Sustainable Private Sector Growth in the Middle East and North Africa: Evidence from the Enterprise Survey. © Luxembourg City: European Investment Bank; London: European Bank for Reconstruction and Development; Washington, DC: World Bank.\\n\\nEllul, Andrew Marco Pagano, and Fabiano Schivardi (2018). “Employment and Wage Insurance within Firms: Worldwide Evidence.” _The Review of Financial Studies_ , 31(4): 1298–1340, 13\", \"datasets\": [{\"mention_name\": {\"text\": \"Firm-Level Data\", \"confidence\": 0.811133861541748, \"start\": 2229, \"end\": 2244, \"clean_text\": \"Firm-Level Data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2229, \"end\": 2244, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2229, \"end\": 2244, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2229, \"end\": 2244}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2229, \"end\": 2244, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.33274003863334656, \"start\": 798, \"end\": 813}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.811133861541748, \"start\": 798, \"end\": 813}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.47165390849113464, \"start\": 798, \"end\": 813}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.47165390849113464, \"start\": 798, \"end\": 813}}, {\"mention_name\": {\"text\": \"Enterprise Survey\", \"confidence\": 0.7512475848197937, \"start\": 2416, \"end\": 2433, \"clean_text\": \"Enterprise Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2416, \"end\": 2433, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2416, \"end\": 2433, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2416, \"end\": 2433}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2416, \"end\": 2433, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9906994700431824, \"start\": 268, \"end\": 285}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7512475848197937, \"start\": 268, \"end\": 285}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7907562255859375, \"start\": 268, \"end\": 285}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7907562255859375, \"start\": 268, \"end\": 285}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"Fama, Eugene F. (1980). “Agency Problems and the Theory of the Firm.” _Journal of Political Economy_ 88(2): 288-307.\\n\\nGosnell, Greer K., John A. List, Robert D. Metcalfe (2020). “The Impact of Management Practices on Employee Productivity: A Field Experiment with Airline Captains.” _Journal of Political Economy_ 128(4):1195-1232.\\n\\nHjort, Jonas, and Jonas Poulsen (2019) “The Arrival of Fast Internet and Employment in Africa.” _American Economic Review_ , 109 (3): 1032-79.\\n\\nIslam, Asif M. and Roberta Gatti (2023). “The Human Capital of Firms and Formal Training.” _IZA Journal of Development and Migration forthcoming_ Jensen, Michael C. and William H. Meckling (1976). “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership.” _Journal of Financial Economics_ 3(4): 305-360.\\n\\nLemos, Renata, and Daniela Scur (2019). “The Ties that Bind: Implicit Contracts and the Adoption of Management Technology in the Firm.” CEPR Discussion Paper 12 Mueller, Holger M., and Thomas Philippon (2011). “Family Firms and Labor Relations.” _American Economics Journal: Macroeconomics_ 3(2):218-245.\\n\\nPaunov, Caroline (2016) “Corruption's Asymmetric Impacts on Firm Innovation.” _Journal of Development Economics_ 118: 216-231 Rijkers, Bob, Caroline Freund, and Antonio Nucifora (2017). “All in the Family: State Capture in Tunisia.” _Journal of Development Economics_ 124: 41-59.\\n\\nScur, Daniela, Rafaella Sadun, John Van Reenen, Renata Lemos, and Nicholas Bloom (2021). “World Management Survey at 18: Lessons and the Way Forward.” IZA Discussion Papers No. 14146.\\n\\nShleifer, Andrei and Lawrence H. Summers (1988). “Breach of Trust in Hostile Takeovers’,” in _Corporate Takeovers: Causes and Consequences_ , Chicago IL, University of Chicago Press, 33-68.\\n\\nShleifer, Andrei and Robert W. Vishny (1986). “Large Shareholders and Corporate Control.” _Journal of Political Economy_ 94: 461-488 Sraer, David and David Thesmar (2007). “Performance and Behavior of Family Firms: Evidence from the French Stock Market.” _Journal of the European Economic Association_ 5(4): 709-751.\\n\\nTsoutsoura, Margarita (2021) “Family Firms and Management Practices.” _Oxford Review of Economic Policy_ 37(3): 323-334.\\n\\nVillalonga, Belen and Raphael Amit (2006). “How do Family Ownership, Control, and Management Affect Firm Value?” _Journal of Financial Economics_ 80(2): 385-417.\\n\\n14\", \"datasets\": [{\"mention_name\": {\"text\": \"World Management Survey\", \"confidence\": 0.5297600030899048, \"start\": 1471, \"end\": 1494, \"clean_text\": \"World Management Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1471, \"end\": 1494, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1471, \"end\": 1494, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1471, \"end\": 1494}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1471, \"end\": 1494, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.999667763710022, \"start\": 816, \"end\": 839}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5297600030899048, \"start\": 816, \"end\": 839}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8313037753105164, \"start\": 816, \"end\": 839}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8313037753105164, \"start\": 816, \"end\": 839}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"# **Table 1: Summary Statistics**\\n\\n||||**Std.**|||\\n|---|---|---|---|---|---|\\n|**Variable**|**Obs**|**Mean **|**dev.**|**Min**|**Max **|\\n|Overall Management Score|9,015|0.515|0.208|0|1|\\n|Share of family management (0 to 1)|9,015|0.407|0.450|0|1|\\n|All management is family Y/N|9,015|0.319|0.466|0|1|\\n|Share of same family ownership (0 to 1)|9,015|0.527|0.464|0|1|\\n|Owner, Manager or Board Political Connection Y/N|9,015|0.076|0.264|0|1|\\n|Female top manager Y/N|9,015|0.148|0.355|0|1|\\n|Firm Formally Registered when Started Operations|9,015|0.959|0.199|0|1|\\n|Top manager experience in sector (years)|9,015|21.296|11.364|1|60|\\n|Log of age of firm|9,015|2.851|0.686|0|5.088|\\n|Log of size|9,015|3.890|0.917|0|11.067|\\n|Direct exports 10% or more of sales Y/N|9,015|0.264|0.441|0|1|\\n|Foreign ownership Y/N|9,015|0.112|0.316|0|1|\\n|Establishment has checking or savings account Y/N|9,015|0.974|0.160|0|1|\\n|Establishment has a line of credit or loan Y/N|9,015|0.476|0.499|0|1|\\n|ISO Certification Ownership Y/N|9,015|0.327|0.469|0|1|\\n|Firm identifying labor regulations as a major or severe constraint Y/N|9,015|0.094|0.292|0|1|\\n|Manufacturing Sector Y/N|9,015|0.393|0.488|0|1|\\n|Firm expected to make payment to get things done Y/N|8,493|0.111|0.314|0|1|\\n|No or One competitor (Y/N) (domestic)|6,856|0.062|0.241|0|1|\\n|Introduced a process innovation Y/N|8,954|0.230|0.421|0|1|\\n|MG1 Problem resolution|9,015|0.707|0.278|0|1|\\n|MG2 Number of performance indicators monitored|9,015|0.428|0.342|0|1|\\n|MG3 Level of ease or difficulty to achieve production or service provision<br>targets|9,015|0.535|0.377|0|1|\\n|MG4 Knowledge of production or service provision targets|9,015|0.353|0.385|0|1|\\n|MG5 Basis of bonuses|9,015|0.380|0.398|0|1|\\n|MG6 Length of focus of production targets|9,015|0.573|0.404|0|1|\\n|MG7 Promotion of non-mangers|9,015|0.729|0.399|0|1|\\n|MG8Dismissal|9,015|0.412|0.459|0|1|\\n\\n15\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"**Table 2: Family Management and Management Practices**\\n\\n|**Model**|**OLS with Country**|**FE (Cross-section)**|\\n|---|---|---|\\n|**Outcome Variable**|**Overall Mana**|**gement Score**|\\n||**(1)**|**(2)**|\\n|Share of family management (0 to 1)|-0.044**||\\n||(0.018)||\\n|All management is family Y/N||-0.045***|\\n|||(0.013)|\\n|Share of same family ownership (0 to 1)|0.017|0.014|\\n||(0.018)|(0.015)|\\n|Owner, Manager or Board Political Connection Y/N|-0.011|-0.010|\\n||(0.020)|(0.020)|\\n|Female top manager Y/N|0.010|0.012|\\n||(0.011)|(0.011)|\\n|Firm Formally Registered when Started Operations|0.047**|0.047**|\\n||(0.022)|(0.022)|\\n|Top manager experience in sector (years)|-0.001|-0.001|\\n||(0.000)|(0.000)|\\n|Log of age of firm|-0.014*|-0.014*|\\n||(0.008)|(0.008)|\\n|Log of size|0.035***|0.035***|\\n||(0.004)|(0.004)|\\n|Direct exports 10% or more of sales Y/N|0.032***|0.031***|\\n||(0.010)|(0.010)|\\n|Foreign ownership Y/N|0.025**|0.025**|\\n||(0.011)|(0.011)|\\n|Establishment has checking or savings account Y/N|0.017|0.020|\\n||(0.022)|(0.022)|\\n|Establishment has a line of credit or loan Y/N|0.028***|0.027***|\\n||(0.009)|(0.009)|\\n|ISO Certification Ownership Y/N|0.054***|0.054***|\\n||(0.011)|(0.011)|\\n|Firm identifying labor regulations as a major or severe constraint Y/N|-0.007|-0.006|\\n||(0.016)|(0.015)|\\n|Manufacturing Sector Y/N|-0.008|-0.008|\\n||(0.009)|(0.009)|\\n|Constant|0.298***|0.295***|\\n||(0.044)|(0.044)|\\n\\n16\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"|Country Fixed Effects|YES|YES|\\n|---|---|---|\\n|Number of observations|9,015|9,015|\\n|Adjusted R2|0.226|0.228|\\n\\nnote: *** p<0.01, ** p<0.05, * p<0.1, Standard errors clustered at the strata level. Survey weights are employed.\\n\\n17\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"# **Table 3: Family Management and Management Practices by Region**\\n\\n|**Model**|||**OLS with Country**|**FE (Cross-sectio**|**n)**||\\n|---|---|---|---|---|---|---|\\n|**Outcome Variable**|||**Overall Man**|**agement Score**|||\\n|||**Eastern**|||||\\n||**Developing**<br>**MENA**|**Europe and**<br>**Central**<br>**Asia**|**Western Europe**<br>**(Italy, Portugal,**<br>**Greece, Cyprus)**|**Developing**<br>**MENA**|**Eastern**<br>**Europe and**<br>**Central Asia**|**Western Europe**<br>**(Italy, Portugal,**<br>**Greece, Cyprus)**|\\n||**(1)**|**(2)**|**(3)**|**(4)**|**(5)**|**(6)**|\\n|Share of family management (0 to 1)|-0.124**|-0.048**|-0.055||||\\n||(0.054)|(0.020)|(0.054)||||\\n|All management is family Y/N||||-0.126***|-0.036**|-0.075**|\\n|||||(0.037)|(0.015)|(0.033)|\\n|Share of same family ownership (0 to 1)|-0.007|0.039*|0.059|-0.021|0.028|0.074|\\n||(0.043)|(0.020)|(0.057)|(0.031)|(0.017)|(0.047)|\\n|Owner, Manager or Board Political<br>Connection Y/N|-0.019|0.019|-0.261***|-0.021|0.021|-0.257***|\\n||(0.041)|(0.019)|(0.082)|(0.040)|(0.019)|(0.085)|\\n|Female top manager Y/N|0.060|0.006|-0.050|0.055|0.008|-0.044|\\n||(0.048)|(0.012)|(0.037)|(0.046)|(0.012)|(0.037)|\\n|Firm Formally Registered when Started<br>Operations|0.012|0.053*|0.125**|0.013|0.053*|0.131***|\\n||(0.043)|(0.030)|(0.050)|(0.042)|(0.030)|(0.049)|\\n|Top manager experience in sector (years)|-0.0005|-0.0005|0.0003|-0.001|-0.001|0.0002|\\n||(0.001)|(0.001)|(0.001)|(0.001)|(0.001)|(0.001)|\\n|Log of age of firm|-0.019|-0.014*|-0.016|-0.015|-0.014|-0.012|\\n||(0.020)|(0.008)|(0.026)|(0.020)|(0.008)|(0.026)|\\n|Log of size|0.005|0.038***|0.048**|0.006|0.039***|0.049**|\\n||(0.013)|(0.005)|(0.020)|(0.014)|(0.005)|(0.019)|\\n|Direct exports 10% or more of sales Y/N|0.026|0.023**|0.111***|0.026|0.022*|0.108***|\\n||(0.022)|(0.011)|(0.036)|(0.021)|(0.011)|(0.035)|\\n|Foreign ownership Y/N|0.032|0.029**|-0.013|0.036|0.030**|-0.010|\\n||(0.028)|(0.012)|(0.041)|(0.028)|(0.012)|(0.040)|\\n|Establishment has checking or savings<br>account Y/N|0.073**|-0.004|-0.148***|0.071**|-0.001|-0.155***|\\n||(0.031)|(0.026)|(0.057)|(0.030)|(0.026)|(0.059)|\\n|Establishment has a line of credit or loan<br>Y/N|0.005|0.028***|0.014|-0.001|0.028**|0.010|\\n||(0.023)|(0.011)|(0.044)|(0.022)|(0.011)|(0.044)|\\n|ISO Certification Ownership Y/N|0.048|0.058***|0.002|0.053|0.058***|-0.004|\\n||(0.039)|(0.011)|(0.047)|(0.039)|(0.011)|(0.047)|\\n|Firm identifying labor regulations as a<br>major or severe constraint Y/N|-0.089**|-0.003|0.044|-0.089**|-0.003|0.057|\\n||(0.037)|(0.017)|(0.051)|(0.035)|(0.017)|(0.051)|\\n\\n18\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"|Manufacturing Sector Y/N|-0.009|-0.009|0.009|-0.014|-0.009|0.010|\\n|---|---|---|---|---|---|---|\\n||(0.026)|(0.010)|(0.035)|(0.025)|(0.010)|(0.035)|\\n|Constant|0.424***|0.257***|0.216**|0.415***|0.249***|0.205*|\\n||(0.102)|(0.053)|(0.108)|(0.105)|(0.053)|(0.106)|\\n|Country Fixed Effects|YES|YES|YES|YES|YES|YES|\\n|Number of observations|1,821|5,982|974|1,821|5,982|974|\\n|Adjusted R2|0.218|0.211|0.289|0.231|0.211|0.301|\\n\\nnote: *** p<0.01, ** p<0.05, * p<0.1, Standard errors clustered at the strata level. Survey weights are employed.\\n\\n19\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"**<u>Table 4: Family Management and Management Practices by Income</u>**\\n\\n|**Model**||**OLS with Countr**|**y FE (Cross-section)**||\\n|---|---|---|---|---|\\n|**Outcome Variable**||**Overall Man**|**agement Score**||\\n||**Low and Middle**<br>**Income Economies**|**High Income**<br>**Economies**|**Low and Middle**<br>**Income Economies**|**High Income**<br>**Economies**|\\n||**(1)**|**(2)**|**(3)**|**(4)**|\\n|Share of family management (0 to 1)|-0.057*|-0.026|||\\n||(0.029)|(0.019)|||\\n|All management is family Y/N|||-0.045**|-0.043***|\\n||||(0.021)|(0.015)|\\n|Share of same family ownership (0 to 1)|0.026|0.006|0.013|0.019|\\n||(0.027)|(0.021)|(0.021)|(0.019)|\\n|Owner, Manager or Board Political Connection Y/N|0.004|-0.051|0.005|-0.049|\\n||(0.024)|(0.037)|(0.024)|(0.037)|\\n|Female top manager Y/N|0.010|0.010|0.011|0.013|\\n||(0.014)|(0.017)|(0.014)|(0.017)|\\n|Firm Formally Registered when Started Operations|0.045|0.044|0.045|0.044|\\n||(0.029)|(0.033)|(0.029)|(0.032)|\\n|Top manager experience in sector (years)|-0.001|-0.0004|-0.001|-0.0004|\\n||(0.001)|(0.001)|(0.001)|(0.001)|\\n|Log of age of firm|-0.015|-0.014|-0.015|-0.015|\\n||(0.010)|(0.013)|(0.010)|(0.013)|\\n|Log of size|0.036***|0.034***|0.036***|0.033***|\\n||(0.005)|(0.008)|(0.005)|(0.008)|\\n|Direct exports 10% or more of sales Y/N|0.023*|0.046***|0.023*|0.045***|\\n||(0.013)|(0.014)|(0.013)|(0.014)|\\n|Foreign ownership Y/N|0.017|0.038**|0.018|0.037**|\\n||(0.016)|(0.015)|(0.016)|(0.015)|\\n|Establishment has checking or savings account Y/N|0.027|-0.026|0.030|-0.023|\\n||(0.022)|(0.059)|(0.022)|(0.060)|\\n|Establishment has a line of credit or loan Y/N|0.021*|0.040***|0.021*|0.039***|\\n||(0.012)|(0.014)|(0.012)|(0.014)|\\n|ISO Certification Ownership Y/N|0.050***|0.064***|0.050***|0.063***|\\n||(0.015)|(0.015)|(0.015)|(0.015)|\\n|Firm identifying labor regulations as a major or severe<br>constraint Y/N|-0.044**|0.030|-0.044**|0.032|\\n||(0.020)|(0.023)|(0.019)|(0.023)|\\n|Manufacturing Sector Y/N|-0.005|-0.012|-0.005|-0.012|\\n||(0.011)|(0.015)|(0.011)|(0.015)|\\n|Constant|0.292***|0.389***|0.288***|0.387***|\\n\\n20\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"||(0.052)|(0.081)|(0.052)|(0.083)|\\n|---|---|---|---|---|\\n|Country Fixed Effects|YES|YES|YES|YES|\\n|Number of observations|5,931|3,084|5,931|3,084|\\n|Adjusted R2|0.202|0.275|0.202|0.280|\\n\\nnote: *** p<0.01, ** p<0.05, * p<0.1, Standard errors clustered at the strata level. Survey weights are employed.\\n\\n21\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"**<u>Table 5: Government Interactions and Family Management Practices</u>**\\n\\n|**Model**||**OLS with Country**|**FE (Cross-section)**||\\n|---|---|---|---|---|\\n|**Outcome Variable**||**Overall Mana**|**gement Score**||\\n||**(1)**|**(2)**|**(3)**|**(4)**|\\n|All Family Management x Political connections|-0.083||||\\n||(0.053)||||\\n|Family Management share x Political connections||-0.092*|||\\n|||(0.051)|||\\n|All Family Management x Informal payments to get<br>things done|||-0.069**||\\n||||(0.031)||\\n|Family Management share x Informal payments to get<br>things done||||-0.074**|\\n|||||(0.031)|\\n|All management is family Y/N|-0.040***||-0.035***||\\n||(0.013)||(0.013)||\\n|Share of family management (0 to 1)||-0.038**||-0.032*|\\n|||(0.018)||(0.018)|\\n|Firm expected to make payment to get things done Y/N|||0.004|0.015|\\n||||(0.015)|(0.017)|\\n|Owner, Manager or Board Political Connection Y/N|0.009|0.017|-0.017|-0.018|\\n||(0.020)|(0.021)|(0.018)|(0.018)|\\n|Share of same family ownership (0 to 1)|0.013|0.015|0.011|0.012|\\n||(0.015)|(0.018)|(0.014)|(0.018)|\\n|Female top manager Y/N|0.011|0.010|0.011|0.009|\\n||(0.011)|(0.011)|(0.011)|(0.011)|\\n|Firm Formally Registered when Started Operations|0.048**|0.047**|0.044**|0.043*|\\n||(0.022)|(0.022)|(0.022)|(0.023)|\\n|Top manager experience in sector (years)|-0.001|-0.001|-0.001*|-0.001|\\n||(0.000)|(0.000)|(0.000)|(0.000)|\\n|Log of age of firm|-0.013*|-0.014*|-0.012|-0.012|\\n||(0.008)|(0.008)|(0.008)|(0.008)|\\n|Log of size|0.035***|0.035***|0.034***|0.034***|\\n||(0.004)|(0.004)|(0.005)|(0.005)|\\n|Direct exports 10% or more of sales Y/N|0.031***|0.031***|0.034***|0.034***|\\n||(0.010)|(0.010)|(0.010)|(0.010)|\\n|Foreign ownership Y/N|0.026**|0.025**|0.024**|0.024**|\\n||(0.011)|(0.011)|(0.012)|(0.012)|\\n|Establishment has checking or savings account Y/N|0.019|0.016|0.019|0.018|\\n||(0.022)|(0.022)|(0.023)|(0.022)|\\n|Establishment has a line of credit or loan Y/N|0.027***|0.028***|0.033***|0.034***|\\n||(0.009)|(0.009)|(0.008)|(0.008)|\\n\\n22\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"|ISO Certification Ownership Y/N|0.054***|0.054***|0.050***|0.050***|\\n|---|---|---|---|---|\\n||(0.011)|(0.011)|(0.011)|(0.011)|\\n|Firm identifying labor regulations as a major or severe<br>constraint Y/N|-0.006|-0.006|-0.006|-0.007|\\n||(0.015)|(0.015)|(0.016)|(0.016)|\\n|Manufacturing Sector Y/N|-0.008|-0.007|-0.010|-0.009|\\n||(0.009)|(0.009)|(0.009)|(0.009)|\\n|Constant|0.292***|0.295***|0.299***|0.300***|\\n||(0.043)|(0.043)|(0.044)|(0.044)|\\n|Country Fixed Effects|YES|YES|YES|YES|\\n|Number of observations|9,015|9,015|8,493|8,493|\\n|Adjusted R2|0.230|0.228|0.224|0.222|\\n\\nnote: *** p<0.01, ** p<0.05, * p<0.1, Standard errors clustered at the strata level. Survey weights are employed.\\n\\n23\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"# **Table 6: Competition, Innovation, and Regulation and Family Management Practices**\\n\\n|**Model**||**OLS w**|**ith Country**|**FE (Cross-sect**|**ion)**||\\n|---|---|---|---|---|---|---|\\n|**Outcome Variable**||**O**|**verall Mana**|**gement Score**|||\\n||**(1)**|**(2)**|**(3)**|**(4)**|**(5)**|**(6)**|\\n|All Family Management x No or One Competitor<br>(domestic)|-0.100**<br>(0.042)||||||\\n|Family Management share x No or One Competitor<br>(domestic)||-0.119**|||||\\n|||(0.047)|||||\\n|All Family Management x Introduced Process<br>Innovation|||0.047**<br>(0.022)||||\\n|Family Management share x Introduced Process<br>Innovation||||0.056**|||\\n|||||(0.023)|||\\n|All Family Management x Labor Regulations<br>Major/Severe Obstacle|||||0.092***<br>(0.032)||\\n|Family Management share x Labor Regulations<br>Major/Severe Obstacle||||||0.125***<br>(0.033)|\\n|All management is family Y/N|-0.037**||-<br>0.050***||-0.054***||\\n||(0.016)||(0.013)||(0.013)||\\n|Share of family management (0 to 1)||-0.035||-0.048***||-0.058***|\\n|||(0.021)||(0.017)||(0.018)|\\n|No or One competitor (Y/N) (domestic)|0.038|0.051|||||\\n||(0.034)|(0.038)|||||\\n|Introduced a process innovation|||0.057***|0.050***|||\\n||||(0.014)|(0.016)|||\\n|Share of same family ownership (0 to 1)|0.009|0.012|0.007|0.007|0.014|0.018|\\n||(0.019)|(0.022)|(0.014)|(0.017)|(0.015)|(0.018)|\\n|Owner, Manager or Board Political Connection Y/N|-0.013|-0.015|-0.011|-0.011|-0.011|-0.011|\\n||(0.023)|(0.023)|(0.020)|(0.021)|(0.020)|(0.020)|\\n|Female top manager Y/N|0.022*|0.021*|0.009|0.008|0.011|0.010|\\n||(0.012)|(0.012)|(0.011)|(0.011)|(0.011)|(0.011)|\\n|Firm Formally Registered when Started Operations|0.043*|0.042*|0.053**|0.053**|0.048**|0.049**|\\n||(0.024)|(0.024)|(0.022)|(0.023)|(0.022)|(0.022)|\\n|Top manager experience in sector (years)|-0.001|-0.001|-0.001|-0.001|-0.001|-0.001|\\n||(0.001)|(0.001)|(0.000)|(0.000)|(0.000)|(0.000)|\\n|Log of age of firm|-0.018**|-0.019**|-0.013*|-0.013*|-0.014*|-0.014*|\\n||(0.009)|(0.009)|(0.008)|(0.008)|(0.008)|(0.008)|\\n\\n24\", \"datasets\": [{\"mention_name\": {\"text\": \"gement Score\", \"confidence\": 0.46042269468307495, \"start\": 234, \"end\": 246, \"clean_text\": \"gement Score\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 234, \"end\": 246, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 234, \"end\": 246, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 234, \"end\": 246}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 234, \"end\": 246, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.6101600527763367, \"start\": 234, \"end\": 246}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.46042269468307495, \"start\": 234, \"end\": 246}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6892645955085754, \"start\": 234, \"end\": 246}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6892645955085754, \"start\": 234, \"end\": 246}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"|Log of size|0.035***|0.035***|0.034***|0.034***|0.035***|0.035***|\\n|---|---|---|---|---|---|---|\\n||(0.005)|(0.005)|(0.004)|(0.004)|(0.004)|(0.004)|\\n|Direct exports 10% or more of sales Y/N|0.029**|0.030**|0.030***|0.030***|0.032***|0.033***|\\n||(0.013)|(0.013)|(0.009)|(0.009)|(0.010)|(0.010)|\\n|Foreign ownership Y/N|0.041***|0.042***|0.025**|0.025**|0.025**|0.023**|\\n||(0.013)|(0.013)|(0.011)|(0.011)|(0.011)|(0.011)|\\n|Establishment has checking or savings account Y/N|0.025|0.022|0.016|0.015|0.022|0.021|\\n||(0.025)|(0.025)|(0.021)|(0.021)|(0.021)|(0.022)|\\n|Establishment has a line of credit or loan Y/N|0.024**|0.024**|0.022**|0.023**|0.027***|0.028***|\\n||(0.011)|(0.011)|(0.009)|(0.010)|(0.009)|(0.009)|\\n|ISO Certification Ownership Y/N|0.062***|0.062***|0.048***|0.048***|0.054***|0.054***|\\n||(0.013)|(0.013)|(0.011)|(0.011)|(0.011)|(0.011)|\\n|Firm identifying labor regulations as a major or<br>severe constraint Y/N|-0.008|-0.008|-0.011|-0.012|-0.035**|-0.058***|\\n||(0.018)|(0.018)|(0.015)|(0.015)|(0.017)|(0.018)|\\n|Manufacturing Sector Y/N|-0.004|-0.004|-0.013|-0.012|-0.008|-0.007|\\n||(0.010)|(0.010)|(0.009)|(0.009)|(0.009)|(0.009)|\\n|Constant|0.296***|0.299***|0.303***|0.304***|0.294***|0.297***|\\n||(0.049)|(0.049)|(0.044)|(0.044)|(0.044)|(0.044)|\\n|Country Fixed Effects|YES|YES|YES|YES|YES|YES|\\n|Number of observations|6,856|6,856|8,954|8,954|9,015|9,015|\\n|Adjusted R2|0.224|0.223|0.250|0.248|0.232|0.232|\\n\\nnote: *** p<0.01, ** p<0.05, * p<0.1, Standard errors clustered at the strata level. Survey weights are employed.\\n\\n25\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"# **Table 7: Family Management and Management Practices Sub-scores**\\n\\n|**Model**||**MG2**|**OLS wit**<br>**MG3 Level of**|**h Country FE (C**<br>**MG4**|**ross-section)**||||\\n|---|---|---|---|---|---|---|---|---|\\n|**Outcome Variable**|**MG1**<br>**Problem**<br>**resolution**|<br>**Number of**<br>**performance**<br>**indicators**<br>**monitored**|**ease or difficulty**<br>**to achieve**<br>**production or**<br>**service provision**<br>**targets**|**Knowledge of**<br>**production or**<br>**service**<br>**provision**<br>**targets**|**MG5**<br>**Basis of**<br>**bonuses**|**MG6 Length**<br>**of focus of**<br>**production**<br>**targets**|**MG7**<br>**Promotion**<br>**of non-**<br>**mangers**|**MG8**<br>**Dismissal**|\\n||**(1)**|**(2)**|**(3)**|**(4)**|**(5)**|**(6)**|**(7)**|**(8)**|\\n|All management is<br>family Y/N|-0.020|-0.082***|-0.033|-0.076***|-0.053**|-0.029|-0.040|-0.026|\\n||(0.020)|(0.019)|(0.024)|(0.024)|(0.027)|(0.027)|(0.027)|(0.027)|\\n|Constant|0.584***|-0.048|0.177**|-0.002|0.382***|0.229**|0.727***|0.313***|\\n||(0.073)|(0.068)|(0.082)|(0.069)|(0.094)|(0.095)|(0.081)|(0.111)|\\n|Controls|YES|YES|YES|YES|YES|YES|YES|YES|\\n|Country Fixed Effects|YES|YES|YES|YES|YES|YES|YES|YES|\\n|Number of observations|9,015|9,015|9,015|9,015|9,015|9,015|9,015|9,015|\\n|Adjusted R2|0.168|0.219|0.129|0.142|0.139|0.168|0.132|0.152|\\n|Share of family<br>management (0 to 1)|0.025|-0.065**|-0.038|-0.098***|-0.052|-0.040|-0.058*|-0.027|\\n||(0.028)|(0.025)|(0.030)|(0.030)|(0.035)|(0.036)|(0.035)|(0.036)|\\n|Constant|0.579***|-0.045|0.179**|0.007|0.385***|0.233**|0.732***|0.315***|\\n||(0.074)|(0.069)|(0.082)|(0.068)|(0.094)|(0.095)|(0.081)|(0.111)|\\n|Controls|YES|YES|YES|YES|YES|YES|YES|YES|\\n|Country Fixed Effects|YES|YES|YES|YES|YES|YES|YES|YES|\\n|Number of observations|9,015|9,015|9,015|9,015|9,015|9,015|9,015|9,015|\\n|Adjusted R2|0.168|0.215|0.128|0.142|0.139|0.168|0.132|0.152|\\n\\nnote: *** p<0.01, ** p<0.05, * p<0.1, Standard errors clustered at the strata level. Survey weights are employed. Controls are the same as in the base regressions in Table 2 26\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [27]}}, {\"page\": 28, \"chunk\": 28, \"input_text\": \"# **Table A1: Characteristics of Family Managed Firms in MENA**\\n\\n||**All**<br>|**Manage**<br>|**ment is Family**<br>|\\n|---|---|---|---|\\n||**YES**|**NO**|**Significance**|\\n|Overall Management Score|0.47|0.53|***|\\n|Medium Firms (%)|81.27|72.10|***|\\n|Large Firms (%)|11.66|23.78|***|\\n|Young Firms (5 years or less)|6.39|6.19||\\n|Manufacturing firms (%)|38.20|39.84||\\n|Exporter (%)|25.26|26.90||\\n|Foreign Owned (%)|5.53|13.89|***|\\n|Female Top Manager (%)|17.57|13.50|**|\\n|Share of same family ownership (%)|96.15|32.32|***|\\n|Owner, Manager or Board Political Connection (% of firms)|5.23|8.65|**|\\n|Top manager experience in sector (years)|23.86|20.09|***|\\n|Checking or savings account (% of firms)|97.26|97.44||\\n|Line of credit or loan (% of firms)|50.08|46.49||\\n|ISO Certification Ownership (% of firms)|31.46|33.32||\\n|Firm identifying labor regulations as a major or severe constraint (% of firms)|0.10|0.09||\\n|Firm Formally Registered when Started Operations|95.98|95.82||\\n|Firm expected to make payment to get things done (% of firms)|10.71|11.33||\\n|No or One competitor (domestic) (% of firms)|5.99|6.26||\\n|Introduced a processinnovation(% of firms)|21.70|23.65||\\n\\n27\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [28]}}, {\"page\": 29, \"chunk\": 29, \"input_text\": \"**Table A2: Frequency of Manager Interactions (% of firms)**\\n\\n|**All Management is F**|**amily**|||\\n|---|---|---|---|\\n||**YES**|**NO**|**Difference**|\\n|**Other Decision Makers (**e.g. COO, CAO, CMO, Board members,<br>Business Unitmanagers, or managersfroma parent company**)**||||\\n|Never|4.32|2.89|1.43|\\n|Once a week|27.13|30.97|-3.85|\\n|Between 2 and 4 times a week|22.49|24.89|-2.40|\\n|Daily|37.68|34.92|2.76|\\n|More than once a day|7.54|4.85|2.69|\\n|**Suppliers**||||\\n|Never|23.85|24.36|-0.51|\\n|Once a week|40.25|45.45|-5.21|\\n|Between 2 and 4 times a week|19.85|17.25|2.60|\\n|Daily|12.31|9.24|3.07|\\n|More than once a day|0.86|0.49|0.37|\\n|**Employees**||||\\n|Never|10.77|9.44|1.33|\\n|Once a week|30.21|33.41|-3.19|\\n|Between 2 and 4 times a week|19.63|21.57|-1.93|\\n|Daily|31.78|30.44|1.34|\\n|More than once a day|0.28|0.22|0.06|\\n\\nNote: *** p<0.01, ** p<0.05, * p<0.1. This information is only available for large firms. None of the differences between family-managed and non-family managed firms are statistically significant 28\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [29]}}, {\"page\": 30, \"chunk\": 30, \"input_text\": \"# **Table A3: Family Management Practices and Firm Characteristics**\\n\\n|**Model**|||**OLS with Co**|**untry FE (Cr**|**oss-section)**|||\\n|---|---|---|---|---|---|---|---|\\n|**Outcome Variable**|||**Overall**|**Management**|**Score**|||\\n||**(1)**|**(2)**|**(3)**|**(4)**|**(5)**|**(6)**|**(7)**|\\n|All Family Management x Share of same<br>family ownership (%)|-0.021<br>(0.072)|||||||\\n|All Family Management x Size (in logs)||-0.0004||||||\\n|||(0.010)||||||\\n|All Family Management x Exporter|||0.005|||||\\n||||(0.021)|||||\\n|All Family Management x Foreign<br>ownership||||0.024||||\\n|||||(0.038)||||\\n|All Family Management x Age (in logs)|||||-0.025*|||\\n||||||(0.015)|||\\n|All Family Management x Manufacturing||||||0.016||\\n|||||||(0.019)||\\n|All Family Management x Female<br>Manager|||||||0.010|\\n||||||||(0.023)|\\n|All management is family Y/N|-0.026|-0.043|-0.046***|-0.046***|0.028|-0.051***|-0.047***|\\n||(0.071)|(0.044)|(0.015)|(0.013)|(0.046)|(0.016)|(0.014)|\\n|Share of same family ownership (0 to 1)|0.015|0.014|0.014|0.014|0.013|0.014|0.014|\\n||(0.015)|(0.015)|(0.015)|(0.015)|(0.015)|(0.015)|(0.015)|\\n|Owner, Manager or Board Political<br>Connection Y/N|-0.010|-0.010|-0.010|-0.010|-0.009|-0.010|-0.010|\\n||(0.020)|(0.020)|(0.020)|(0.020)|(0.020)|(0.020)|(0.020)|\\n|Female top manager Y/N|0.011|0.012|0.011|0.011|0.011|0.012|0.008|\\n||(0.011)|(0.011)|(0.011)|(0.011)|(0.011)|(0.011)|(0.013)|\\n|Firm Formally Registered when Started<br>Operations|0.047**|0.047**|0.047**|0.048**|0.046**|0.046**|0.047**|\\n||(0.022)|(0.022)|(0.022)|(0.022)|(0.022)|(0.022)|(0.022)|\\n|Top manager experience in sector (years)|-0.001|-0.001|-0.001|-0.001|-0.001|-0.001|-0.001|\\n||(0.000)|(0.000)|(0.000)|(0.000)|(0.000)|(0.000)|(0.000)|\\n|Log of age of firm|-0.014*|-0.014*|-0.014*|-0.014*|-0.006|-0.014*|-0.014*|\\n||(0.008)|(0.008)|(0.008)|(0.008)|(0.008)|(0.008)|(0.008)|\\n|Log of size|0.035***|0.035***|0.035***|0.035***|0.035***|0.035***|0.035***|\\n||(0.004)|(0.005)|(0.004)|(0.004)|(0.004)|(0.004)|(0.004)|\\n|Direct exports 10% or more of sales Y/N|0.031***|0.031***|0.030**|0.031***|0.031***|0.032***|0.031***|\\n||(0.010)|(0.010)|(0.012)|(0.010)|(0.010)|(0.009)|(0.010)|\\n|Foreign ownership Y/N|0.025**|0.025**|0.025**|0.021*|0.026**|0.026**|0.025**|\\n||(0.011)|(0.011)|(0.011)|(0.012)|(0.011)|(0.011)|(0.011)|\\n\\n29\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [30]}}, {\"page\": 31, \"chunk\": 31, \"input_text\": \"|Establishment has checking or savings<br>account Y/N|0.019|0.019|0.020|0.019|0.020|0.021|0.019|\\n|---|---|---|---|---|---|---|---|\\n||(0.022)|(0.022)|(0.022)|(0.022)|(0.022)|(0.022)|(0.022)|\\n|Establishment has a line of credit or loan<br>Y/N|0.027***|0.027***|0.027***|0.027***|0.028***|0.027***|0.027***|\\n||(0.009)|(0.009)|(0.009)|(0.009)|(0.009)|(0.009)|(0.009)|\\n|ISO Certification Ownership Y/N|0.054***|0.054***|0.054***|0.054***|0.054***|0.054***|0.054***|\\n||(0.011)|(0.011)|(0.011)|(0.011)|(0.011)|(0.011)|(0.011)|\\n|Firm identifying labor regulations as a<br>major or severe constraint Y/N|-0.006|-0.006|-0.006|-0.006|-0.007|-0.006|-0.006|\\n||(0.015)|(0.015)|(0.015)|(0.015)|(0.015)|(0.015)|(0.015)|\\n|Manufacturing Sector Y/N|-0.008|-0.008|-0.008|-0.008|-0.009|-0.013|-0.008|\\n||(0.009)|(0.009)|(0.009)|(0.009)|(0.009)|(0.010)|(0.009)|\\n|Constant|0.296***|0.295***|0.295***|0.295***|0.274***|0.296***|0.295***|\\n||(0.044)|(0.046)|(0.044)|(0.044)|(0.043)|(0.044)|(0.044)|\\n|Country Fixed effects|YES|YES|YES|YES|YES|YES|YES|\\n|Number of observations|9,015|9,015|9,015|9,015|9,015|9,015|9,015|\\n|Adjusted R2|0.228|0.228|0.228|0.228|0.230|0.228|0.228|\\n\\nnote: *** p<0.01, ** p<0.05, * p<0.1, Standard errors clustered at the strata level. Survey weights are employed.\\n\\n30\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [31]}}, {\"page\": 32, \"chunk\": 32, \"input_text\": \"# **Table A4: Management Practices Scoring**\\n\\n|**MG1 Problem resolution (r1)**|**Score**|\\n|---|---|\\n|**Action when problem in the production/service provision arose**||\\n|Most structured: We fixed it and took action to make sure that it did not happen again, and had a<br>continuousimprovement process to anticipate problemslike theseinadvance|1|\\n|Second most structured: We fixed it and took action to make sure it did not happen again|0.667|\\n|Second least structured: We fixed it but did not take further action|0.333|\\n|Least structured: No action was taken|0|\\n|**MG2 Number of performance indicators monitored (r3)**|**Score**|\\n|**Number of production or service provision performance indicators monitored**||\\n|10 or more indicators|1|\\n|3-9 indicators|0.667|\\n|1-2 indicators|0.333|\\n|No indicators|0|\\n|**MG6 Length of focus of production targets**|**Score**|\\n|**Focus of production targets**||\\n|Combination of short-term and long-term targets|1|\\n|long-term only|0.667|\\n|short-term only|0.333|\\n|No targets ortargetsnot achieved|0|\\n|**MG3 Level of ease or difficulty to achieve production or service provision targets (r6)**|**Score**|\\n|**Level of ease or difficulty to achieve targets**||\\n|No targets or targets not achieved|0|\\n|Achieved without much effort|0.2|\\n|Only achieved with extraordinary effort|0.4|\\n|Achieved with some effort|0.6|\\n|Achieved with normal amount of effort|0.8|\\n|Achieved with more than normal effort|1|\\n\\n|**MG4 Knowledge of production or service provision targets (r7)**|**Score**|\\n|---|---|\\n|**Personnel's knowledge of production or service provision targets**||\\n|All managers andmost workers|1|\\n|Most managers and most workers|0.667|\\n|Most managers and some workers|0.333|\\n|Only senior managers|0|\\n|No targets|0|\\n|**MG5 Basis of bonuses (r9)**|**Score**|\\n|**What managers' performance bonuses were usually based on**||\\n\\n31\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [32]}}, {\"page\": 33, \"chunk\": 33, \"input_text\": \"|Their own performance as measured by targets|1|\\n|---|---|\\n|Their team or shift performance as measured by targets|0.75|\\n|Their establishment’s performance as measured by targets|0.5|\\n|Their company’s performance as measured by targets|0.25|\\n|No performance bonuses|0|\\n|**MG7 Promotion of non-mangers**|**Score**|\\n|**Basis for promoting non-mangers**||\\n|Based solely onperformance and ability|1|\\n|Based partly on performance and ability, and partly on other factors (for example, tenure or family<br>connections)|0.667|\\n|Based mainly on factors other than performance and ability (for example, tenure or family connections)|0.333|\\n|Non-managers are normally not promoted|0|\\n\\n|**MG8 Dismissal**|**Score**|\\n|---|---|\\n|**When underperforming managers were dismissed or reassigned**||\\n|Within6months ofunderperformance|1|\\n|After 6 months|0.5|\\n|Rarely or never|0|\\n\\nNote: “Don’t know” responses are removed from the scores 32\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [33]}}, {\"page\": 34, \"chunk\": 34, \"input_text\": \"**Figure 1: Overall Management Score**\\n\\n<!-- Start of picture text --> 0.80<br>0.70<br>0.60<br>0.50<br>0.40<br>0.30<br>0.20<br>0.10<br>0.00<br>Figure 2: All Family Management (% of Firms)<br>100<br>90<br>80<br>70<br>60<br>50<br>40<br>30<br>20<br>10<br>0<br>Portugal Italy Poland Kosovo Tajikistan Turkey Kazakhstan Uzbekistan Lebanon West Bank and Gaza Romania Montenegro Morocco Albania Lithuania Egypt, Arab Rep. Hungary Georgia Slovak Republic Greece Jordan Moldova Armenia Croatia Kyrgyz Republic Bosnia and Herzegovina Russian Federation Ukraine North Macedonia Cyprus Slovenia Estonia Belarus Tunisia Serbia Mongolia Czech Republic Latvia Malta Bulgaria Azerbaijan<br>Russian Federation Georgia Azerbaijan Tunisia Egypt, Arab Rep. Kyrgyz Republic Montenegro Jordan Armenia Morocco Turkey Belarus Kazakhstan Albania Serbia Ukraine Poland Uzbekistan Tajikistan Mongolia West Bank and Gaza Moldova Latvia Bosnia and Herzegovina Romania Italy Kosovo Bulgaria Slovenia Estonia Slovak Republic Lithuania Malta Hungary Czech Republic Greece Lebanon Croatia Portugal Cyprus North Macedonia<br><!-- End of picture text --> 33\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [34]}}, {\"page\": 35, \"chunk\": 35, \"input_text\": \"# **Figure 3: Percentage of Family Management (%)**\\n\\n<!-- Start of picture text --> 100<br>90<br>80<br>70<br>60<br>50<br>40<br>30<br>20<br>10<br>0<br>Russian Federation Georgia Azerbaijan Egypt, Arab Rep. Kyrgyz Republic Montenegro Jordan Kazakhstan Armenia Turkey Serbia Ukraine Tajikistan Mongolia Uzbekistan Belarus Poland Tunisia Albania Morocco West Bank and Gaza Bosnia and Herzegovina Romania Moldova Latvia Italy Bulgaria Slovenia Estonia Slovak Republic Kosovo Hungary Lithuania Greece Lebanon Malta Czech Republic Croatia Portugal Cyprus North Macedonia<br><!-- End of picture text --> 34\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099626001292432099/pdf/IDU16da65b2c1ef8914b1a19d5e1ac4acf945a1b.pdf\", \"pages\": [35]}}]}"}}},{"rowIdx":90,"cells":{"doc_id":{"kind":"string","value":"34248405"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34248405\", \"last_modified_date\": \"2024-02-09T00:00:00Z\", \"authors\": {\"0\": {\"author\": \"Cortina Toro,Magdalena\"}, \"1\": {\"author\": \"Jimenez,Juan Miguel\"}, \"2\": {\"author\": \"Rozo Villarraga,Sandra Viviana\"}}, \"count\": \"World\", \"count_key\": \"517191\", \"docna\": {\"0\": {\"docna\": \"Little Nomads : Economic and Social Impacts\\n of Migration on Children\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"DECRG: Poverty & Inequality (DECPI)\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34248405\"}, \"repnb\": \"WPS10686\", \"docdt\": \"2024-01-30T00:00:00Z\", \"datestored\": \"2024-01-30T02:18:34Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"This paper reviews the main findings\\n from 113 studies produced between 1990 and 2023, focusing on\\n the impact of migration on various child groups affected\\n through the migration path, including left-behind, migrant\\n (voluntary and forced), and native children. The findings\\n reveal that migration influences children’s\\n outcomes in complex and context-dependent ways, and it\\n interacts dramatically with household demographics and\\n public policies. Key results include the following: (i)\\n left-behind children benefit from remittances but experience\\n dramatic declines in their cognitive and non-cognitive\\n development due to parental absence; (ii) immigrant children\\n generally fare better than those in their origin countries\\n but still underperform compared to native children in host\\n countries; and (iii) the impacts of migration on native\\n children largely depend on the adjustment of public service\\n supply to meet increased demand. In cases where education\\n services expand to meet rising demand, the effect on native\\n children can be minimal or even positive. This paper\\n emphasizes the need for more experimental or\\n quasi-experimental research to examine the effectiveness of\\n programs that support migrant and minor host children, and\\n it calls for longitudinal data collection to better\\n understand the challenges and needs of migrant children,\\n particularly in developing countries.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10686; PEOPLE\", \"display_title\": \"Little Nomads : Economic and Social\\n Impacts of Migration on Children\", \"disclosure_date\": \"2024-01-30T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099418301302435890/text/IDU113cacc80159c21491318dca10779c4990d0b.txt\", \"listing_relative_url\": \"/research/2024/01/34248405/little-nomads-economic-social-impacts-migration-children\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099418301302435890/IDU113cacc80159c21491318dca10779c4990d0b\", \"new_url\": \"2024/01/34248405/IDU113cacc80159c21491318dca10779c4990d0b\", \"ext_pub_date\": \"2024-01-30T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Child Development m1327025 2151,Child Labor and Education m1327744 1543,Migration Policies and Jobs m1504161 2836\", \"docm_id\": \"IDU-13cacc80-59c2-4913-8dca-0779c4990d0b\", \"chronical_docm_id\": \"IDU-13cacc80-59c2-4913-8dca-0779c4990d0b\", \"publishtoextweb_dt\": \"2024-01-30T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10686\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"guid\": \"099418301302435890\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"46\", \"author_keys\": {\"0\": {\"author_key\": \"2346931\"}, \"1\": {\"author_key\": \"NA\"}, \"2\": {\"author_key\": \"1932755\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099418301302435890/IDU113cacc80159c21491318dca10779c4990d0b\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10686 Little Nomads Economic and Social Impacts of Migration on Children _Magdalena Cortina Juan Miguel Jimenez Sandra V. Rozo_ Development Economics Development Research Group January 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10686\\n\\n# **Abstract**\\n\\nThis paper reviews the main findings from 113 studies produced between 1990 and 2023, focusing on the impact of migration on various child groups affected through the migration path, including left-behind, migrant (voluntary and forced), and native children. The findings reveal that migration influences children’s outcomes in complex and context-dependent ways, and it interacts dramatically with household demographics and public policies. Key results include the following: (i) left-behind children benefit from remittances but experience dramatic declines in their cognitive and non-cognitive development due to parental absence; (ii) immigrant children generally fare better than those in their origin countries but still underperform compared to native children in host countries; and (iii) the impacts of migration on native children largely depend on the adjustment of public service supply to meet increased demand. In cases where education services expand to meet rising demand, the effect on native children can be minimal or even positive. This paper emphasizes the need for more experimental or quasi-experimental research to examine the effectiveness of programs that support migrant and minor host children, and it calls for longitudinal data collection to better understand the challenges and needs of migrant children, particularly in developing countries.\\n\\nThis paper is a product of the Development Research Group, Development Economics. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at sandrarozo@worldbank.org.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [{\"mention_name\": {\"text\": \"longitudinal data collection\", \"confidence\": 0.30963125824928284, \"start\": 1294, \"end\": 1322, \"clean_text\": \"longitudinal data collection\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1294, \"end\": 1322, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1294, \"end\": 1322, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1294, \"end\": 1322}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1294, \"end\": 1322, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.5503836870193481, \"start\": 384, \"end\": 412}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.30963125824928284, \"start\": 384, \"end\": 412}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5439035892486572, \"start\": 384, \"end\": 412}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5439035892486572, \"start\": 384, \"end\": 412}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# **Little Nomads: Economic and Social Impacts of Migration on Children**<sup>*</sup>\\n\\nMagdalena Cortina<sup>†</sup> Juan Miguel Jimenez<sup>‡</sup> Sandra V. Rozo<sup>§</sup> Originally published in the Policy Research Working Paper Series on _January 2024._ This version is updated on _February 2024._ To obtain the originally published version, please email prwp@worldbank.org.\\n\\n**JEL Classification:** F22, O15, R23 **Keywords:** Children, Migration.\\n\\n> *This research project received financial support from the Research Support Budget of the Development Vicepresidency of the World Bank. We are grateful to Catalina Amuedo-Dorantes for suggestions.\\n\\n> †Robert S. McNamara Fellow, World Bank. E-mail: mcortinatoro@worldbank.org\\n\\n> ‡Vancouver School of Economics. E-mail: jmjimen@student.ubc.ca\\n\\n> §World Bank, Development Research Group. Corresponding author, E-mail: sandrarozo@worldbank.org\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"“ _People made fun of me because I used different words; despite the fact we all spoke Spanish, they could tell something was different as soon as I opened my mouth. Hence, I learned to speak Colombian fast, I had to. It was not a choice._ ” Venezuelan migrant child in Colombia (age 16), Dec 2023\\n\\n# **I INTRODUCTION**\\n\\nApproximately 3% of the global population are migrants (World Bank, 2023). Of this share, almost 9% are children under 14 years of age, amounting to about 23.3 million individuals. While significant, these numbers do not account for all the children impacted by migration—including those left behind and native-born children in hosting countries—which highlights the relevance of this issue. It is crucial to understand the effects of migration on children because this significant event can profoundly shape life trajectories. Moreover, specifically for children whose origins or destinations are in developing countries, migration can alter their mobility out of or into poverty. Although migration destinations are balanced between the northern and southern hemispheres, only 15% of the world’s migrant children originate in the north, whereas 85% migrate from the south (World Bank, 2023). There is substantial research on the impacts of migration on adult migrants and hosts but considerably less regarding its effects on children.\\n\\nThis paper aims to provide a comprehensive review of research findings on the impact of migration on children’s development with an emphasis on insights from Economics and Psychology. It summarizes migration effects on various groups of children by category based on their relationship to the migratory journey, including:\\n\\n1. _Left-behind children:_ Children who remain in their home country while one or more parents migrate. We focus on how the absence of a parent impacts their outcomes.\\n\\n2. _Migrant children:_ Those who have migrated themselves or in some cases are second-generation migrants. We examine how the experience of relocating to a new environment affects their outcomes.\\n\\n3. _Native children:_ Non-migrant children in the host country who interact with migrant children. We assess how their exposure to peers from different backgrounds influences their outcomes.\\n\\n2\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"4. _Forcibly displaced children_ : These migrate due to involuntary conditions such as conflict and natural disasters. We explore their unique challenges, including trauma, disruption in education, and adaptation to new environments.\\n\\nThe review encompasses papers published in the top five economic journals and specialized journals in migration, education, and early childhood development.<sup>1</sup> We completed the search by consulting Google Scholar to capture any significant articles missed initially. Additionally, we examined references within each selected article’s literature review, focusing on the most relevant ones. This approach yielded a robust collection of 113 pertinent scholarly works.\\n\\nThis literature review discusses various methodologies, including experimental, quasi-experimental, and qualitative approaches, among others. Three papers employed experimental methods, specifically randomized controlled trials. Quasi-experimental designs were used in 95 papers, indicating causal studies not strictly experimental in nature. Within these, 51 papers used ordinary least squares, 19 used instrumental variables, 13 applied difference-in-difference approaches, 4 implemented regression discontinuity designs, and 25 incorporated fixed effects. Other methodologies included the Cox proportional hazard model, the Todd and Wolpin (2023) value-added model, the Blinder-Oaxaca decomposition technique, and survival analysis. Notably, only 3 qualitative studies were cited. (Refer to the appendix for detailed descriptions of each paper).\\n\\nThis review comprises five additional sections. Section II delves into research on the effects of migration on _migrant children_ , focusing on education, health, early childhood development, and the influence of immigration policy. Generally, this research shows that migrant children fare better than those in their origin countries but still fall behind native children in host countries. This gap often extends to poorer long-term labor outcomes, partly due to lower educational attainment and language barriers. Age at migration significantly influences the disparity between native and migrant children, and immigration policies are pivotal in narrowing or widening this gap.\\n\\nSection III examines the impact of migrant parents on their _left-behind children_ . The evidence reveals several key points. First, remittances usually benefit the education and physical health of these children by stabilizing income. This stability tends to reduce child labor, bolster school re-\\n\\n> 1 Included journals are the _International Migration Review_ , _Journal of Development Economics_ , _Journal of Population Economics_ , _Economics of Education Review_ , and _Early Childhood Research Quarterly_ .\\n\\n3\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"tention, and improve nutrition, potentially enhancing long-term economic well-being. Second, a parent’s migration often shifts intra-household dynamics—especially in bargaining—which positively influences girls’ educational achievements. Third, the absence of a parent due to migration severely affects children’s mental health and socio-emotional development. Finally, young children left behind experience a notable cognitive delay, primarily due to reduced parental involvement.\\n\\nSection IV describes the effects of migration on _native children_ . Most evidence for this population concentrates on education outcomes, native flight, and youth employment. The main findings suggest that the impact of migration on native children’s educational performance stems from the ability and willingness of educational institutions to adapt and expand resources in response to increased demand from migrant populations. Positive effects are observed in cases where resources adjust quickly to higher demand for education, and negative effects are observed when resources don’t change. Moreover, studies point to large flight effects among native students in schools where the share of immigrant students increases. Furthermore, concerning youth employment, there is some evidence of detrimental effects of youth migration on youth employment for Black Americans in the United States.\\n\\nSection V examines research regarding the effects of migration on forcibly displaced children. Work on refugees represents the majority of scholarship in this area. Refugee children are among the most vulnerable children worldwide. They not only endure the migration experience but do so involuntarily, and it typically accompanies traumatic events such as conflicts and natural disasters. Moreover, refugee children, especially adolescents, face a high risk of not returning to school since they often have to work inside or outside the household due to the extremely high vulnerability of their families. Existing research generally confirms that forcibly displaced children lag behind native children. It also establishes that—as in the case of voluntary migration—children’s outcomes depend considerably on how public service supply responds to higher demand. Emerging work has examined the effects of public programs that support refugee populations on children’s outcomes. These studies highlight promising outcomes from cash transfer initiatives, teacher training programs, language instruction, and programs that promote social cohesion.\\n\\nFinally, section VI discusses areas to advance research. A summary table with all the papers 4\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"reviewed in this study is presented in the appendix.\\n\\n# **II MIGRANT CHILDREN**\\n\\nThis section outlines the evidence concerning the impact of migration on migrant children. It reviews areas such as education, health, and early childhood development, and it highlights the critical role of immigration policies.\\n\\n# **II.A Education outcomes**\\n\\nExisting work in this area documents lower educational performance and attainment of immigrant children relative to native children. These differences have long-term consequences for immigrants’ economic success, including career choices and wages. Documented drivers of these disparities include higher socioeconomic vulnerability, parental decisions, stereotypes held by teachers, school sorting, and immigration enforcement.\\n\\n_Immigrants exhibit lower performance relative to hosts_ . Studies consistently indicate that immigrant children tend to perform worse than their native counterparts in terms of school GPA and standardized test scores. The performance gap is more pronounced among boys, whereas girls tend to perform relatively better (B ̈ohlmark, 2008).\\n\\nEarly economic research tested whether the age of arrival plays a crucial role in immigrant children’s education outcomes. The premise was that the age of arrival correlates with a stage of children’s physical and cognitive development in which younger immigrants adapt and learn faster due to the brain’s plasticity at early ages. These studies showed a noticeable decline in performance among children who arrive after a certain age threshold (typically around 7–9 years old); this decline is particularly worse on verbal tests (Bleakley and Chin, 2010). Specifically, B ̈ohlmark (2008) finds a strong negative impact of immigrating after age nine on immigrant children’s GPA in Sweden, and Cahan _et al._ (2001) show that starting at the arrival age of seven, there was a monotonic decrease in performance for immigrant children in Israel between 1952 and 1970, with worse results on verbal tests than on mathematical ones. Similarly, Cortes (2006) shows that the test score gap between immigrant children and second-generation children narrows according to how long the former have been residing in the United States.\\n\\nMoreover, research on this topic underscores the importance of the critical period for language 5\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"acquisition, emphasizing that the timing of arrival can significantly impact the integration of immigrant children when they become adults. Both Aslund _et al._ (2015) and Bleakley and Chin (2010) find in Sweden and the United States, respectively, that English proficiency (measured indirectly by the age of childhood arrival) increases the probability of marrying a native, having a higher fraction of work colleagues who are natives, and residing outside of an ethnic enclave.\\n\\nImmigrant children also face disparities in terms of school attendance rates and attainment. Colding _et al._ (2009) find that the dropout rates of children of immigrants in Denmark who are enrolled in vocational upper secondary school are higher than those of natives. Additionally, McKenzie and Rapoport (2011) show that living in a migrant household decreases children’s likelihood of completing school by 13%–15% in the Mexican context.\\n\\nSimilar to performance, the age at which immigrant children arrive can impact their attendance and attainment levels. In Germany, Lemmermann and Riphahn (2018) show attainment decreases for children who arrive after age six, suggesting immigrant children face more language barriers after this threshold. In fact, Foged _et al._ (2022) and Foged _et al._ (2023) find that language training offered by the Danish government to immigrant adults has significant intergenerational spillovers on their children in terms of higher completion rates from lower secondary school.\\n\\nFinally, there is evidence that lower educational performance of immigrant children has long-term consequences, particularly in terms of career choices and wages. In Italy, Barban and White (2011) observe that immigrant children are more likely to enroll in vocational and polytechnic schools due to inferior outcomes on the middle school exam. Moreover, research on immigrant children’s labor outcomes consistently shows that adults who arrive at an older age during the immigration process tend to have lower earnings in their host country (B ̈ohlmark, 2009). Language skills are a critical channel through which age at arrival influences adult earnings. Bleakley and Chin (2004) suggest that immigrants who arrive later in the United States may have lower language proficiency, which can significantly hinder their ability to secure higher-paying jobs and integrate into the labor market effectively.\\n\\n_Potential driver 1: parental investment decisions_ . Another group of articles has explained the consequences of migration on the parental investment decisions of migrants concerning the human 6\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"capital of their children. The main finding is that the socioeconomic status of migrant families upon arrival is on average lower than that of native families, which directly translates into fewer monetary and time investments in their children. For instance, Rangvid (2007) finds that 50% of the gap between immigrant and native students in Denmark is explained by differences in the family socioeconomic status of each comparison group. Likewise, Schnepf (2007) finds in a crosscountry analysis of OECD members that in European countries, socioeconomic background is the main determinant for the immigrant-versus-native performance gap in tests like PISA, TIMSS, and PIRLS. Furthermore, Dahl _et al._ (2020) show that in Germany, migrant parents from backgrounds with traditional cultural norms (e.g., Muslims) tend to invest less in their daughters’ schooling. Consequently, their daughters receive worse grades in school.\\n\\n_Potential driver 2: school sorting._ Another explanation for the lower performance of immigrant children compared to native peers is the potential sorting of this population into schools with unfavorable characteristics. Rangvid (2007) finds that even though immigrant students in Denmark have access to school resources comparable to native students—including class size, student-teacher ratio, language lessons per week, and physical and educational infrastructure—other factors contribute to the immigrant students’ low performance on the PISA test. These determinants include teachers’ expectations, encouragement, and pressure to achieve, as well as a scarcity of specialized teachers compared to the number available for native students of the same socioeconomic status. Also, Liu _et al._ (2010) note that immigrant families often locate in areas with lower-quality schools than those attended by native counterparts.\\n\\n_Potential driver 3: stereotypes held by teachers._ Recent literature has shown how teachers’ perceptions and stereotypes can influence immigrant children’s education outcomes. Triventi (2020) finds that Italian teachers tend to assign lower grades to students with an immigrant background compared to native students, suggesting that teachers may harbor implicit biases. Alesina _et al._ (2018) confirmed that intuition, finding that Italian teachers who hold stronger stereotypes (higher Implicit Association Test scores - IAT) assign lower grades to immigrant students compared to native peers. However, teachers increase grading when they are randomly informed about the stereotypes they hold.\\n\\nSimilarly, Carlana _et al._ (2022) find that Italian teachers give inaccurate recommendations about 7\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"educational careers to immigrant children, who consequently enroll disproportionately in vocational high schools, as opposed to natives who go more to technical and academically oriented high schools. The behavior is particularly observed among female immigrant students (Aktas ̧ _et al._ , 2022). Nevertheless, the authors show that when career counseling is randomly provided to highly skilled immigrant students, their likelihood of enrolling in the high track matches that of native students.\\n\\n# **II.B Health outcomes**\\n\\nEvidence in this area indicates that migrant children generally have better health outcomes than children in their origin countries. However, they still lag behind native children. Additionally, the evidence shows that access to health services is key to enhancing migrant children’s health. This leads to more efficient health care utilization, with greater use of preventive services and less reliance on emergency services.\\n\\n_Health outcomes_ . Migrant children often exhibit better anthropometric measures, such as weightfor-age (WfA) and height-for-age (HfA) than children from their places of origin, yet they experience higher mortality rates compared to native children. Cockx (2018) demonstrates that children who relocate from rural to urban areas in Tanzania exhibit higher HfA z-scores and reduced stunting rates than those from sending areas. The study suggests this improvement may stem from an increase in maternal income that in turn contributes to improved long-term food security. However, in addition to the income effect hypothesis, there is evidence that the improvement in HfA and stunting rates might be partly due to changes in parenting nutritional habits. Specifically, Stillman _et al._ (2012) find that dietary change is one of the main determinants of the improvement in anthropometric measures for immigrant children aged three to six years in New Zealand. However, the authors also note that these children face a higher risk of obesity compared to peers in their places of origin.\\n\\nInterestingly, Thomas (2007) finds that in the lowest quintiles of the socioeconomic distribution in South Africa, migrants have a higher likelihood of child mortality than natives. However, in the highest socioeconomic quintiles, child mortality is more prevalent among the native population.\\n\\n_Health service access_ . Access to health care is crucial for improving migrant children’s health out8\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"comes. Bronchetti (2014) and Currie (2000) find that changes in health insurance eligibility in the United States reduce the likelihood that a child in an immigrant family will go without a doctor’s visit for more than 12 months by 7 to 12 percentage points (pp). They also decrease the likelihood of an emergency room (ER) visit in the past year by 4 to 6 pp. These results collectively suggest that eligibility for public insurance leads to increased utilization of more efficient health care services such as preventive and ambulatory care while reducing reliance on costly ER visits for children in immigrant families. Furthermore, the findings also indicate that eligibility for public insurance may lead to modest improvements in certain child health outcomes that typically respond to ambulatory or preventive care. These include, for example, a lower likelihood of experiencing an asthma attack in the past 12 months (Bronchetti, 2014).\\n\\n# **II.C Early childhood development**\\n\\nLiterature on the early development of immigrant children is very limited. Available evidence in Psychology suggests that immigrant children frequently exhibit disparities compared to their native counterparts, with potential delays in cognitive and language skills (De Feyter and Winsler, 2009).\\n\\nRegarding language proficiency among young immigrant children, Palermo and Mikulski (2014) show that factors such as preschool attendance significantly influence the acquisition of communication skills. They observe that positive peer interactions and increased levels of English exposure during young ages positively correlate with the English vocabulary and letter-word skills of Spanish-speaking immigrant children in the United States.\\n\\n# **II.D The central role of immigration policy**\\n\\nResearch often investigates how immigration policies impact migrant children’s outcomes. The findings consistently show that strict enforcement against immigrants harms these children’s outcomes. Conversely, policies that improve service access and integration typically lead to positive outcomes for them.\\n\\n_Deportation enforcement_ . The impact of deportation enforcement measures on immigrant children is a subject of broad interest, especially in the U.S. context. Regarding the effect on early childhood development, the enforcement of immigration laws can lead to a decline in the English9\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"language skills of these children (Arenas Arroyo and Schmidpeter, 2022). Moreover, Santillano _et al._ (2020) also find that in the United States, enrollment of Hispanic immigrant children in the Head Start program—which provides comprehensive early childhood education, health, and nutrition—decreases by 10% after a law enforcement raid. Primarily, this decline occurs because parents are deterred from appearing in public spaces.\\n\\nConcerning the education outcomes of immigrant children, Amuedo-Dorantes and Lopez (2015) and Amuedo-Dorantes and Lopez (2017) find that in U.S. areas with intensified immigration law enforcement, the probability of school-year repetition and dropping out for Hispanic children between ages 14 and 18 years increases by 14% and 18%, respectively. Likewise, Dee and Murphy (2020) find that new ICE partnerships lead to a 10% decrease in the Hispanic immigrant student population within a span of two years. Furthermore, the learning environment plays an important role in immigrant children’s motivation, focus, and parental involvement. Amuedo-Dorantes _et al._ (2023) use data from the “Between the Lines” project to show that immigrant students who are enrolled in U.S. school districts implementing safe-zone policies<sup>2</sup> demonstrate improvements in several outcomes. For instance, they are 17% less likely to report difficulties in concentrating during school hours, 65% more inclined to express diligence in their studies, and 31% more optimistic about their prospects of achieving higher education, relative to other students. Additionally, these policies enhance caregiver participation in parent-teacher meetings by 34%.\\n\\nSimilar results are observed in the health care domain. Particularly, Watson (2014) finds that an increase in federal immigration enforcement acts leads to a decrease in Medicaid enrollment and usage among children of non-citizens, even if the children themselves are U.S. citizens. Additionally, Amuedo-Dorantes _et al._ (2022) find that more intense local U.S. immigration enforcement measures, particularly during the third trimester of pregnancy, raise the probability of infants being born with low birth weight (less than 2,500 grams). They note increased maternal stress and a decrease in prenatal-care medical visits as underlying factors.\\n\\nIn terms of economic and family impacts, Amuedo-Dorantes _et al._ (2018) show that deportation enforcement raises the likelihood of children living in impoverished households by 4%, primarily due to reduced working hours among parents. Furthermore, Amuedo-Dorantes and Arenas-\\n\\n> 2A safe-zone policy within a school district implies that immigration authorities have limited entry or access.\\n\\n10\", \"datasets\": [{\"mention_name\": {\"text\": \"Between the Lines\", \"confidence\": 0.348238468170166, \"start\": 1122, \"end\": 1139, \"clean_text\": \"Between the Lines\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1122, \"end\": 1139, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1122, \"end\": 1139, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1122, \"end\": 1139}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1122, \"end\": 1139, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.5846769213676453, \"start\": 309, \"end\": 326}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.348238468170166, \"start\": 309, \"end\": 326}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8017099499702454, \"start\": 309, \"end\": 326}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8017099499702454, \"start\": 309, \"end\": 326}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"Arroyo (2019) show a concerning 19% increase in family separations, particularly affecting children whose parents are undocumented and/or who are sent to live with relatives or friends threatened by deportation.\\n\\n_Public service access_ . Foged _et al._ (2023) underscore that refugee children in Denmark benefit from higher completion rates when their parents are exposed to language training programs, suggesting a positive correlation between language skills acquisition within refugee families and educational success. In a related context, Foged _et al._ (2022) indicate lower crime rates among refugee populations, possibly as a consequence of improved access to educational opportunities and integration programs. Furthermore, studies in the U.S. context by Bronchetti (2014) and Currie (2000) emphasize the importance of eligibility for public health insurance, revealing that it increases child preventive care utilization and positively impacts health outcomes such as the management of conditions like asthma.\\n\\nMoreover, Villarreal and Gonzalez (2016) show that participation in extracurricular activities in U.S. schools predicts positive social behaviors for immigrant children. Specifically, participation in sports-related activities is associated with an increase in school membership feelings of Hispanic immigrant children and higher peer prosocial orientation behavior. Nevertheless, treatments might not induce homogeneous effects across population groups. For instance, Dahl _et al._ (2020) find that girls from traditional cultures are socially isolated and less likely to self-identify as German, which suggests they might need tailored support.\\n\\nFinally, evidence in the economic field shows how targeted interventions can mitigate higher criminal propensity of migrant youth as documented in Vazsonyi and Killias (2001). Foged _et al._ (2022) and Foged _et al._ (2023) highlight that language training programs for refugees can lead to a reduction in crime rates among the children of treated parents. Both articles suggest that equipping parents with language skills contributes to keeping children in schools and possibly increases the opportunity cost of criminal activities for immigrant children.\\n\\n_Citizenship rights_ . Felfe _et al._ (2020) find that the introduction of birthright citizenship for immigrants in Germany is associated with increased attendance in non-compulsory educational stages such as preschool and secondary school. Moreover, it also caused immigrant children to progress 11\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"faster through primary school and raised the likelihood they would pursue the academic track of secondary school. All these effects underscore the positive impact of citizenship on immigrant children’s educational trajectories and opportunities. Furthermore, Felfe _et al._ (2021) delve into the social dimension, revealing that complete citizenship rights in Germany can foster prosocial behavior, including out-group cooperation. This finding suggests that individuals who enjoy full citizenship rights may exhibit a higher level of engagement and cooperation with members of different social or cultural groups, potentially facilitating assimilation into the broader community.\\n\\nNevertheless, Dahl _et al._ (2020) highlight that the birthright citizenship policy in Germany can have a differential effect on immigrant girls from traditional cultures, leading to a decrease in their measures of life satisfaction and self-esteem by approximately 0.32 and 0.25 standard deviations, respectively. As noted above, the study also reveals that Muslim girls who acquire birthright citizenship exhibit reduced integration into German society. The authors suggest that immigrant Muslim parents invest less in educating their daughters, resulting in lower academic performance among girls born after the implementation of the birthright citizenship reform.\\n\\n# **III LEFT-BEHIND CHILDREN**\\n\\nThe term left-behind children refers to minors who remain in their places of origin or permanent residence while one or both parents migrate to another country, typically for employment purposes. This circumstance often arises due to financial constraints, legal impediments, or concerns regarding the children’s welfare that prevent the parents from bringing them along. Additionally, it encompasses situations wherein parents relocate to urban areas while their children remain in rural residences. These children are typically cared for by a lone parent, other family members, acquaintances, community members, childcare institutions, or independently. In many instances, parents provide financial support through remittances while experiencing prolonged physical separation. This scenario produces a range of positive and negative impacts encompassing cognitive, emotional, psychological, physical, and social aspects. This section summarizes the work about the effects of a migrant parent on left-behind children in terms of education, health outcomes, and early childhood development.\\n\\n12\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"# **III.A Education outcomes**\\n\\nCurrent evidence highlights two main findings. First, remittances generally have a positive impact on the education outcomes of left-behind children through the stabilization of household income. However, in households where remittances fail to stabilize income, child labor tends to increase, leading to lower educational attainment for left-behind children. These effects have significant consequences and have been shown to extend into long-term changes in economic wellbeing. Second, when a parent migrates, it often leads to changes in intra-household bargaining dynamics for the remaining household members. Typically, these changes positively influence girls’ educational achievement.\\n\\n_Remittances_ . Remittances are the most prevalent determinant of improvements in educational outcomes. The most common documented channels include increased household income and higher financial stability that lead to a reduction in child labor and an increase in school retention (Amuedo Dorantes _et al._ , 2010). Edwards and Ureta (2003), for instance, find that remittances sent to El Salvador lower the risk of school dropout in rural areas. Kandel (2003) finds that sibling internal migration in Mexico was associated positively with the educational attainment of leftbehind children. Furthermore, Wassink and Viera (2021) show that Mexican left-behind children are more likely to complete lower-secondary school, enter upper-secondary school, and complete upper-secondary school. Notably, these associations are most pronounced among children whose parents did not complete primary school and those who reside in rural areas.\\n\\nWhen remittances fail to substantially increase household financial stability, in the short run, leftbehind boys increase working hours and reduce educational attainment. Antman (2011) finds that in Mexico, the most affected group is boys aged 12 to 15 years. This highlights the immediate impact of parental migration on the educational and labor trajectories of these children. Moreover, the reduction in human capital accumulation that results from these adjustments can lead to lower income levels in adulthood, as observed in the Chinese context (Feng _et al._ , 2022). This suggests that the consequences of parental migration during childhood can extend into the individual’s later years, affecting their economic well-being in the long run.\\n\\n_Changes in intra-household bargaining_ . Another important channel relates to intra-household bar13\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"gaining after one parent migrates. Antman (2012) observes a reallocation of resources among siblings when the male head of household migrates. Thus, the female parent allocates more resources to girls and a lower fraction to boys, increasing the educational attainment of the former group by up to one year. These results underscore a significant connection between international migration and gender-based discrimination within the household, empowering spouses to exert greater control over the allocation of resources and thereby favoring investment in girls’ education. Similarly, Saleemi (2023) finds that in Pakistan, when the father is absent, women participate more in the resource allocation decision and thus expenditures on girls increase by 31%. Lee and Park (2010) document similar results for the Chinese context.\\n\\n# **III.B Health outcomes**\\n\\nResearch on the effects of a migrant parent on left-behind children generally indicates positive outcomes for physical growth, which are associated with the positive income effect caused by remittances and better nutritional practices. In contrast, the parental absence is highly detrimental to children’s mental health and socioemotional skill development.\\n\\n_Anthropometrics_ . Overall, the evidence shows that parental migration has positive effects on anthropometric measures of left-behind children. Hildebrandt _et al._ (2005) show that in the Mexican context, this is due to two factors. First, remittances allow households to spend additional resources on food and health services. Second, the authors argue that migrants learn better care and nutritional practices in the United States that later transfer to their children in Mexico. Despite these positive effects, left-behind children are less likely to be breastfed, fully vaccinated, or taken to the doctor in the first year of life. Likewise, Mu and De Brauw (2015) find that in China, parental migration improves the WfA z-score because an increase in income translates into better access to tap water for left-behind children. Nonetheless, there is evidence that the income effect has a differential impact by gender. For instance, Vikram (2023) observes that in India, the effect of parental migration is greater for boys than for girls, as father’s migration is associated with higher HfA z-scores only for boys.\\n\\nIn terms of measures at the time of birth, Khan _et al._ (2023) find that women in Bangladesh with a migrant spouse are more likely to have had prenatal care, but this is not linked to delivering at a professional health care facility or the presence of a qualified attendant for birth labor. Neverthe14\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"less, there is also evidence that remittances alone are insufficient to increase left-behind children’s utilization of qualified health care. In fact, some results are puzzling: among children who report a recent illness, Cambodian households that receive remittances are less likely to utilize qualified health care providers (Treleaven, 2019).\\n\\n_Mental health_ . Most research suggests that left-behind children exhibit poorer measures related to mental health outcomes compared to other children. Lee and Park (2010) show that left-behind children in China demonstrate negative effects with regard to psychosocial well-being, while Zheng _et al._ (2022) find they have significantly higher levels of depression than migrant children. This difference is more prominent for boys than for girls. Many of the articles attribute these results to the deterioration of the parent-child relationship that occurs when families are separated. For instance, Amuedo-Dorantes _et al._ (2023) find that immigrant children in the United States whose parents have been deported are more likely to experience post-traumatic stress disorder (PTSD) symptoms as well as higher levels of internalizing (anxiety and depression) and externalizing (aggression) problems relative to children with legal resident parents or even children of undocumented parents who have not encountered immigration enforcement.\\n\\nMoreover, Wang and Liu (2021) and Wang and Zhu (2021) find adverse spillover effects on the mental health of classmates when there is a higher proportion of left-behind children in the class. This is primarily attributed to a deterioration in the learning environment within Chinese schools. In fact, the authors observe a greater occurrence of peer victimization (bullying) and even a rise in cases of self-injury.\\n\\nEvidence on this front also shows that parents’ absence can lead to severe psychological development issues, an increase in risk-loving behavior, and lower career aspirations. Lee and Park (2010) find that Chinese left-behind children tend more to experience drastic changes in behavior, becoming more anxious and depressed and more prone to impulsivity and aggression. In addition, Kandel and Kao (2001) find that left-behind children tend to have lower career aspirations, which are associated with a reduced desire to pursue university degrees. Moreover, Adunts (2021) observes that in Ukraine, children with a father abroad exhibit lower levels of perseverance compared to other children. This is evident in their tendency to avoid choosing challenging tasks with potentially high rewards. Liu _et al._ (2021) show that, in the long run, Chinese left-behind 15\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"children with migrant mothers tend to exhibit reduced levels of conscientiousness and grit along with higher levels of neuroticism (or lower emotional stability). In the short term, when mothers migrate, children typically display decreased levels of conscientiousness, agreeableness, and openness. Furthermore, Luo and Wang (2023) also observe that Chinese left-behind children show less prosocial behavior, which encompasses attributes such as fairness, trust, trustworthiness, and cooperation.<sup>3</sup> Finally, Su _et al._ (2013) find that infrequent parent-child communication correlates with higher feelings of loneliness in Chinese left-behind children.\\n\\nThe limited available evidence also indicates a correlation between parental absence due to migration and an increased probability of criminal behavior by left-behind children later in life. According to Cameron _et al._ (2022), this association is observed among Chinese adult males. The authors suggest that childhood experiences of being left behind contribute to diminished educational achievements and a tendency toward risk-taking behavior, both of which are considered contributing factors toward criminal activities.\\n\\n# **III.C Early childhood development**\\n\\nThe general consensus and evidence are that parental migration causes cognitive delay in leftbehind children. For instance, Bai _et al._ (2022) show that maternal migration in the Chinese context leads to a 6 pp reduction in children’s cognitive skills compared to children in the same village without a migrant parent, and Yue _et al._ (2016) also find negative effects on cognitive development and dietary quality. Even though rural-to-urban migration in China has been key to economic growth, it might entail a significant human capital cost for the next generations.<sup>4</sup> In both studies, the primary determinants of this decline are decreases in parental time and the quality of investment in stimulating activities, along with a reduction in dietary diversity. Similarly, in Mexico, Powers (2011) estimates a structural model to argue that the drop in cognitive skills among left-behind children stems from a shift in their allocation of time towards activities unrelated to cognitive development. Furthermore, Zheng _et al._ (2023) find that left-behind children in China are not only at a disadvantage compared with other children in the same village, but also com-\\n\\n> 3Notwithstanding the preceding point, the paper underscores the positive impact of having access to one parent.\\n\\n> 4In the Chinese context, the reallocation of labor from rural to urban areas has been a key driver of China’s prosperity in recent decades. Because of hukou reforms, migrants face limitations in accessing fundamental public services like public education for their children. As a result, migrant parents are compelled to leave their children in their hometowns, producing tens of millions of “split families” and “left-behind children.”\\n\\n16\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"pared with migrant children. They argue that parental supervision and closeness are a vital part of children’s cognitive development.\\n\\nContrary to the psychological evidence presented earlier, the limited economic evidence offers a different perspective. Specifically, in Nicaragua, Macours and Vakis (2010) find that having a migrant father in the household has no significant impact on cognitive skills, whereas there is a positive effect when the mother migrates. The authors suggest that female empowerment enables a higher proportion of remittances to be allocated to children’s development.\\n\\n# **IV NATIVE CHILDREN**\\n\\n# **IV.A Education outcomes**\\n\\nPrevious work indicates migration has varied impacts on the educational performance of native children. Effects are generally negative in situations where increased demand for educational services (driven by immigrant flows) does not meet a corresponding increase in supply. This discrepancy often results in overcrowding and strains educational resources, adversely affecting native students. On the other hand, in regions where resources are augmented in tandem with the arrival of immigrants, effects on native children’s education outcomes are either negligible or positive. This suggests that the key underlying factor is the ability and willingness of educational institutions to adapt and expand resources to meet greater demand from immigrant populations.\\n\\n_Negative effects_ . Studies exploring the adverse effects of migration on native children’s education outcomes identify two primary factors: increased classroom overcrowding and a decline in the quality of the learning environment. The degradation of the learning environment is evident in several cases. For example, Imberman _et al._ (2012)’s study on Hurricane Katrina evacuees reveals that their arrival in Houston led to higher absenteeism and more disciplinary issues in secondary schools. Ohinata and Van Ours (2013)’s research suggests that Dutch schools experience more bullying incidents when the share of immigrants per classroom is higher. Similarly, Figlio and ̈Ozek (2019) find that after the influx of Haitian migrants to Florida, disciplinary incidents and school suspensions rose. Nevertheless, the last two articles do not find strong evidence of negative spillover effects on the academic performance of native children.\\n\\nConcerning the overcrowding channel, studies suggest that an excessive number of students per 17\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"classroom is associated with a decline in the performance of native students, particularly when there is no compensatory investment to address new educational demand in schools that receive a substantial influx of immigrants. Brunello and Rocco (2013), for example, study the case of 19 countries where immigration waves doubled the number of students in public secondary schools between 2000 and 2009. This shock led to a decrease in PISA scores for natives. Furthermore, these negative effects are shown to be concentrated among male students, children with disadvantaged backgrounds, and underperforming students. Similarly, Green and Iversen (2020) observe that negative effects are concentrated in students at the most risk of underperformance in Norwegian schools. Finally, Contreras and Gallardo (2022) provide evidence that Venezuelan migration led to a decrease in math and reading standardized test scores for male native students in Chile. As a consequence, there has been an increase in native flight by high-skilled students to private schools.\\n\\n_Positive effects_ . Two key mechanisms are documented in the set of studies identifying the positive impacts of migration on native children’s education outcomes. The first is the beneficial spillover from having high-achieving immigrant peers, contingent on the availability of compensatory school funding (Imberman _et al._ , 2012). For instance, Morales (2022)’s research in Georgia illustrates this, showing that a 1% increase in refugee students boosts native math scores by 0.01 standard deviation, an effect amplified by increased funding for schools. In the U.S. context, these positive spillovers predominantly benefit students from disadvantaged backgrounds. This improvement is attributed to exposure to immigrant children who are less disruptive and exhibit better behavior, as noted by Figlio _et al._ (2021). Similarly, in T ̈urkiye, Tumen (2021) finds that positive impacts on native students stem from competitive incentives. This competition seems to particularly motivate underperforming native students to improve their performance.\\n\\nThe second significant migration-related factor that contributes to positive educational outcomes for native students involves increased perceived returns to education following a rise in immigrant student populations. Hunt (2017) highlights this in the U.S. context, where white students showed a higher likelihood of staying in high school and completing 12 years of education in response to increased immigrant presence. This trend was also observed in T ̈urkiye, where native youth noticeably shifted from balancing work and school to focusing solely on education.\\n\\nHowever, it is still unclear which specific sociodemographic group benefits most from this shift in 18\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"educational values. C ̧ akır _et al._ (2023) present evidence that the most advantaged group comprises native students with more educated parents, while Tumen (2018) argues that it is male students with less-educated parents.\\n\\n_Neutral effects_ . A set of articles documents the negligible effects of migration on the educational outcomes of native children, particularly when adequate compensatory funding and appropriate selection mechanisms are in place. For example, Figlio and Ozek<sup> ̈</sup> (2019)’s research on the arrival of Haitian refugees in the United States exemplifies this. The study suggests that schools effectively accommodated the influx of students without detriment to natives, partly due to preemptive measures like employing Haitian-Creole-speaking counselors. Similarly, findings by Van der Werf (2021) further support this argument for the resettlement of Vietnamese students in the United States. The study finds that schools, when provided with additional resources for each Vietnamese student, could prevent the diversion of funds from native students and avoid the negative effects of overcrowding. Lastly, Assaad _et al._ (2023)’s research on Jordanian schools during the Syrian refugee crisis reveals effective governmental strategies. The opening of new school shifts and the establishment of schools in refugee camps were key measures that successfully mitigated the potential adverse effects of overcrowding. These studies collectively suggest that with thoughtful planning and sufficient resources, schools can manage increased student populations due to immigration, ensuring that the education outcomes of native students are not affected.\\n\\nMoreover, the negative correlation observed in raw data between an increase in immigrant students and the academic performance of native students can often be attributed to selection issues. Immigrants tend to be placed in areas with schools possessing less desirable characteristics, as documented in Van der Werf (2021). Geay _et al._ (2013)’s work in England corroborates this argument, finding that immigrant students are typically located in schools with a higher proportion of underperforming native students and a higher teacher-student ratio. By accounting for these factors, the apparent negative impact of immigration on the education outcomes of native students becomes negligible. These results suggest the initial perceived detrimental effects are more due to the characteristics of schools where immigrants are enrolled, rather than the presence of immigrant students per se.\\n\\n19\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"# **IV.B Native flight from schools**\\n\\nIn the context of education, native flight occurs when native parents respond to immigrant inflows by switching their children from public to private schools, or even to public schools with a lower share of immigrant children. This phenomenon often leads to residential and educational isolation of immigrants by reducing native demand for public schools (Cascio and Lewis, 2012).\\n\\nIn the United States, results from various studies suggest native flight is significant. For instance, Betts and Fairlie (2003) estimate that for every four new immigrants in a public school, one U.S.born student switches to a private school. Similarly, Cascio and Lewis (2012) calculate that a 10% increase in immigrant children leads to a 7.8% decrease in native students’ enrollment in Californian public schools. These effects seem to be driven by white children and worsen when immigrants are not English speakers at home. These findings align with the strong selection of U.S.-born students out of schools with higher proportions of immigrants. This highlights why these compositional effects can obscure the interpretation of regression analyses that do not consider this sorting mechanism (Figlio _et al._ , 2021).\\n\\nThis phenomenon is not unique to the U.S. context. In Denmark, Gerdes (2010) shows how in small municipalities a rise in the share of immigrants increases the propensity of Danish parents to enroll children in private schools.<sup>5</sup> The author suggests that segregation is easier in small municipalities than in large ones where enrolling a child in a private school can be more difficult. Similarly, in Per ́u, Mart ́ınez (2022) finds that native flight is used as a mitigation strategy by parents to counteract the negative effects of immigration on incumbent students. Specifically, they discover that students who switch to private schools tend to be male students with low academic achievement in primary school or female students with high academic achievement in secondary school. Furthermore, cream-skimming effects are also observed in Spain, where more educated native households with higher-achieving children are more likely to switch them to private schools in response to immigration (Farre _et al._ , 2018).\\n\\n> 5The article refers to private schools as free schools.\\n\\n20\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"# **IV.C Youth employment**\\n\\nIn addition to studies focusing on the impact of immigration on youth employment rates through changes in returns to education, as documented by C ̧ akır _et al._ (2023), Hunt (2017), and Tumen (2021), some articles investigate the effects of immigration on labor markets independent of education outcomes. In the U.S. context, for example, Smith (2012) documents that one primary factor contributing to the youth employment rate decline among Black teenagers was the increase in immigration between 1990 and 2005. Specifically, the author estimates that a 10% increase in the share of low-skill immigrant workers led to a 5 pp decrease in the employment rate of Black teenagers. Nevertheless, Winegarden and Khor (1991) argue that this declines in youth employment rates does not emerge from an increase in undocumented immigrants. They posit that these immigrants are not a labor substitute for native youths since the former usually rely on labor markets most favorable to the undocumented population to find a job.\\n\\n# **V FORCIBLY DISPLACED CHILDREN**\\n\\nAt the end of 2022, there were more than 108 million forcibly displaced individuals in the world; 20% were children below 17 years of age (UNHCR, 2023). This section reviews the research concerning the impacts of forced migration on children. The distinction between forced and voluntary migrants hinges on the reasons for relocation. While voluntary migrants relocate by choice, forced migrants move due to involuntary factors such as conflicts, natural disasters, or other compelling circumstances. Children who are forcibly displaced merit specific attention due to their heightened vulnerability. They frequently endure violence and disasters, experience traumatic and abrupt migration, leave everything behind, and arrive in environments characterized by uncertainty about their rights and potential discrimination.\\n\\nRecent evidence on the effects of forced migration largely focuses on child refugees and international forced migrants.<sup>6</sup> Therefore, we limit our review to this group. This section includes three parts. The first part examines the impacts of international forced migration on children. The evidence suggests that refugee children are extremely vulnerable and lag behind native children in terms of education and health outcomes (at least in the short term). Moreover, refugee children—\\n\\n> 6A notable exception is Chiovelli _et al._ (2021), who examine the effects of conflict-driven displacement on human capital for Internally Displaced Individuals in Mozambique. The authors find that displaced children register higher investments in education.\\n\\n21\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"especially adolescents—face high risk of school dropout, as their time could be used to generate income, support household chores, or enter into marriage.\\n\\nThe second part investigates how refugee children affect native populations. Results suggest that, as in the case of voluntary migration, the impact of refugee migration on natives largely depends on policy responses to address the consequent challenges. Yet, underperforming native students are at greater risk of being negatively affected by the potential overcrowding effect caused by the arrival of new refugee students in hosting communities. The last part examines findings on interventions to enhance outcomes for both refugee and native children.\\n\\n# **V.A Impacts of forced migration on refugee children**\\n\\nThis section describes work concerning the disparities between refugee and native children in outcomes such as education, health, and child labor. Native children are the most common comparison group for refugee children as it is difficult to obtain data on peers in their places of origin. The majority of scholarship suggests that refugee children lag behind native children.\\n\\n_Education._ Krafft _et al._ (2022) present evidence on the educational outcomes of Syrian refugee children in Jordan. The study reveals that these children are less likely to enroll in Jordanian schools compared to native children of similar socioeconomic status. Additionally, Syrian refugee children, once enrolled, have a higher propensity to drop out. This trend is more pronounced among boys, primarily because they often seek employment to alleviate their families’ financial burdens. However, older girls also face significant disadvantages. They are incentivized to leave school due to increased household responsibilities or the prospect of marriage.\\n\\nWhen it comes to comparing educational disparities among refugee children, Ginn (2018) evaluates the educational gaps between Syrian refugees located in camps and those outside. The author documents that enrollment and attendance rates for children in camps are higher. These results are partly attributed to the direct establishment of new schools in camps, coupled with higher transportation costs and enrollment fees at schools outside camps. Interestingly, increased school attendance in camps does not necessarily equate to better academic achievement. This could be explained by the higher repetition rates in schools in camps, indicating a possible concentration of underperforming students in these schools compared to those outside, and the challenges of 22\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"hiring and training teachers for new schools in camps, as highlighted by Salemi _et al._ (2018).\\n\\n_Health._ Demirci _et al._ (2022) study the disparities in health and nutrition among native and Syrian refugee children in T ̈urkiye. Although the authors do not observe differences in infant or child mortality between refugee children born in T ̈urkiye and native children, they do find that refugee infants born in T ̈urkiye have lower birth weight, age-adjusted weight, and age-adjusted height than native infants. When decomposing the effects of weight at birth before and after migration, they find the difference before relocation is greater, suggesting that the persistent adverse effect of being born in a crisis outweighs the possible lack of health supply in the destination country.\\n\\nOther research has studied the impacts of refugee camps on health outcomes. Rashad _et al._ (2018) find that in Jordan, Palestinian children in refugee camps have lower age-adjusted height scores relative to refugee children living outside camps. The study attributes this difference primarily to the lower asset levels of families in the camps. This economic disadvantage, in turn, leads to reduced investment in the developmental needs of their children.\\n\\nLastly, according to S ́anchez-Ariza _et al._ (2023), children in households internally displaced by the Colombian conflict exhibit lower mental health levels than other children. This assessment is based on measurements using the Symptom Checklist-90 and the Child Trauma Symptom Checklist for Young Children scales. Interestingly, the authors suggest this phenomenon also extends to children who were not directly affected by the conflict but whose caregivers experienced its consequences. This implies the possibility of an intergenerational transmission of the adverse effects of displacement on children, highlighting the need for intervention.\\n\\n_Child labor._ The issue of child labor among refugee children, particularly its extent and impact, remains underresearched due to challenges in gathering reliable data. However, it is evident that refugee children are more susceptible to entering the child labor market, although the magnitude of this phenomenon varies according to the methodologies employed to measure child labor. Sieverding _et al._ (2018) note that in T ̈urkiye, Syrian refugee boys aged 10 to 14 years are 2% more likely to engage in paid work than their peers. On the other hand, Dayıo ̆glu _et al._ (0) find significantly higher incidence of child labor among Syrian refugee boys in Jordan. The study reports that refugee boys aged 12 to 14 years are 17% more likely to be involved in child labor than native 23\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"boys. This discrepancy becomes even more pronounced in the 15 to 17 year-old group, where the likelihood increases to 45% compared to native counterparts, highlighting a grave concern regarding the involvement of older refugee boys in the labor market.\\n\\n_Child marriage._ Forced displacement and conflict often prompt gains in child marriage. Parents may decide to marry off their daughters for reasons such as family insurance, protection, increased freedom, and adherence to cultural norms. However, research on this issue is quite limited, mostly emerging from development contexts and relying heavily on anecdotal or one-time assessments (Mazurana and Marshak, 2019). Hunersen _et al._ (2020) suggest a potential uptick in early marriage rates among refugees after displacement but highlight that education acts as a protective factor.\\n\\n# **V.B Effects on native children**\\n\\nThe majority of research focuses on evaluating the impact of refugee children on host communities, with mixed outcomes. However, it is clear that negative consequences are more likely to occur in the absence of adequate measures to mitigate the challenges associated with increased demand for public services, especially in the education sector. This is primarily due to the “overcrowding effect” resulting from the influx of refugee children, which strains existing educational resources and infrastructure.\\n\\n_Studies documenting null effects._ Van der Werf (2021) reports that the presence of refugee students in educational settings does not prompt adverse effects on the educational outcomes of native children. The result is particularly noticeable when compensatory measures—such as increased funding, additional counselors, and the provision of night shifts—are implemented. This observation is supported by Figlio and Ozek<sup> ̈</sup> (2019) and Assaad _et al._ (2023), who note that these measures act as safeguards to mitigate potential negative consequences for native children.\\n\\n_Studies documenting positive effects._ Some studies find positive outcomes associated with the presence of refugee students. Morales (2022) emphasizes the beneficial impact on educational performance, largely due to the spillover effects of compensatory funding. This suggests that resources allocated to support students from forced migrant backgrounds can also positively influence native students. Furthermore, as mentioned earlier, Tumen (2021) suggests that the increased presence of Syrian refugee students can lead to higher motivation among native children, likely driven 24\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"by heightened competition in the academic context. Furthermore, studies by C ̧ akır _et al._ (2023) and Tumen (2021) highlight changes in parental decision-making about their children’s education. These changes include higher enrollment rates, influenced by shifts in labor market dynamics and the perceived value of education.\\n\\n_Studies documenting negative effects_ . The research also uncovers several negative aspects associated with forced migration, similar to those observed in voluntary migration scenarios. Gerdes (2010) and Mart ́ınez (2022), for example, document the phenomenon of “Native Flight,” which is also common in the context of forced migration. Furthermore, Green and Iversen (2020) point out negative spillover effects on educational performance, with a particular impact on low-achieving students. These findings underscore the necessity for well-thought-out policies and strategies that cater to the diverse needs of different student groups within the educational system, ensuring that all students are adequately supported and none are disproportionately disadvantaged.\\n\\n# **V.C Evidence on the effectiveness of interventions to support refugee children**\\n\\nThis section outlines the effects of interventions to support refugee children. Most of this research has examined the effects of programs on the refugee population in general without focusing specifically on children. In doing so, however, authors often also examine the effects of those programs on a number of children’s outcomes as secondary or exploratory analyses. Results from this work suggest positive effects of cash transfers, programs to facilitate integration (e.g., language courses), and programs that promote social cohesion and diversity awareness.\\n\\n_Evidence on the impacts of cash transfers._ De Hoop _et al._ (2019) examine the impacts of the “No Lost Generation” cash transfer program on the education outcomes of Syrian refugees in Lebanon. This program provided cash assistance to families to cover school commuting costs and compensate households for the potential income their children could have earned. The authors took advantage of the program’s random allocation to treated and control schools and found that the cash transfer resulted in a 20% increase in school attendance in treated schools compared to the control group. These findings are significant, particularly because the cash transfer was not contingent on children’s attendance and was therefore not monitored by the Lebanese government.\\n\\nMoreover, Ozler<sup> ̈</sup> _et al._ (2021) use a regression discontinuity design to examine the effects of the 25\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"“Emergency Social Safety Net” in T ̈urkiye, a cash transfer program for Syrian refugees. The authors show that the program caused changes in household composition, with a movement of primarily school-aged children from ineligible to eligible households. This was ultimately reflected in a reduction in poverty and inequality for the overall refugee population. Similar effects on housing recomposition are documented by Miguel _et al._ (2023), who examine the effects of the Norwegian Refugee Council’s Urban Shelter Program in Jordan through a randomized controlled trial. The program supports Syrian refugees by granting them housing subsidies. In this case, treated households experienced an increment in the number of boys ages 13 to 17 years in the short run. These effects dissipate in the medium to long run.\\n\\n_Programs facilitating integration_ . Evidence suggests that policies to facilitate migrant integration, even if focused on the adult member of the household, can have positive effects on children’s outcomes. Rude (2023), for example, finds that the introduction of birthright citizenship for Venezuelan children in Colombia has positively affected their educational prospects. The study reveals that enrollment rates in public institutions for Venezuelan children under six years improved by 50%. Similarly, Foged _et al._ (2022) and Foged _et al._ (2023) provide compelling evidence of positive intergenerational benefits resulting from language training programs offered by the Danish government to immigrant adults after they were granted refugee status in 1999. These programs led to higher completion rates among their children in lower secondary school.\\n\\n_Programs promoting social cohesion between students._ Alan _et al._ (2021) evaluate an educational program in T ̈urkiye that was designed to foster perspective-taking skills among students with the aim of enhancing social cohesion among peers. Perspective taking, in this context, refers to a cognitive skill that enables individuals to comprehend the goals and intentions of others. In practice, this skill empowers students to consider the perspectives and opinions of their peers, ultimately promoting greater tolerance and acceptance of differences, particularly with respect to refugee students. The findings revealed a dual impact: it reduced incidents of violence, peer victimization, and exclusion in schools with a higher proportion of Syrian refugee children while also producing significant improvements in prosocial behavior and altruism.\\n\\nMoreover, Tumen _et al._ (2023) evaluate a teacher training program implemented in Turkish schools. This program aimed to enhance teachers’ diversity awareness by improving their knowledge of 26\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [27]}}, {\"page\": 28, \"chunk\": 28, \"input_text\": \"the immediate needs of Syrian refugee students, enabling the teachers to provide mentoring and create a more conducive classroom learning environment. The authors employ a regression discontinuity design, as assignment into the treatment group depended on the proportion of refugee students in a school. They show that the program effectively halved absenteeism among refugee children during the year of implementation and continued to have a positive impact in the subsequent year. This work exemplifies the importance of social inclusion in diverse populations.\\n\\n_Programs improving mental health and early childhood development._ Moya _et al._ (2022) implemented a randomized trial intervention in Colombia that offers Child-Parent Psychotherapy (CPP) to children and caregivers affected by conflict and forced displacement. The goal of CPP is to restore the mental health of patients and strengthen the child-parent relationship to enhance care and child development. The authors observe positive effects of CPP on maternal mental health, as mothers appeared to be less anxious and depressed, among other conditions. More importantly, CPP was found to enhance child-mother interactions, resulting in a 0.17 standard deviation improvement in child mental health and a 0.21 standard deviation improvement on the early childhood development index.\\n\\nSimilarly, Annan _et al._ (2017) implemented a related intervention for the Burmese displaced population in Thailand. The intervention was designed after the Strengthening Families Program (SFP) to train Burmese parents in caregiving skills while teaching refugee children social skills. The primary objective was to enhance children’s psychological outcomes by improving the quality of parental care and communication. The authors found the program cut the likelihood of externalizing problems (e.g., aggressive behavior) and child attention problems. Compared to the control group, the reductions were 0.22 standard deviation and 0.11 standard deviation, respectively.\\n\\n# **VI DISCUSSION**\\n\\nThis paper examines the impact of migration on children, offering a comprehensive analysis of its effects throughout the process. This includes children left behind, migrant children (including those forcibly displaced), and native children. The findings reveal that migration’s impact on children’s outcomes is complex and context dependent, influenced by household demographics and particularly by policies to support migrant populations and their host communities. Policies that facilitate service adjustments in response to migrant arrivals and that promote integration 27\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [28]}}, {\"page\": 29, \"chunk\": 29, \"input_text\": \"are highly beneficial for all children involved. Conversely, enforcement-oriented policies tend to hinder children’s development, particularly among immigrant children, by reducing service uptake and damaging the mental health of children and their parents.\\n\\nThe review also draws attention to the vulnerability of left-behind children, who despite observing positive income shocks might still be negatively affected by parental absence. Notably, the evidence highlighted the large negative effects of parental absence on cognitive outcomes, mental health, and socioemotional behaviors. Future research and programming should cover effective ways to support these children. Specifically, the Chinese case raises a crucial question about the trade-off between relocation policies to spur growth and the human capital costs of such policies.\\n\\nForcibly displaced children are one of the most vulnerable groups since they not only experience migration but do so unexpectedly, often after traumatic events. Yet, without representative and longitudinal data to comprehensively measure children’s outcomes, it is difficult to diagnose the precise challenges they face. This includes an effective evaluation of their cognitive and noncognitive development relative to native children and its impact on future outcomes, such as adult income. These issues are particularly important in developing countries and present a significant opportunity to expand our knowledge in the future.\\n\\nFinally, more evaluation of programs that effectively target migrant and forced migrant children’s outcomes will be extremely valuable in guiding policymakers about the most effective programs to support these populations and their hosts. This is particularly crucial for long-term outcomes because the majority of existing work concerns the immediate impact of such programs due to various constraints. Even though the forcibly displaced population requires urgent attention, there is wide scope for action in understanding the long-term impacts of these experiences during their teenage years and into adulthood.\\n\\n28\", \"datasets\": [{\"mention_name\": {\"text\": \"representative and longitudinal data\", \"confidence\": 0.5969276428222656, \"start\": 1015, \"end\": 1051, \"clean_text\": \"representative and longitudinal data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1015, \"end\": 1051, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1015, \"end\": 1051, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1015, \"end\": 1051}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1015, \"end\": 1051, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.22171489894390106, \"start\": 1015, \"end\": 1051}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5969276428222656, \"start\": 1015, \"end\": 1051}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.858281135559082, \"start\": 1015, \"end\": 1051}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.858281135559082, \"start\": 1015, \"end\": 1051}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099418301302435890/pdf/IDU113cacc80159c21491318dca10779c4990d0b.pdf\", \"pages\": [29]}}]}"}}},{"rowIdx":91,"cells":{"doc_id":{"kind":"string","value":"34248497"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34248497\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"admreg\": \"Middle East and North Africa,Middle East and North Africa\", \"admreg_key\": \"119229,119229\", \"authors\": {\"0\": {\"author\": \"Amjad,Beenish\"}, \"1\": {\"author\": \"Carrasco Nunez,Haydeeliz\"}, \"2\": {\"author\": \"Finn,Arden Jeremy\"}, \"3\": {\"author\": \"Goldman,Maya Scott\"}}, \"count\": \"West Bank and Gaza\", \"count_key\": \"82715\", \"docna\": {\"0\": {\"docna\": \"Fiscal Policy, Poverty, and Inequality in a\\n Constrained Environment : The Case of the West Bank and Gaza\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"EFI-MNA-POV-Poverty and Equity (EMNPV)\", \"projn\": \"GZ-Poverty, Equity And Welfare Monitoring In\\n The West Bank And Gaz -- P179253\", \"subsc\": \"Central Government (Central Agencies)\", \"trustfund\": \"TF0B9468-Bringing a gender lens to the\\n impact of fiscal policy and social progra\", \"trustfund_key\": \"2211562\", \"theme\": \"Data production, accessibility and use,Gender,Human Development and Gender,Data Development and Capacity Building,Economic Policy,Rural Development,Fiscal Policy,Institutional strengthening and capacity building,Public Sector Management,Fiscal sustainability,Urban and Rural Development,Geospatial Services\", \"prdln\": \"Advisory Services & Analytics\", \"prdln_key\": \"1416466\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34248497\"}, \"repnb\": \"WPS10687\", \"docdt\": \"2024-01-30T00:00:00Z\", \"datestored\": \"2024-01-30T04:46:40Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"This report analyzes the distributional\\n impacts of the main taxes and transfers on\\n households’ welfare in the West Bank and Gaza.\\n The analysis uses the Commitment to Equity methodology,\\n enabling comparison of the results to other countries where\\n this framework has been applied. The report assesses the\\n effects of government taxation, social expenditure, and\\n indirect subsidies on poverty and inequality in the West\\n Bank and Gaza. The results indicate that the combination of\\n taxes and transfers modelled in the West Bank and Gaza\\n reduces inequality by 6.5 Gini points but increases the\\n national poverty headcount by 8.4 percentage points. These\\n fiscal policy outcomes on poverty and inequality reduction\\n are below average in terms of desirability compared to other\\n lower-middle-income countries. The taxes and transfers\\n modelled in the West Bank and Gaza achieve most inequality\\n reduction through in-kind benefits from public basic\\n education and public hospitals, followed by the Cash\\n Transfer Program and the value-added tax (VAT). Their large\\n impact on inequality reduction is explained by a combination\\n of their progressivity and their size relative to household\\n income. The redistributive effect of direct taxes, customs\\n duties, and indirect subsidies is zero or close to zero.\\n Indirect taxes represent the fiscal interventions\\n contributing most to the increase in national poverty;\\n customs duties followed by VAT represent the largest burden\\n on households’ incomes. Direct transfers from\\n social protection cannot offset the impoverishment effect\\n from indirect taxes because they have very limited coverage.\\n Only the poorest decile is a net cash beneficiary after\\n paying taxes and receiving cashable transfers. The rest of\\n the deciles are net payers to the fiscal system. To decrease\\n poverty and inequality in the West Bank and Gaza, the most\\n significant policy recommendation to emerge from the\\n analysis is to expand direct transfers to the second and\\n third deciles to compensate for indirect tax burdens.\\n Financing this reform is feasible through domestic tax\\n mobilization or through rationalization of inefficient fuel\\n and electricity subsidies that benefit the top income\\n deciles most.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10687; RRR; PROSPERITY\", \"display_title\": \"Fiscal Policy, Poverty, and\\n Inequality in a Constrained Environment : The Case of the\\n West Bank and Gaza\", \"disclosure_date\": \"2024-01-30T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099646401302432923/text/IDU11fba1aab1fb111482e1a1441122a743dfde3.txt\", \"listing_relative_url\": \"/research/2024/01/34248497/fiscal-policy-poverty-inequality-constrained-environment-case-west-bank-gaza\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099646401302432923/IDU11fba1aab1fb111482e1a1441122a743dfde3\", \"new_url\": \"2024/01/34248497/IDU11fba1aab1fb111482e1a1441122a743dfde3\", \"ext_pub_date\": \"2024-01-30T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Poverty m1326687 695,Fiscal Incidence Analysis m1326940 711,Taxation m1327615 1121,Inequality and Shared Prosperity m1327655 700\", \"docm_id\": \"IDU-1fba1aab-fb11-482e-a144-122a743dfde3\", \"chronical_docm_id\": \"IDU-1fba1aab-fb11-482e-a144-122a743dfde3\", \"publishtoextweb_dt\": \"2024-01-30T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10687\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"projectid\": \"P179253\", \"guid\": \"099646401302432923\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"79\", \"author_keys\": {\"0\": {\"author_key\": \"1959425\"}, \"1\": {\"author_key\": \"1599475\"}, \"2\": {\"author_key\": \"1594966\"}, \"3\": {\"author_key\": \"1712768\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099646401302432923/IDU11fba1aab1fb111482e1a1441122a743dfde3\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10687 Fiscal Policy, Poverty, and Inequality in a Constrained Environment The Case of the West Bank and Gaza _Beenish Amjad Haydeeliz Carrasco Arden Finn Maya Goldman_ Poverty and Equity Global Practice January 2024 A verified reproducibility package for this paper is available at http://reproducibility.worldbank.org, click **here** for direct access.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10687\\n\\n# **Abstract**\\n\\nThis report analyzes the distributional impacts of the main taxes and transfers on households’ welfare in the West Bank and Gaza. The analysis uses the Commitment to Equity methodology, enabling comparison of the results to other countries where this framework has been applied. The report assesses the effects of government taxation, social expenditure, and indirect subsidies on poverty and inequality in the West Bank and Gaza. The results indicate that the combination of taxes and transfers modelled in the West Bank and Gaza reduces inequality by 6.5 Gini points but increases the national poverty headcount by 8.4 percentage points. These fiscal policy outcomes on poverty and inequality reduction are below average in terms of desirability compared to other lower-middle-income countries. The taxes and transfers modelled in the West Bank and Gaza achieve most inequality reduction through in-kind benefits from public basic education and public hospitals, followed by the Cash Transfer Program and the value-added tax (VAT). Their large impact on inequality reduction is explained by a combination of their progressivity and their size relative to household income. The redistributive effect of direct taxes, customs duties, and indirect subsidies is zero or close to zero. Indirect taxes represent the fiscal interventions contributing most to the increase in national poverty; customs duties followed by VAT represent the largest burden on households’ incomes. Direct transfers from social protection cannot offset the impoverishment effect from indirect taxes because they have very limited coverage. Only the poorest decile is a net cash beneficiary after paying taxes and receiving cashable transfers. The rest of the deciles are net payers to the fiscal system. To decrease poverty and inequality in the West Bank and Gaza, the most significant policy recommendation to emerge from the analysis is to expand direct transfers to the second and third deciles to compensate for indirect tax burdens. Financing this reform is feasible through domestic tax mobilization or through rationalization of inefficient fuel and electricity subsidies that benefit the top income deciles most.\\n\\nThis paper is a product of the Poverty and Equity Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at afinn1@worldbank.org. A verified reproducibility package for this paper is available at http://reproducibility.worldbank. org, click **here** for direct access.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# Fiscal Policy, Poverty, and Inequality in a Constrained Environment: The Case of the West Bank and Gaza<sup>1</sup>\\n\\nBeenish Amjad Haydeeliz Carrasco Arden Finn<sup>2</sup> Maya Goldman This paper was finalized prior to the conflict in the Middle East that started in October 2023. Given the rapid pace of events and the possible changes to the enabling environment, some of the article’s analytics, assessments, and policy recommendations may not fully reflect the latest developments, and in a number of cases, the challenges raised in the paper may have been exacerbated by the hostilities. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily reflect the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.\\n\\n> 1 **Acknowledgements** : This report was written under the overall guidance of Johannes Hoogeveen and Alan Fuchs. We are grateful to Alia Jane Aghajanian, Eduardo Malasquez, Gianluca Mele and Nur Nasser Eddin for their contributions, and to Jessica Marie Anderson, Hamza Mighri, Sami Miaari, Haytham Abushaham and Ntuthuko Hlela for their support with data collection and research assistance. We are also thankful to the many World Bank colleagues and staff at the Palestinian Ministry of Finance and the Palestinian Central Bureau of Statistics who generously provided the information and data needed to finalize this work. Jon Jellema, Gabriela Inchauste and Matthew Wai-Poi provided insightful peer review comments that have greatly improved this report.\\n\\n> 2 Corresponding author: afinn1@worldbank.org\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **Table of Contents**\\n\\n|1<br>Introduction ...................................................................................................................................... 5|\\n|---|\\n|1.1<br>Six Key Findings ............................................................................................................................. 7|\\n|2<br>Tax revenue and public expenditure .............................................................................................. 10|\\n|2.1<br>Overview of the fiscal system ..................................................................................................... 10|\\n|2.2<br>Pensions ...................................................................................................................................... 12|\\n|2.3<br>Tax revenues ............................................................................................................................... 13|\\n|2.3.1<br>Direct taxes ......................................................................................................................... 13|\\n|2.3.2<br>Indirect taxes ....................................................................................................................... 14|\\n|2.4<br>Transfers ..................................................................................................................................... 15|\\n|2.4.1<br>Direct transfers ................................................................................................................... 15|\\n|2.4.2. Indirect subsidies .......................................................................................................................... 16|\\n|2.5<br>In-kind benefits ........................................................................................................................... 18|\\n|2.6<br>International comparison ........................................................................................................... 18|\\n|3<br>Data and methodology ................................................................................................................... 21|\\n|3.1<br>Data sources ................................................................................................................................ 21|\\n|3.2<br>Commitment to Equity (CEQ) framework ................................................................................... 21|\\n|4<br>Methodologies for allocating taxes and transfers .......................................................................... 26|\\n|4.1<br>Pensions ...................................................................................................................................... 26|\\n|4.2<br>Direct transfers ........................................................................................................................... 27|\\n|4.3<br>Direct taxes ................................................................................................................................. 27|\\n|_4.3.1_<br>_Personal income tax (PIT)_.................................................................................................... 27|\\n|_4.3.2. Property tax_............................................................................................................................... 29|\\n|4.4. Indirect taxes .................................................................................................................................... 29|\\n|4.4.1. Customs..................................................................................................................................... 29|\\n|4.4.2. VAT ............................................................................................................................................ 29|\\n|4.4.3. Excises ....................................................................................................................................... 31|\\n|4.5. Indirect subsidies ............................................................................................................................. 31|\\n|4.5.1. Domestic electricity .................................................................................................................. 31|\\n|4.5.2. Fuel subsidies ............................................................................................................................ 32|\\n|4.6<br>In-kind benefits ........................................................................................................................... 33|\\n|5<br>Headline results .............................................................................................................................. 35|\\n\\n1\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"|5.1<br>Poverty and inequality ................................................................................................................ 35|\\n|---|\\n|5.2<br>Fiscal impoverishment ................................................................................................................ 37|\\n|5.3<br>Net cash beneficiaries ................................................................................................................. 39|\\n|6<br>Marginal contributions and progressivity ....................................................................................... 41|\\n|6.1<br>Marginal contributions ............................................................................................................... 41|\\n|6.1.1<br>Inequality ............................................................................................................................ 41|\\n|6.1.2<br>Poverty ................................................................................................................................ 43|\\n|6.2<br>Progressivity ................................................................................................................................ 45|\\n|7<br>Incidence and concentration shares by decile ................................................................................ 48|\\n|7.1.<br>Taxes ........................................................................................................................................... 48|\\n|7.2.<br>Transfers ..................................................................................................................................... 50|\\n|8<br>International comparison ............................................................................................................... 53|\\n|8.1<br>Inequality .................................................................................................................................... 53|\\n|8.2<br>Poverty ........................................................................................................................................ 53|\\n|9<br>Macro-validation and limitations of the analysis ............................................................................ 55|\\n|9.1<br>Macro-validation ......................................................................................................................... 55|\\n|9.2<br>Limitations................................................................................................................................... 57|\\n|9.2.1<br>Direct taxes and pension contributions .............................................................................. 57|\\n|9.2.2<br>Indirect taxes ....................................................................................................................... 58|\\n|9.2.3<br>Indirect subsidies ................................................................................................................ 58|\\n|9.3<br>Direct transfers ........................................................................................................................... 59<br> <br>|\\n|9.3.1<br>Education ............................................................................................................................ 59|\\n|9.3.2<br>Health .................................................................................................................................. 59|\\n|10<br>Conclusion and next steps .............................................................................................................. 60|\\n|10.1<br>Conclusion ................................................................................................................................... 60|\\n|10.2<br>Next steps ................................................................................................................................... 61|\\n|References .................................................................................................................................................. 62|\\n|Annex A: Pensions as deferred income versus pensions as a government transfer .................................. 66|\\n|Annex B: Methodology for allocating PIT deductions ................................................................................ 69|\\n|Annex C: Marginal Contributions at Market, Disposable and Consumable income ................................... 75|\\n\\n2\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"|Figure 1-1 Poverty and inequality are higher in Gaza than in the West Bank .............................................. 6|\\n|---|\\n|Figure 2-1 Tax revenues as a share of GDP are higher than the sample average at 20.7 percent ............. 19|\\n|Figure 2-2 Social expenditures as a share of GDP are higher than the sample average at 14.1 percent ... 20|\\n|Figure 3-1 Income concepts in the CEQ framework ................................................................................... 24|\\n|Figure 5-1 The national poverty headcount increases by 8.4 percentage points from Market to<br>Consumable Income ................................................................................................................................... 36|\\n|Figure 5-2 The Gini coefficient decreases by 6.5 Gini points from Market to Final Income ...................... 37|\\n|Figure 5-3 Over two thirds of the market poor population experiences net fiscal gains due to fiscal policy<br>.................................................................................................................................................................... 39|\\n|Figure 5-4 Households in decile 1 are net cash beneficiaries of fiscal policy in the West Bank and Gaza . 40|\\n|Figure 6-1 Almost all fiscal instruments in the West Bank and Gaza are either inequality-reducing or have|\\n|no effect ...................................................................................................................................................... 43|\\n|Figure 6-2 In the West Bank and Gaza, direct transfers and the direct effects of electricity subsidies are<br>the only poverty-reducing instruments ...................................................................................................... 44|\\n|Figure 6-3 All transfers are progressive, except for fuel subsidies and pension income ........................... 46|\\n|Figure 6-4 All taxes are progressive. The PIT is the most progressive tax, however the direct effects of<br>fuel excises are the most progressive component of the fiscal system. .................................................... 47|\\n|Figure 7-1 Of all taxes, the Customs Tax and VAT have the largest incidence across the distribution ...... 49|\\n|Figure 7-2 Cash transfers provide the largest benefits to the poorest decile ............................................ 51|\\n|Figure 8-1 Total inequality reduction in the West Bank and Gaza is moderate relative to countries in<br>other CEQ assessments ............................................................................................................................... 53|\\n|Figure 8-2 Total poverty impact at the $5.50 line is severe relative to countries in other CEQ assessments<br>.................................................................................................................................................................... 54|\\n|Table 2-1 Total revenue and government expenditure in the West Bank and Gaza in 2017 is 27.0 and<br>27.4 percent, respectively, of GDP ............................................................................................................. 10|\\n|Table 2-2 There are three pension schemes applicable in the West Bank and Gaza, and one pension<br>scheme for Palestinian employees working in Israel and the settlements ................................................ 12|\\n|Table 2-3 The income tax framework in the West Bank and Gaza has only three brackets, and statutory<br>rates range between 5 and 15 percent....................................................................................................... 14|\\n|Table 2-4 The price of electricity per kWh for domestic consumers ranges between 0.458 and 0.6731 in<br>the West Bank and Gaza ............................................................................................................................. 16|\\n|Table 4-1: Scaling is applied to deductions to prevent an under-allocation of PIT in the survey .............. 28|\\n|Table 4-2 The estimated electricity subsidy per kWh ranges from 0.178 to -0.0371 in the West Bank and<br>Gaza............................................................................................................................................................. 32|\\n\\n3\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"Table 9-1 The ratio of the amount allocated in the survey to the amount recorded in the budget ranges from 43% (for the electricity subsidy) to 124% (for pension income) ........................................................ 56 4\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"# **1 Introduction**\\n\\n## **Fiscal space for the Palestinian Authority (PA) has been shrinking for much of the past two decades and**\\n\\n**is reaching a critically low level.** Economic shocks from COVID-19 and the Russian invasion of Ukraine have compounded negative effects from systemic constraints, recurring armed conflict, and declining external assistance. Despite relatively stronger public revenue mobilization in 2022, the PA faces persistent financing gaps. It has so far filled the gaps by accruing arrears to the private sector and public pension funds and by paying partial salaries to public employees, but these coping strategies are reaching their limits.<sup>3</sup> Recent decisions by the new Israeli government to increase deductions from clearance revenues<sup>4</sup> and withhold funds from previously collected revenues will likely increase the strain on PA budgets.<sup>5</sup>\\n\\n**Fiscal policy choices in a constricted resource context affect the day-to-day lives and wellbeing of Palestinians.** In one current example, the PA halted payments to beneficiaries of the main social assistance program (the National Cash Transfer Program) for much of 2022—still delayed at the time of writing this report—to divert resources to other parts of the national budget. Meanwhile, the value of PA fuel subsidies jumped 191 percent in 2022 following the rise in global oil prices.<sup>6</sup> While this cost has since come down and remains modest compared to other countries in the region, this remains a significant fiscal risk. On taxation, efforts to expand fiscal space by increasing public revenue mobilization could risk increasing poverty and inequality if they impose too heavy a burden on lower-income taxpayers.\\n\\n**Distributional analysis of fiscal choices offers valuable insights on how policies affect poverty and equity.** This is in line with calls to improve the fairness, effectiveness, and efficiency of public expenditure while prioritizing support to the most vulnerable.<sup>7</sup> The analysis complements ongoing reform initiatives under the World Bank’s current Development Policy Grant. Moreover, analyzing the distributional impact of fiscal policies on households’ welfare is an important part of monitoring the Sustainable Development Goal 10.4.2, which measures how countries’ fiscal systems contribute to inequality reduction.\\n\\n**This report represents the first comprehensive fiscal incidence analysis in the West Bank and Gaza.** It uses the Commitment to Equity (CEQ) methodology to assess distributional repercussions of the main tax interventions and public social expenditures, including direct and indirect taxes, direct transfers and indirect subsidies, and in-kind benefits from the provision of public health and education services. The analysis is anchored in data from the most recent nationally representative Palestinian Expenditure and Consumption Survey (PECS) (2016/2017), as well as administrative data for the main taxes and social expenditure items in the West Bank and Gaza.\\n\\n## **The analysis presented in this report seeks to answer four key questions:**\\n\\n- How does the combination of taxes and transfers modeled affect poverty and inequality in the West Bank and Gaza?\\n\\n- Which fiscal interventions are the most important in driving these effects?\\n\\n- How progressive or regressive is each component of the fiscal system?\\n\\n> 3 For further detail on the fiscal situation, see: World Bank (2022a), World Bank (September 2022).\\n\\n> 4 Clearance revenues refer to revenue collected by Israel on behalf of and transferred to the Palestinian Authority.\\n\\n> 5 World Bank (March 2023).\\n\\n> 6 World Bank (March 2023).\\n\\n> 7 World Bank (September 2022).\\n\\n5\", \"datasets\": [{\"mention_name\": {\"text\": \"Palestinian Expenditure and Consumption Survey\", \"confidence\": 0.8320466876029968, \"start\": 2869, \"end\": 2915, \"clean_text\": \"Palestinian Expenditure and Consumption Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2869, \"end\": 2915, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2869, \"end\": 2915, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2869, \"end\": 2915}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2869, \"end\": 2915, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.733452320098877, \"start\": 460, \"end\": 506}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8320466876029968, \"start\": 460, \"end\": 506}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5136283040046692, \"start\": 460, \"end\": 506}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5136283040046692, \"start\": 460, \"end\": 506}}, {\"mention_name\": {\"text\": \"administrative data\", \"confidence\": 0.5380475521087646, \"start\": 2947, \"end\": 2966, \"clean_text\": \"administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2947, \"end\": 2966, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2947, \"end\": 2966, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2947, \"end\": 2966}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2947, \"end\": 2966, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.733452320098877, \"start\": 538, \"end\": 557}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5380475521087646, \"start\": 538, \"end\": 557}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5136283040046692, \"start\": 538, \"end\": 557}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5136283040046692, \"start\": 538, \"end\": 557}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"- How can we predict and model potential policy reforms using this study as a base?\\n\\n**The analysis comes at a time when poverty in the Palestinian territories remains elevated after the COVID-19 shock and has become increasingly concentrated in Gaza.** Based on the PECS 2016/2017, about 29.2 percent of the Palestinian population lived on household consumption levels below the national poverty line of NIS 2,470 per month for a reference household of two adults and three children (equivalent to about US$7.92 in 2017 PPP per day per capita). About 20.5 percent lived on household consumption levels below the Upper Middle Income Class Poverty Line of US$6.85 (2017 PPP) per day per capita. However, poverty rates differed substantially by region: 13.9 percent of Palestinians in the West Bank (including East Jerusalem) and 53.0 percent of Palestinians in Gaza lived below the national poverty line (Figure 1-1).<sup>8</sup> Consequently, the Palestinian poor have increasingly concentrated in Gaza. GDP per capita measured in 2017 PPP dollars had been increasing steadily since 2004, but poverty-reduction has been achieved primarily in the West Bank. Poverty rates in Gaza remain persistently higher and increased markedly from 2011 to 2016-17. Inequality as measured by the Gini coefficient was relatively low at 0.30 nationally in 2016-17, and slightly higher in Gaza with respect to the West Bank. Since 2016-17, poverty rate estimates spiked during the peak of the COVID-19 pandemic, and while they have gradually declined, they have not yet recovered to pre-pandemic levels.\\n\\n# **Figure 1-1 Poverty and inequality are higher in Gaza than in the West Bank**\\n\\nPoverty and Inequality in the West Bank and Gaza, 2017 Panel a: Poverty by location Panel b: Inequality by location <!-- Start of picture text --> 53.0<br>0.3<br>0.27 0.28<br>29.2<br>13.9<br>West Bank Gaza National West Bank Gaza National<br>Gini coefficient (0-1)<br>Below national poverty line (%)<br><!-- End of picture text --> Source: World Bank calculations based on PECS 2016/17.\\n\\nNote: Poverty calculated based on real consumption per capita. National poverty line establishes different thresholds based on different combinations of numbers of adults and children in the data. Gini calculated based on real consumption per adult equivalent.\\n\\n**The growing regional poverty disparity reflects geographic, administrative, and economic fragmentation.** East Jerusalem has been under Israeli administration and domestic law since 1967, and it\\n\\n> 8 Figures and trends in this paragraph drawn from: World Bank (October 2022), World Bank (2022a), Palestinian Central Bureau of Statistics (PCBS) (2018), PCBS (2011), PCBS (2007), PCBS (2006), PCBS (2004).\\n\\n6\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS 2016/2017\", \"confidence\": 0.6915069818496704, \"start\": 267, \"end\": 281, \"clean_text\": \"PECS 2016/2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 267, \"end\": 281, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 267, \"end\": 281, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 267, \"end\": 281}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 267, \"end\": 281, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9013588428497314, \"start\": 267, \"end\": 281}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6915069818496704, \"start\": 267, \"end\": 281}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7425004243850708, \"start\": 267, \"end\": 281}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7425004243850708, \"start\": 267, \"end\": 281}}, {\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.7616845369338989, \"start\": 2042, \"end\": 2046, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.42249467968940735, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2042, \"end\": 2046, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2042, \"end\": 2046}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2042, \"end\": 2046, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.985640287399292, \"start\": 498, \"end\": 502}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7616845369338989, \"start\": 498, \"end\": 502}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.624189555644989, \"start\": 498, \"end\": 502}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.624189555644989, \"start\": 498, \"end\": 502}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"is isolated from the remainder of the West Bank by the separation barrier and expanding settlements.<sup>9</sup> The West Bank does not function as a single contiguous administrative system. The Oslo Accords designated discontinuous areas to be assigned to the civil and security (or joint security) jurisdiction of the PA—Areas A and B—and left the more than 60 percent of the territory remaining primarily under Israeli control as Area C.<sup>10</sup> The PA is responsible for service provision in Area C, but infrastructure development and taxation fall within Israeli authority.<sup>11</sup> Gaza meanwhile is under a sea and air blockade, and has been under the _de facto_ authority of Hamas since 2007. The PA and the _de facto_ authorities in Gaza do not coordinate administrative and fiscal systems.<sup>12</sup> Palestinians living in these distinct contexts experience very different fiscal and socioeconomic realities, and it is difficult to move between the physically separated territories.<sup>13</sup>\\n\\n## **The PA operates within the policy constraints defined by this fragmentation and by the Paris Protocol.**\\n\\nIntended as a temporary arrangement, over 25 years after its signing the Protocol remains the general framework governing, shaping, and constraining Palestinian trade relations, and trade, macroeconomic, and fiscal policies, including economic and fiscal relations between the PA and the Government of Israel (GOI).<sup>14</sup> Among other provisions, it ties Palestinian regulatory frameworks, indirect tax rates, and trade policy to those of Israel.<sup>15</sup> It also established the revenue clearance system by which the GoI is supposed to collect and transfer certain public revenues to the PA, including import taxes, taxes on Palestinian workers in Israeli-administered territories, and taxes on economic activity in Area C.<sup>16</sup> Some provisions of the Paris Protocol are outdated, and sometimes inconsistently implemented. In addition, under the current _de facto_ authority Gaza is not integrated into the Protocol. The fiscal difficulties this situation creates for the PA are substantial: the administration spent about one-third of its 2021 budget in East Jerusalem and Gaza but collects almost no revenue from these areas or from Area C.<sup>17</sup>\\n\\n## **Acknowledging these constraints, this analysis aims to understand the distributional impacts of PA fiscal**\\n\\n**policies.** We aim to analyze the overall household welfare effects of a complex fiscal environment and identify areas where policy change could reduce poverty and advance equity.\\n\\n# **1.1 Six Key Findings**\\n\\n**1. Fiscal policies reduce inequality but increase poverty.** The final model in the analysis finds that the combination of taxes and public transfers in the 2017 fiscal year reduces inequality by 6.5 Gini points and increases the national poverty headcount by 8.4 percentage points. Much of the reduction in inequality is driven by in-kind benefits (basic public education and public health), while the poverty increase stems from the fact that the social protection system in the West Bank and Gaza is too small to compensate for the impoverishing effects of indirect taxes such as customs duties and the Value-Added Tax (VAT). The\\n\\n> 9 United Nations General Assembly (September 14, 2022).\\n\\n> 10 United Nations General Assembly (September 14, 2022).\\n\\n> 11 United Nations General Assembly (September 14, 2022); United Nations Conference on Trade and Development (2019).\\n\\n> 12 Office of the United Nations Special Coordinator for the Middle East peace process (September 22, 2022).\\n\\n> 13 See for example World Bank (September 2022); United Nations Office for the Coordination of Humanitarian Affairs (OCHA) (January 25, 2023), OCHA (January 23, 2023), OCHA: occupied Palestinian territory (OCHA oPt) (June 2020).\\n\\n> 14 Office of the United Nations Special Coordinator for the Middle East peace process (September 22, 2022).\\n\\n> 15 World Bank (September 2022), Office of the United Nations Special Coordinator for the Middle East peace process (September 22, 2022).\\n\\n> 16 World Bank (September 2022), United Nations Conference on Trade and Development (2019).\\n\\n> 17 International Monetary Fund (2022b).\\n\\n7\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"reduction in inequality also reflects substantial redistribution from the West Bank to Gaza, as the model assumes that residents of Gaza pay few taxes but benefit more from public transfers.\\n\\n**2. Nationally, most fiscal interventions are progressive relative to Market Income.** This means that the rich pay a larger share of total taxes than their share of market income, and the poor receive a larger share of total transfers than their market income share. The exceptions are fuel subsidies, pension income, and property tax on land which are all regressive.\\n\\n**3. While indirect taxes reduce inequality in the West Bank and Gaza, they nevertheless constitute a burden for the poor and so require compensation.** Given that the poor tend to consume more from informal retail outlets, indirect taxes like the VAT become _de facto_ progressive (Bachas et al. 2020). The large size of VAT and its progressivity reduces inequality through greater taxation of the rich as a share of income. Nonetheless, indirect taxes like the VAT still constitute a burden for the poor that increases poverty, which should be compensated for in the form of greater social protection.\\n\\n**4. To reduce poverty, more redistribution to deciles 2 and 3 needs to occur.** At the national level, only the poorest decile is a net cash beneficiary of the fiscal system; that is, they receive more from the system in the form of direct transfers and indirect subsidies than they pay in the form of direct taxes, pension contributions, and indirect taxes. All the remaining deciles are net payers to the fiscal system. The fact that deciles 2 and 3 are net payers is consistent with the increase in the national poverty headcount due to fiscal policy, and that the richest decile has a larger net cash loss is consistent with inequality reduction.\\n\\n**Expanding targeted social protection cash transfers to offset the tax burden would reduce the povertyincreasing effects of the fiscal system.** Including the Cash Transfer Program (CTP)<sup>18</sup> reduces poverty by 2.2 percentage points, more than any other instrument, and only pension income benefits come close to this at 2.0 percentage points. While the first three deciles live below the poverty line, only 26 percent of total transfers (pensions, direct transfers and indirect subsidies) currently go to the first decile, with decile 2 receiving 8 percent and decile 3 receiving 5 percent.<sup>19</sup>\\n\\n# **5. Indirect electricity and fuel subsidies are an inefficient way of compensating the poor, hence they**\\n\\n**should be rationalized to generate fiscal space and expand more redistributive social programs.** In absolute terms, the rich benefit more from electricity subsidies than the poor, although they are progressive. Changing the consumer tariff structure from incremental block tariff to volumetric could increase the progressivity of the electricity subsidies and enhance their redistributive effect, although an untargeted subsidy will never be as effective as a targeted cash transfer. Fuel subsidies are regressive, benefitting the rich more than the poor both in absolute and relative terms. Reducing the size of fuel subsidies in favor of a targeted cash transfer would represent a greater benefit to the poor.\\n\\n# **6. Given the relatively flat structure of the Personal Income Tax, it is only slightly more progressive than**\\n\\n**the VAT (measured by a Kakwani Index of 14.1 and 12.6 respectively).** Revising the structure of the PIT to increase the number of brackets, and implementing a more progressive set of rates would increase the progressivity of the tax, and the overall impact on inequality-reduction.\\n\\n> 18 Fiscal instrument refers to the ensemble of taxes and transfers which a government has at its disposal for redistribution and other objectives.\\n\\n> 19 If we exclude pension income, then 59 percent of direct transfers go to the first decile, 10 percent go to the second decile, and 7 percent to the decile 3.\\n\\n8\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"The rest of this report is organized as follows: Section 2 describes the fiscal system in the West Bank and Gaza, providing context on each of the fiscal instruments modeled in this assessment and an overview of the share of taxes and revenues modeled. Section 3 describes the fiscal administrative and microdata sources used for our model and the Commitment to Equity (CEQ) framework used in the analysis. Section 4 describes methodologies used to allocate each of the fiscal instruments, while Section 5 analyzes headline results. Section 6 presents the marginal contribution from each fiscal intervention to poverty and equity, while Section 7 presents distributional analysis by decile. Section 8 analyzes how fiscal interventions in the West Bank and Gaza compare to other countries in terms of distributional effects. Section 9 provides macro-validation, totals allocated in the survey compared with the official fiscal administrative data and reflects on the main limitations of the analysis. Finally, Section 10 provides concluding remarks.\\n\\n9\", \"datasets\": [{\"mention_name\": {\"text\": \"fiscal administrative data\", \"confidence\": 0.75227290391922, \"start\": 919, \"end\": 945, \"clean_text\": \"fiscal administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 919, \"end\": 945, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 919, \"end\": 945, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 919, \"end\": 945}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 919, \"end\": 945, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8740114569664001, \"start\": 919, \"end\": 945}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.75227290391922, \"start\": 919, \"end\": 945}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4315759539604187, \"start\": 919, \"end\": 945}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4315759539604187, \"start\": 919, \"end\": 945}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"# **2 Tax revenue and public expenditure**\\n\\n## **This section provides an overview of tax and public expenditure instruments in the West Bank and Gaza.**\\n\\nThis report assesses 89 percent of all tax revenue the PA collects and 42 percent of primary government expenditures.\\n\\n- **2.1 Overview of the fiscal system**\\n\\n## **Table 2-1 shows the disaggregation of tax revenue and government expenditure for the West Bank and Gaza in 2017, based on official government sources.**\\n\\n## **Tax revenue represents 20.7 percent of GDP in 2017, and this report models around 89 percent of this**\\n\\n**(equivalent to 18.5 percent of GDP).** About 19 percent of GDP stems from indirect taxes levied on consumption, production, and trade activities, while direct taxes represent 1.7 percent of GDP. Indirect tax collections stem mostly from taxes on international trade (7.1 percent of GDP), followed by VAT (5.9 percent of GDP), and excises (5.2 percent of GDP). Within direct taxes, Corporate Income Tax (CIT) represents the largest source of revenues (0.9 percent of GDP), followed by payroll taxes (0.4 percent of GDP), and the PIT (0.3 percent of GDP). We model 95.8 percent of indirect taxes (VAT, customs duties, and excises)—omitting only taxes on use of goods, and permissions—but only 18.3 percent of direct taxes (PIT and property tax). The small share of direct taxes modeled is mainly due to the inability of a CEQ assessment to measure CIT.<sup>20</sup>\\n\\n## **Total primary government expenditure is 27.4 percent of GDP in 2017, and we model 42 percent of that**\\n\\n**(to the value of 11.6 percent of GDP).** Social expenditure is 14.1 percent of GDP, while non-social government expenditure represents 13.3 percent of GDP. We model 81 percent of social expenditures and about 1.5 percent of non-social government expenditure by including the indirect subsidies on electricity and fuel. The main government social expenditure categories included in this study are education (5.0 percent of GDP), social protection (4.5 percent of GDP) and health (3.9 percent of GDP). We model 79 percent of education expenditures, 92 percent of social protection expenditures, and 84 percent of health expenditures. According to available information, indirect subsidies reached only 0.2 percent of GDP (Table 2-1). Most of the non-social expenditure stems from public order and safety (7.3 percent of GDP) and general public services (4.4 percent of GDP excluding debt transactions), which are not included in the current fiscal incidence model. The latter categories are all public goods for which the CEQ framework has no standard methodology for allocation. (See Box 3.2 on caveats and limitations of the methodology).\\n\\n## **Table 2-1 Total revenue and government expenditure in the West Bank and Gaza in 2017 is 27.0 and 27.4 percent, respectively, of GDP**\\n\\nTotal revenue and government expenditure, West Bank and Gaza (2017)\\n\\n|**Year of Budget Data: 2017**|**W**|**est Bank and**|**Gaza**|\\n|---|---|---|---|\\n|**Government Revenue & Expenditure**<br>**(Million NIS, unless otherwise specified)**|Total (%<br>of GDP)|Included<br>(Yes/No)|Total<br>included as<br>% of GDP|\\n\\n> 20 The CIT is better modelled with tax administrative data. Also, it is a challenge to distribute CIT burdens across households.\\n\\n10\", \"datasets\": [{\"mention_name\": {\"text\": \"tax administrative data\", \"confidence\": 0.8034429550170898, \"start\": 3193, \"end\": 3216, \"clean_text\": \"tax administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3193, \"end\": 3216, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3193, \"end\": 3216, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3193, \"end\": 3216}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3193, \"end\": 3216, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9980567097663879, \"start\": 486, \"end\": 509}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8034429550170898, \"start\": 486, \"end\": 509}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7594022750854492, \"start\": 486, \"end\": 509}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7594022750854492, \"start\": 486, \"end\": 509}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"|**Total Revenue & Grants**|27.0%|||\\n|---|---|---|---|\\n|Revenue|22.5%||18.5%|\\n|Tax Revenue|20.7%||18.5%|\\n|Direct taxes of which|1.7%||0.3%|\\n|Taxes on income, profits and capital gains, of which:|1.2%|Partial|0.3%|\\n|Corporate Income Tax|0.9%|No|0.0%|\\n|Personal Income Tax|0.3%|Yes|0.3%|\\n|Taxes on payroll and workforce|0.4%|No|0.0%|\\n|Taxes on property|0.0%|Yes|0.0%|\\n|Total Contributions to Social Insurance|1.4%|Yes|1.4%|\\n|Indirect Taxes of which:|19.0%||18.2%|\\n|General Tax on goods and services (VAT)|5.9%|Yes|5.9%|\\n|Excise|5.2%|No|0.0%|\\n|Taxes on international trade and import duties|7.1%|Yes|7.1%|\\n|Taxes on other goods, permissions|0.8%|No||\\n|**Primary Government Expenditure**|27.4%||11.6%|\\n|Social Expenditure|14.1%|Partial|11.4%|\\n|Social Assistance of which:|4.5%|Yes|2.8%|\\n|Family and children|0.0%|n/a|n/a|\\n|Social protection n.e.c.|4.5%|Yes|2.7%|\\n|Social contributions (employee benefits)|1.8%|Partial|1.4%|\\n|Education of which:|5.0%|Partial|4.0%|\\n|Basic|2.2%|Yes|2.2%|\\n|Secondary|1.7%|Yes|1.7%|\\n|Health of which:|3.9%|Partial|3.3%|\\n|Primary healthcare facilities|0.8%|Yes|0.8%|\\n|Hospital healthcare facilities|2.5%|Yes|2.5%|\\n|Non-Social Expenditure, of which:|13.3%||0.2%|\\n|Subsidies*|0.2%|Yes|Less than<br>0.2%**|\\n\\nSource: Tax revenue from PCBS. Government Financial Statistics for 2017 prepared by the Ministry of Finance. For further disaggregation, data was supplemented with reports from the Ministry of Education (for education expenditure disaggregation) and National Health Accounts (for health disaggregation).\\n\\n* Disaggregation for fuel and water subsidies is not available in the Government Financial Statistics and amounts in other technical documents included both explicit and implicit subsidies therefore it is not included here, but instead used in the macro-validation section.\\n\\n** We cannot estimate the proportion of the total subsidy amount that we cover given that we do not know the share of total subsidies allocated to types of subsidies such as non-fuel and electricity subsidies, water.\\n\\n11\", \"datasets\": [{\"mention_name\": {\"text\": \"Government Financial Statistics for 2017\", \"confidence\": 0.4355926215648651, \"start\": 1260, \"end\": 1300, \"clean_text\": \"Government Financial Statistics for 2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1260, \"end\": 1300, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1260, \"end\": 1300, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1260, \"end\": 1300}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1260, \"end\": 1300, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.89557284116745, \"start\": 133, \"end\": 173}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.4355926215648651, \"start\": 133, \"end\": 173}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8275892734527588, \"start\": 133, \"end\": 173}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8275892734527588, \"start\": 133, \"end\": 173}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"# **2.2 Pensions**\\n\\n**There are three different pension schemes in the Palestinian Territories: the unified pension scheme, the Gaza civil servant pension scheme, and the West Bank civil servant pension scheme; there is one scheme for those contributing to the Israeli pension system** (see Table 2-2). The pension system in the West Bank and Gaza is being unified through Public Pension Law No.7 of 2005, which covers public servants, security forces agents and personnel, employees of local authorities and other public institutions, Non-Governmental Organization (NGO) workers, and private sector employees. The most recent amendment is Presidential Decree No.5 of 2007, which made the pension system mandatory for private sector workers. The Palestine Pension Authority (PPA) administers the pension system, with the size of the contribution and benefits varying depending on whether the employee is subscribed to the new unified system or to the local system for civil servants in the West Bank and Gaza.\\n\\n**The unified public pension scheme covers six main categories of beneficiaries:** (i) civil servants paid by the Palestine National Authority (PNA), (ii) security personnel who are less than 45 years old as of 9/1/2006 paid by the PNA, (iii) the broad-based Palestine Liberation Organization (PLO) members paid from the PNA general budget, (iv) local authorities and public institution employees, (v) NGO employees, and (vi) private-sector workers. Contributions to the unified pension system have three main components: a defined benefit (DB) component, a defined contribution (DC) component, and a lumpsum welfare part. The minimum contribution period is 15 years of service. The benefits are disbursed as a lifetime pension following retirement. The legal age for retirement is 60 years old, but early retirement is authorized at 55 years old for men and 50 for women, provided a minimum of 15 years of service has been met.\\n\\n**In addition to the unified pension scheme, two separate pension schemes apply to civil servants in the West Bank and Gaza who were 45 years of age or older in 2006.** The Gaza civil servant pension scheme covers all the below categories of people older than age 45 as of 09/01/2006: (i) Gaza civil servants, (ii) Gaza municipality employees, (iii) Village councils’ employees, (iv) employees of the electricity distribution company in Gaza, and (v) civil servants working in the West Bank appointed after 2000 by decree of President Yasser Arafat. The legal age for retirement under this scheme is 60 with at least 15 years of active service. Early retirement is allowed with at least 20 years of service and subject to a pro-rated deduction in the pension until the age of 60. The West Bank civil servant pension scheme covers public servants who were appointed before the year 2000 to the service in the West Bank and are older than age 45 as of 09/01/2006.\\n\\n**The National Insurance Law, first introduced in 1953, governs the Israeli state pension, which comprises a universal insurance pension combined with means-tested income support.** Mandatory contributions to defined contribution pension funds were introduced in January 2008. (OECD, n.d.). The National Insurance Institute administers the pension system, which collects contributions and pays out benefits. A major reform recently enacted has been raising of the retirement age for recipients of old-age allowance from age 65 for men and age 60 for women to 67 for men and 64 for women.\\n\\n**Table 2-2 There are three pension schemes applicable in the West Bank and Gaza, and one pension scheme for Palestinian employees working in Israel and the settlements**\\n\\nSummary of pension scheme rates in the West Bank and Gaza, 2017\\n\\n|**Scheme**<br>**Simulation**|**Total**|**Employee**<br>**Employer**|\\n|---|---|---|\\n\\n12\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"|Unified<br>pension<br>system|Eligible and under 55 years of age in 2016.<br>Defined Benefit and Defined Contribution +<br>welfare lump sum of 700 NIS. We only<br>allocate the Defined Benefit and welfare<br>lump-sum, as according to the 2016 Public<br>Expenditure Review (PER). The Defined<br>Contribution proportion is not yet<br>effective.|DB<br>DC|16%<br>6%|7%<br>3%|9%<br>3%|\\n|---|---|---|---|---|---|\\n|Gaza civil<br>servant<br>pension<br>scheme|Over 55 years of age in 2016, resident in<br>Gaza & working in the public sector.|22.|5%|10%|12.5%|\\n|West Bank<br>civil servant<br>pension<br>scheme|Over 55 years of age in 2016, resident in<br>the West Bank & working in the public<br>sector.|2|%|2%|0%|\\n|Israeli scheme|Allocated to employees only who claim<br>that their employers make pension<br>contributions.|17.|5%|5.5%|12%|\\n\\nSource: World Bank Public Expenditure Review, 2016.\\n\\n# **2.3 Tax revenues**\\n\\n**All taxes are collected by the central government’s Ministry of Finance (MoF).** Local government units do not collect any type of tax, except for “Education Tax or Fees”, and a building tax for operators<sup>21</sup> (90 percent of the tax is distributed to municipalities, and 10 percent is remitted to the national accounts) (Abdelkarim & Jaber, 2018). Two public departments within the MoF are responsible for tax collection (except for property tax, which is collected by the General Administration of Property Tax): the General Administration of Customs and VAT and the General Directorate of Income Tax (GDIT). Collection of tax at source from salaries and wages is a smooth process with relatively few administrative complications; employers deduct the tax directly from employees’ incomes and transfer these to the MoF. Private taxpayers follow the same procedures. Israel collects clearance revenues<sup>22</sup> and customs duties and then transfers them to the PA after deducting amounts owed by the Palestinians for Israeli electricity and water utilities.\\n\\n## **2.3.1 Direct taxes**\\n\\n## **Personal income tax (PIT)**\\n\\n**The applicable tax law (Income Tax Law number 8 of 2011 and its amendments) states that “all realized incomes in Palestine from any source for any person are taxable unless there is a clear tax exemption in the Income Tax Law”.** Palestinian residents and nonresidents are taxed only on income sourced in the Palestinian Territories. West Bank residents pay PIT to the PA, unless they work in Israel, in which case\\n\\n> 21 A tax building operators pay used to finance infrastructure expenditure on public schools.\\n\\n> 22 Clearance revenues include import taxes, taxes on Palestinian workers in Israeli-administered territories, and taxes on economic activity in Area C. Source: World Bank (March 2023).\\n\\n13\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"they first pay the PIT to the GoI and then it is transferred to the PA as part of clearance revenues. East Jerusalem residents pay PIT to the GoI, unless they work in the West Bank, in which case they pay to the PA. Gaza residents do not pay PIT, except for PA workers in Gaza since the tax is deducted at source.\\n\\n**Taxable income is composed of income from all sources after deducting expenses.** Unless specifically exempt by law, all income is included, less allowable expenses incurred to produce the income and the standard deduction (carryover losses, exemptions, and donations). The individual income tax rate charged is progressive, ranging from 5 to 15 percent (Table 2-3) in the Palestinian Territories and from 1 to 5 percent in Israel. In Israel a 10 percent flat rate is applied to rental income.\\n\\n**Table 2-3 The income tax framework in the West Bank and Gaza has only three brackets, and**\\n\\n**statutory rates range between 5 and 15 percent**\\n\\nStatutory income tax rates for Israel and the Palestinian Territories\\n\\n|**Israel**||**Palestinian Territ**|**ories**|\\n|---|---|---|---|\\n|<br>**Bracket**|**Rate**|**Bracket**|**Rate**|\\n|1 1 - 62,640|0.1|1 - 75,000|0.05|\\n|2 62,641 - 107,040|0.14|75,000 - 150,000|0.1|\\n|3 107,041 - 166,320|0.21|> 150,000|0.15|\\n|4 166,321 - 237,600|0.31|||\\n|5 237,601 - 496,920|0.34|||\\n|6 496,921 - 803,520|0.48|||\\n|7 > 803,521|0.5|||\\n\\nSource: Harris Consulting & Tax Ltd. (2017); PWC (2021)\\n\\n# **Property tax**\\n\\n**Property tax is collected according to the PA General Administration of Property Tax and then transferred to municipalities.** The Jordanian Property Tax Law no. 11 of 1954 regulates the property tax for properties in the West Bank, while the Property Tax Law in Cities No. 42 of 1940 and the Property Tax Law in Villages, No. 5 of 1942 regulate properties in Gaza. Property tax is imposed on owners of the property or building, both resident and nonresident, and varies for land (unconstructed property) and building. The tax rate varies by municipality and by occupation status of the building (rented, prepared in preparation for rental, and inhabited by the owner). For buildings, property tax is imposed at a rate of 17 percent of the annual rental value of the building after depreciation/lease cost (20 percent of value). The tax on unconstructed buildings is 10 percent of the property valuation, based on the unit value of the land.\\n\\n# **2.3.2 Indirect taxes**\\n\\n14\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"## **Customs duties**\\n\\n**Imports customs regulations and procedures in the Palestinian Territories and Israel are determined within the framework of Israeli customs laws, international customs agreements, and the Paris Protocol signed by the GoI and the Palestinian Liberation Organization (PLO) in 1994.** Valuation of commodities for customs purposes is based upon the 1994 General Agreement on Tariffs and Trade (GATT). The Palestinian customs system is laid out in Customs Law No. 1 of 1962, which specifies its functions, procedures, and transparency and integrity standards.\\n\\n## **Value-Added Tax (VAT)**\\n\\n**VAT is levied on the sale of goods, the provision of services, and imports.** The PA does not currently collect any domestic VAT in Gaza. Domestic VAT is only collected in the West Bank. Import VAT—one of the taxes making up clearance revenues that Israel collects on behalf of the PA—is transferred to the PA’s MoF in the West Bank. In 2017 (the year of analysis), the statutory VAT rate was 16 percent. Under the Paris Protocol, The PA VAT rate should be within the range of maximum (minimum) 2 percentage points above (below) the Israeli VAT rate. Regarding preferential VAT, only crop production and commercial transactions carried out by charitable societies and donor organizations are subject to “zero rates”. In 2017 the main VAT exemptions were applied to tourism services (purchases by foreigners), fresh fruit and vegetables, and exports.\\n\\n## **Excises**\\n\\n**In the Palestinian territories, excise taxes are applied domestically to a selection of products** . This includes alcoholic beverages, petroleum, vehicles, and tobacco (regular and flavored hookah tobacco).<sup>23</sup> West Bank residents must pay excises to the PA (according to PA rates) for alcoholic beverages and tobacco purchased in the West Bank and for fuel bought by the PA from Israel; also, PA tax rates are applied for imported goods from third countries. East Jerusalem residents pay to the GoI according to the Israeli tax rates for goods imported from third countries or purchased in East Jerusalem; they pay to the PA directly on alcohol and tobacco purchased in the West Bank. In 2017 the Israeli excise rates for the fuel products were 2.89 NIS/L for gasoline, NIS 3.02 NIS/L for diesel and 2.89 NIS/L for kerosene; it is expected that PA excise rates are similar to avoid arbitrage across countries.\\n\\n# **2.4 Transfers**\\n\\n## **2.4.1 Direct transfers Social Protection**\\n\\n**The PA’s Ministry of Social Development (MoSD) is responsible for provision of basic social protection to the most vulnerable families in the West Bank and Gaza through various cash and near cash programs.** There are four main components of social protection programs in the West Bank and Gaza: social assistance, social insurance, social services, and social justice.\\n\\n**The primary government-funded social protection program is the national Cash Transfer Program (CTP).** In the year 2017, the CTP represented 63 percent of the total MoSD budget.<sup>24</sup> The CTP aims to reach households below the extreme poverty line and marginalized households between the national and\\n\\n> 23 Import Laws and Procedures, 2022, Investment Promotion and Industrial Estates Agency, <u>Import Laws and Procedures (pipa.ps).</u>\\n\\n> 24 Annual Statistical Report (2017), Ministry of Social Development.\\n\\n15\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"extreme poverty lines.<sup>25</sup> Eligibility for the beneficiaries is determined by a Proxy Means Test (PMT) based on household expenditure and consumption data, including 34 variables related to “characteristics of the household head (gender, health, and disability status), marital status (divorced, widowed), household size and composition (children, elderly), employment status, assets (cars, livestock, and property including condition and building materials), and health of household members” (Wong et al 2019, 12). Participating households receive cash transfers quarterly of between NIS 750 and NIS 1,800<sup>26,27</sup> directly into their bank accounts. In addition to the CTP, the MoSD administers a number of other cash assistance programs including cash assistance to orphans and distressed families, cash assistance to discharged government employees, and emergency cash assistance for women. Near-cash social assistance includes food assistance/daily rations to the beneficiaries.\\n\\n# **2.4.2. Indirect subsidies**\\n\\n**A fundamental characteristic of the West Bank and Gaza’s energy supply is its reliance on imports, particularly from Israel, Jordan, and the Arab Republic of Egypt.** Almost all fossil fuels are imported.\\n\\n## **Electricity**\\n\\n**Residents of the West Bank and Gaza rely on imported electricity, primarily from Israel.** The imported share is about 99 percent of total supply in the West Bank and 64 percent in Gaza.<sup>28</sup> The Palestinian Energy and Natural Resources Authority (PENRA) is responsible for provision of electricity and for energy policy formulation. There are five electricity distribution companies (DISCOs) and a single buyer of electricity, the recently established Palestinian Electricity Transmission Company (PETL). The PETL sources electricity from the Palestine Power Generation Company (PPGC), Israel, and other neighboring countries. In Gaza, the Palestine Electric Company (PEC) operate the Gaza Power Plant at partial capacity.\\n\\n**The current electricity tariff structure for domestic consumers (Table 2.4) was promulgated in 2017 in Article 5, Ministerial decision N 03/137/17.** It applies in all regions of the Palestinian Territories except for residents of Jericho and the Palestinian Jordan Valley. Access to the electricity is widespread with 98 percent of the PECS survey respondents in all the regions of the Palestinian Territories reporting having access in 2017. Most consumers (approximately 65 percent) consume in slab 2 (161-250 kWh) and slab 3 (251-400 kWh) of the tariff structure.\\n\\n**Table 2-4 The price of electricity per kWh for domestic consumers ranges between 0.458 and 0.6731 in the West Bank and Gaza**\\n\\nPrice of electricity by consumption slab, West Bank and Gaza\\n\\n|**Consumption slabs**|**Price**|\\n|---|---|\\n|**(kWhper month)**|**(NISper kWh)**|\\n|1- 160|0.458|\\n\\n> 25 As of its 2017 Annual Statistical Report, the MoSD (2018) estimated the extreme poverty line at NIS 1,970 per month and the national poverty line at NIS 2,470 per month for a family of two adults and three children, with the specific amount determined by the PMT.\\n\\n> 26 Ministry of Social Development (Jan 2018); Ministry of Social Affairs (Jan 2017); The Office of the European Union Representative (Oct 2019); The World Bank (Dec 2018).\\n\\n> 27 The specific amount depends on households PMT score.\\n\\n> 28 <u>https://www.worldbank.org/en/country/westbankandgaza/brief/securing-energy-for-development-in-west-bank-and-gazabrief</u>\\n\\n16\", \"datasets\": [{\"mention_name\": {\"text\": \"household expenditure and consumption data\", \"confidence\": 0.5623053908348083, \"start\": 121, \"end\": 163, \"clean_text\": \"household expenditure and consumption data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 121, \"end\": 163, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 121, \"end\": 163, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 121, \"end\": 163}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 121, \"end\": 163, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.22910325229167938, \"start\": 121, \"end\": 163}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5623053908348083, \"start\": 121, \"end\": 163}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7792448401451111, \"start\": 121, \"end\": 163}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7792448401451111, \"start\": 121, \"end\": 163}}, {\"mention_name\": {\"text\": \"PECS survey\", \"confidence\": 0.5900849103927612, \"start\": 2335, \"end\": 2346, \"clean_text\": \"PECS survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2335, \"end\": 2346, \"clean_text\": \"\"}, \"producer\": {\"text\": \"MoSD\", \"confidence\": 0.46575045585632324, \"start\": 2907, \"end\": 2911, \"clean_text\": \"MoSD\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2335, \"end\": 2346}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2335, \"end\": 2346, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9984906911849976, \"start\": 23, \"end\": 34}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5900849103927612, \"start\": 23, \"end\": 34}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8771663904190063, \"start\": 23, \"end\": 34}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8771663904190063, \"start\": 23, \"end\": 34}}, {\"mention_name\": {\"text\": \"households PMT score\", \"confidence\": 0.32420358061790466, \"start\": 3337, \"end\": 3357, \"clean_text\": \"households PMT score\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3337, \"end\": 3357, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3337, \"end\": 3357, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3337, \"end\": 3357}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3337, \"end\": 3357, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.6733443140983582, \"start\": 436, \"end\": 456}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.32420358061790466, \"start\": 436, \"end\": 456}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6061109900474548, \"start\": 436, \"end\": 456}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6061109900474548, \"start\": 436, \"end\": 456}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"|161-250|0.4938|\\n|---|---|\\n|251- 400|0.5695|\\n|401-600|0.6089|\\n|More than 600|0.6731|\\n\\nSource: Prime Minister’s Office (2017).\\n\\n**The electricity tariffs are low relative to the cost-of-service provision, leading to subsidies (explicit and**\\n\\n**implicit) in the sector.** The subsidy to domestic consumers is mainly provided in the form of a Tariff Differential Subsidy (TDS). The regulatory authority, PERC, has set a uniform tariff of NIS 0.53– 0.56 (US$0.14–0.15) per kWh for the Palestinian DISCOs for purchase of electricity. However, financial analysis of the sector suggests that the full cost of service provision—given current inefficiencies—ranges from about NIS 0.66 to NIS 1.42 (US$0.18–0.39) per kWh, depending on the DISCO. The mechanism currently used to safeguard affordability for some consumers is a “Rising Block Tariff or Incremental Block Tariff”, whereby the tariff of the previous slab is granted to all the units consumed up to the lower bound of the slab in which their consumption falls. In the West Bank and Gaza, with the first block of 160 kWh per month currently set at NIS 0.43 (US$0.11) per kWh, matching the lowest income decile’s ability to pay. However, given that average residential electricity consumption in the West Bank and Gaza is only 200– 300 kWh per month, this means that those wealthier Palestinian consuming most of the electricity benefit the most from this subsidized rate.\\n\\n# **Fuel**\\n\\n**As for electricity, the Palestinian population relies on fuel imports.** Under the Paris Economic Protocol (PEP), the West Bank and Gaza is allowed to import petroleum from sources other than the Israeli supplier provided that the final consumer price in the Palestinian Territories is less than 15 percent lower than in Israel. The clause further prescribes that Israel shall collect taxes and customs at crossing points.\\n\\n**The General Directorate of Petroleum (GDoP) is charged with the responsibility of regulating imports of oil and petrochemicals.** The GDoP purchases an average of 80 million liters of fuel a month, including 25 million liters of gasoline, 55 million liters of diesel, and 14,000 tons of cooking gas and other byproducts like kerosene and gas. It is also the monitoring and licensing agency that sets petroleum and petrochemical prices.\\n\\n**Fuel prices in Israel depend on the world oil price index.** Average prices are calculated monthly based on the last five workdays of the previous month. The GDoP is provided with the Israeli refineries’ costs, profit margin, and the logistic expenses. A deduction is applied for the global gasoline evaporation rate (0.005), and addition of the Blu Tax and VAT taxes.\\n\\n**Fuel prices in the Palestinian Territories do not fully reflect fluctuations in global oil prices due to the addition of taxes and subsidies, and exchange rate fluctuations with the dollar.** In general, taxes represent an average of 68 percent of total purchase price. Government fuel subsidies on the consumption of petrol, gas, and other fuel products help alleviate price increases for Palestinian consumers. The GDoP calculates fuel prices at the beginning of every month after consultation with the Minister of Finance to determine the size of financial subsidies, which depends on the financial position of the government, liquidity available in the Treasury, and fluctuations in global oil prices. Retail prices in the West Bank and Gaza are lower than the global price due to the subsidies, on top of which the retailer\\n\\n17\", \"datasets\": [{\"mention_name\": {\"text\": \"world oil price index\", \"confidence\": 0.4421447515487671, \"start\": 2339, \"end\": 2360, \"clean_text\": \"world oil price index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2339, \"end\": 2360, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2339, \"end\": 2360, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2339, \"end\": 2360}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2339, \"end\": 2360, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.5549001693725586, \"start\": 291, \"end\": 312}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4421447515487671, \"start\": 291, \"end\": 312}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8387428522109985, \"start\": 291, \"end\": 312}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8387428522109985, \"start\": 291, \"end\": 312}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"adds a fixed profit margin. The global price, less the subsidy, plus the retailer’s profit margin determines the final price consumers pay at the pump (Prime Minister’s Office, 2017).\\n\\n# **2.5 In-kind benefits**\\n\\n## **Health**\\n\\n**The Ministry of Health (MoH) is responsible for provision of health services in the West Bank and Gaza.** The Palestinian health system consists of four main sectors including: (i) the government health sector (the Palestinian Ministry of Health and Military Medical Services), (ii) the United Nations Relief and Works Agency (UNRWA), (iii) NGOs, and (iv) the private sector. Primary healthcare provision is shared across the different sectors, while the provision of secondary and tertiary healthcare is led by the government health sector. From the supply side, the government health sector provides about 63 percent of the primary healthcare facilities, while in Gaza most residents tend to access UNRWA health facilities.<sup>29</sup>\\n\\n**Health care services are provided to citizens at three levels: primary, secondary, and tertiary.** Primary health care is provided by centers classified into four categories based on the combination of preventive and curative services offered, in addition to mobile clinics (Ministry of Health, 2017).\\n\\n**All residents of Gaza enjoy free health insurance under article (2) of the presidential decree issued on 26/6/2007, while health insurance in the West Bank is for participating families only, although contributions are mainly symbolic.** Health insurance available in the West Bank includes a compulsory health insurance scheme for public sector employees, municipalities, and retirees, as well as other types of health insurance including voluntary insurance, insurance for workers within the Green Line,<sup>30</sup> group insurance, social affairs insurance, and prisoners’ insurance.\\n\\n## **Education**\\n\\n**The Ministry of Education and Higher Education (MoEHE) manages public schools, including regulatory overview of private schools.** This includes pre-school, primary (or basic) and secondary school (grades 1 through 12), technical and vocational, non-formal, and higher education. The MoEHE is the official body responsible for running, organizing, and developing the educational sector either through direct management or supervision. It also spearheads the national education strategic planning process. The Schools’ Census 2016-17 showed that there were about 2,963 schools in the Palestinian Territories, 73 percent of them administered by the government, and about 1,229,756 students were enrolled, about 65 percent of them in government schools.\\n\\n# **2.6 International comparison**\\n\\n**Tax revenue in the West Bank and Gaza as a share of GDP is high at about 20.7 percent.** This compares to an average of 18.1 percent for all countries included in the CEQ Institute’s data center. It is also high compared to the average of lower middle-income countries of about 17.9 percent of GDP (indicated by the dashed orange line in Figure 2-1).\\n\\n> 29 While UNRWA’s mandate is to provide services to refuges, its health services are mainly focused on primary health care, and it purchases secondary/tertiary services for a limited number of cases.\\n\\n> 30 Workers within the Green line refers to individuals working in Israel. The Green Line is the demarcation line set out in the 1949 Armistice Agreements between the armies of Israel and those of its neighbors (Egypt, Jordan, Lebanon and the Syrian Arab Republic) after the 1948 War.\\n\\n18\", \"datasets\": [{\"mention_name\": {\"text\": \"Schools’ Census 2016-17\", \"confidence\": 0.4835119843482971, \"start\": 2395, \"end\": 2418, \"clean_text\": \"Schools’ Census 2016-17\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2395, \"end\": 2418, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2395, \"end\": 2418, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2395, \"end\": 2418}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2395, \"end\": 2418, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.9748937487602234, \"start\": 928, \"end\": 951}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.4835119843482971, \"start\": 928, \"end\": 951}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7823134064674377, \"start\": 928, \"end\": 951}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7823134064674377, \"start\": 928, \"end\": 951}}, {\"mention_name\": {\"text\": \"CEQ Institute’s data center\", \"confidence\": 0.30812016129493713, \"start\": 2843, \"end\": 2870, \"clean_text\": \"CEQ Institute’s data center\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2843, \"end\": 2870, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2843, \"end\": 2870, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2843, \"end\": 2870}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2843, \"end\": 2870, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.2875747084617615, \"start\": 571, \"end\": 598}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.30812016129493713, \"start\": 571, \"end\": 598}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6564701795578003, \"start\": 571, \"end\": 598}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6564701795578003, \"start\": 571, \"end\": 598}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"**Figure 2-1 Tax revenues as a share of GDP are higher than the sample average at 20.7 percent**\\n\\n<!-- Start of picture text --> International comparability of tax revenue<br>40%<br>West Bank and Gaza<br>35%<br>(2017), 20.7%<br>30%<br>25%<br>20%<br>15%<br>10%<br>5%<br>0%<br>Country (CEQ assessment year of survey)<br>Tax revenue as % of GDP (ranked lowest to highest) Iraq (2018) Mexico (2012) Pakistan(2018) Guatemala (2011) Sri Lanka (2009) Ethiopia Indonesia (2012) Ecuador (2011) Paraguay (2014) Costa Rica (2010) Ghana (2012) Honduras (2011) Burkina Faso (2014) Russia (2010) El Salvador (2011) Colombia (2014) Peru (2009) Peru (2011) Chile (2013) Armenia (2011) Lesotho (2017) China (2014) El Salvador (2015) Tunisia (2010) South Africa (2015) Argentina (2012) Turkey (2014) Georgia (2013) United States (2016) Romania (2016) Bolivia (2009) South Africa (2010) Namibia (2016) Tajikistan (2015)<br>Dominican Republic (2013) West Bank and Gaza (2017)<br><!-- End of picture text --> Source: CEQ Data Center, Government Finance Statistics Manual West Bank and Gaza\\n\\n**West Bank and Gaza’s social expenditure as a share of GDP is also a high 14.1 percent.** This compares to the average among both lower middle-income countries (9.4 percent) and upper middle-income countries (11.5 percent) in the sample of countries included in CEQ Institute’s Data Center.<sup>31</sup> It is also high compared to the average of 10.3 percent of all countries in the CEQ data center sample (indicated by the dashed red line in Figure 2-2).\\n\\n> 31 CEQ Data Center on Fiscal Redistribution (https://commitmentoequity.org/datacenter) accessed June 2023.\\n\\n19\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"**Figure 2-2 Social expenditures as a share of GDP are higher than the sample average at 14.1 percent**\\n\\n<!-- Start of picture text --> International comparability of government social expenditure<br>25%<br>West Bank and Gaza<br>(2017), 14.1%<br>20%<br>15%<br>10%<br>5%<br>0%<br>Country (CEQ assessment year of survey)<br>Ivory Coast (2015) Iran (2011) Uganda (2016) Indonesia (2012) Guatemala (2014) Albania (2015) Ethiopia Tanzania (2011) Iraq (2018) Togo (2015) Ghana (2012) Armenia (2011) Egypt (2015) Tunisia (2010) Mexico (2014) Colombia (2010) El Salvador (2011) Paraguay (2014) Ecuador (2011) El Salvador (2013) Romania (2016) Mexico (2010) Chile (2013) Honduras (2011) South Africa (2015) Georgia (2013) Croatia (2014) Venezuela (2013) Belarus (2015) Bolivia (2009) Russia (2010) Brazil (2009) Lesotho (2017) Argentina (2012) Botswana (2010)<br>Social expenditure as % of GDP (ranked lowest to highest)<br>West Bank and Gaza (2017)<br><!-- End of picture text --> Source: CEQ Data Center, Government Finance Statistics Manual West Bank and Gaza.\\n\\n20\", \"datasets\": [{\"mention_name\": {\"text\": \"CEQ assessment year of survey\", \"confidence\": 0.3097810745239258, \"start\": 288, \"end\": 317, \"clean_text\": \"CEQ assessment year of survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 288, \"end\": 317, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 288, \"end\": 317, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 288, \"end\": 317}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 288, \"end\": 317, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.5700691938400269, \"start\": 288, \"end\": 317}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3097810745239258, \"start\": 288, \"end\": 317}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.558051347732544, \"start\": 288, \"end\": 317}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.558051347732544, \"start\": 288, \"end\": 317}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"# **3 Data and methodology**\\n\\n### **The objective of this fiscal incidence analysis is to analyze the impact of taxes and social expenditure on**\\n\\n**households’ welfare, poverty, and inequality in the West Bank and Gaza.** Although designing and implementing fiscal policies should consider different dimensions—including fiscal sustainability, economic efficiency, externalities, and equity—the motivation for performing our analysis is fourfold:\\n\\n- (i) There has not been a comprehensive distributional assessment of the country’s fiscal system.\\n\\n- (ii) The current government is operating in an environment of increasing fiscal pressure, with an urgent need to mobilize revenue and increase fiscal space.\\n\\n- (iii) Given the above, findings from the fiscal incidence analysis create an evidence base for how to promote tax and expenditure reforms consistent with both fiscal sustainability and fiscal equity.\\n\\n- (iv) In addition to helping assess distributional repercussions of existing fiscal policies, developing detailed microsimulation models for fiscal incidence analysis also creates a platform to simulate distributional impact of potential policy reforms.\\n\\n## **3.1 Data sources**\\n\\n### **The fiscal incidence analysis in the Palestinian Territories uses the following data sources:**\\n\\n- The most recent, nationally representative household survey, the Palestinian Expenditure and Consumption Survey (PECS) 2016/17, is a multi-purpose survey on household budget and living standards used to estimate official poverty statistics for the Palestinian Territories. PECS was implemented by the Palestinian Central Bureau of Statistics (PCBS). The PECS 2016/2017 is a 12month survey covering 3739 households and implemented from October 2016 to September 2017. The survey has information on income, employment, consumption, health, and education.\\n\\n- Tax legislation and comprehensive administrative data on direct and indirect taxes, tax rates, tax collection, and executed government expenditure data from MoF and World Bank teams.\\n\\n- Social protection data from the main social programs on budget, beneficiaries, and allocation rules from the MoSD and the World Bank Social Protection team.\\n\\n- Data to calculate the subsidy value for domestic electricity distribution and fuel from the Ministry of Energy and the World Bank Energy team.\\n\\n- Education expenditures and student enrollment from the Ministry of Education and World Bank Education team.\\n\\n- Health expenditure and health visits from the Ministry of Health and the World Bank Health team\\n\\n- Supply and use tables from 2017 for the West Bank and Gaza from the World Bank Macroeconomics, Trade, and Investment team.\\n\\n- The 2017 Labor Force Survey (LFS) on labor informality and public sector employees, also implemented by the PCBS.\\n\\n## **3.2 Commitment to Equity (CEQ) framework**\\n\\n### **This study uses the Commitment to Equity (CEQ) methodology to implement a fiscal incidence analysis**\\n\\n**of the main taxes and public social transfers of the West Bank and Gaza in 2017** . A CEQ Assessment is a rigorous and standardized fiscal incidence methodology that permits systematic analysis of how different tax and public expenditure options affect the welfare distribution. CEQ uses a common framework\\n\\n21\", \"datasets\": [{\"mention_name\": {\"text\": \"Palestinian Expenditure and Consumption Survey\", \"confidence\": 0.7572322487831116, \"start\": 1363, \"end\": 1409, \"clean_text\": \"Palestinian Expenditure and Consumption Survey\"}, \"acronym\": {\"text\": \"PECS\", \"confidence\": 0.5991594195365906, \"start\": 1411, \"end\": 1415, \"clean_text\": \"PECS\"}, \"producer\": {\"text\": \"Palestinian Central Bureau of Statistics\", \"confidence\": 0.8113417029380798, \"start\": 1599, \"end\": 1639, \"clean_text\": \"Palestinian Central Bureau of Statistics\"}, \"reference_year\": {\"text\": \"October 2016 to September 2017\", \"confidence\": 0.30747005343437195, \"start\": 1733, \"end\": 1763}, \"geography\": {\"text\": \"Palestinian Territories\", \"confidence\": 0.7074848413467407, \"start\": 1236, \"end\": 1259, \"clean_text\": \"Palestinian Territories\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9302723407745361, \"start\": 575, \"end\": 621}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7572322487831116, \"start\": 575, \"end\": 621}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9354488849639893, \"start\": 575, \"end\": 621}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9354488849639893, \"start\": 575, \"end\": 621}}, {\"mention_name\": {\"text\": \"PECS 2016/2017\", \"confidence\": 0.3956655263900757, \"start\": 1652, \"end\": 1666, \"clean_text\": \"PECS 2016/2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1652, \"end\": 1666, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1652, \"end\": 1666, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1652, \"end\": 1666}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1652, \"end\": 1666, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9991981387138367, \"start\": 864, \"end\": 878}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3956655263900757, \"start\": 864, \"end\": 878}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.720209538936615, \"start\": 864, \"end\": 878}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.720209538936615, \"start\": 864, \"end\": 878}}, {\"mention_name\": {\"text\": \"tables from 2017\", \"confidence\": 0.33290016651153564, \"start\": 2571, \"end\": 2587, \"clean_text\": \"tables from 2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2571, \"end\": 2587, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2571, \"end\": 2587, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2571, \"end\": 2587}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2571, \"end\": 2587, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.28962263464927673, \"start\": 23, \"end\": 39}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.33290016651153564, \"start\": 23, \"end\": 39}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.43504202365875244, \"start\": 23, \"end\": 39}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.43504202365875244, \"start\": 23, \"end\": 39}}, {\"mention_name\": {\"text\": \"2017 Labor Force Survey\", \"confidence\": 0.7575470209121704, \"start\": 2686, \"end\": 2709, \"clean_text\": \"2017 Labor Force Survey\"}, \"acronym\": {\"text\": \"LFS\", \"confidence\": 0.4586111009120941, \"start\": 2711, \"end\": 2714, \"clean_text\": \"LFS\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2686, \"end\": 2709, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"October 2016 to September 2017\", \"confidence\": 0.4642725884914398, \"start\": 1733, \"end\": 1763}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2686, \"end\": 2709, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9993205070495605, \"start\": 1024, \"end\": 1047}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7575470209121704, \"start\": 1024, \"end\": 1047}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8020520806312561, \"start\": 1024, \"end\": 1047}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8020520806312561, \"start\": 1024, \"end\": 1047}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"developed by the CEQ Institute and presented in the CEQ Handbook (Lustig, 2022). CEQ Assessments have been implemented in over 60 countries at the time of writing this report.\\n\\n# **The first step of fiscal incidence analysis is allocation of the main taxes and transfers to households in**\\n\\n**the survey.** The main modeling techniques in this report include direct identification, inference, imputation, simulation, and prediction. These concepts are described in Box 3.1, below. Typically, fiscal interventions such as direct taxes, indirect taxes, and indirect subsidies are modeled based on simulation, combining tax rules and characteristics of individual’s employment and consumption patterns, respectively.\\n\\n**Box 3.1: Tax and transfer allocation methods**\\n\\n<mark>We allocate each tax and transfer by drawing from a set of methods described in detail in Chapter</mark> 3 of the CEQ Handbook, namely:\\n\\n- **Direct identification:** The survey asks whether a tax (transfer) was paid (received), and the amount.\\n\\n- **Inference:** Whether a tax (transfer) was paid (received), or the amount, is determined based on administrative data of the taxes (transfers) and other relevant secondary information self-reported by the household.\\n\\n- **Imputation:** Taxpayers (beneficiaries) are directly identified in the survey, but the amount of a tax (subsidy) paid (received) is imputed based on program rules.\\n\\n- **Simulation:** Both the identification of taxpayers (beneficiaries) and the amount of a tax (subsidy) paid (received), is estimated based on program rules.\\n\\n- **Prediction:** A regression model is applied to predict who receives benefits or pays taxes, and/or the amount.\\n\\nAdapted from Chapter 3 of Lustig (2022).\\n\\n**The building block of fiscal incidence analysis is the construction of income concepts.** Conceptually, the analysis starts from a household’s pre-fiscal income<sup>32</sup> —that is, any private income that excludes any government intervention—and each new income concept is constructed by including different layers of the fiscal system, or categories of taxes and transfers, as shown in Figure 3-1, until a household’s total post-fiscal income is estimated (income once all taxes and transfers are included).\\n\\n**Once completed, we can estimate the distributional impacts of any tax or transfer decisions on the welfare of a country’s households** . Once all fiscal instruments and income concepts are constructed, the analysis measures the main distributional repercussions, such as changes in number of households living under the poverty line, identification of groups which are net cash-beneficiaries, and impact of any tax and/or government’s transfers on inequality and progressivity, as well as the marginal contribution of each fiscal intervention to poverty and inequality reduction.\\n\\n> 32 Pre-fiscal income could be equivalent to Market Income or Market Income plus pensions depending on the pension scenario chosen for the baseline. This is explained in more detail later in this section.\\n\\n22\", \"datasets\": [{\"mention_name\": {\"text\": \"administrative data\", \"confidence\": 0.5971437692642212, \"start\": 1117, \"end\": 1136, \"clean_text\": \"administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1117, \"end\": 1136, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1117, \"end\": 1136, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1117, \"end\": 1136}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1117, \"end\": 1136, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9913018941879272, \"start\": 346, \"end\": 365}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5971437692642212, \"start\": 346, \"end\": 365}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6196249723434448, \"start\": 346, \"end\": 365}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6196249723434448, \"start\": 346, \"end\": 365}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"**In countries where consumption is a more reliable measure of welfare than income (as in the West Bank and Gaza), the analysis starts by setting Disposable Income equal to the consumption welfare aggregate and adds or removes certain fiscal interventions to calculate the other income concepts.** Given that Disposable Income already includes certain interventions of the fiscal system, the analysis can move “backwards” to remove fiscal interventions until reaching “pre-fiscal income”, a measure of households’ total income from private sources only. Also, the analysis works “forwards” from Disposable income, adding additional fiscal interventions until reaching “post-fiscal income” (income once all taxes and transfers modeled here have been included). Figure 3-1 displays the different steps of the income construction process.\\n\\n**The analysis applies this framework to the West Bank & Gaza’s fiscal system as follows:** Net Market Income is calculated from Disposable Income by removing direct social protection transfers (government cash aid and government near-cash aid). Market Income plus pensions is constructed by adding direct taxes (PIT, property tax, and payroll tax) back into Net Market Income. Market Income is calculated from Market Income plus pensions by subtracting pension income and adding pension contributions. The analysis then constructs the rest of the income concepts forward. Consumable Income is constructed by subtracting indirect taxes (the direct and indirect effects of VAT and fuel excises, and the direct effects of customs duties) from Disposable Income and by adding indirect subsidies (the direct effects of electricity subsidies and the direct and indirect effects of fuel subsidies). Lastly, Final Income is constructed by adding in-kind benefits from public education and public health to Consumable Income.<sup>33</sup>\\n\\n**In the case of the West Bank and Gaza, this fiscal incidence analysis uses the “Pensions as Government Transfer” (PGT) scenario from the CEQ framework.** As the name suggests, in this scenario pensions are treated as a government transfer—a tool for redistribution from one segment of society to another. In the PGT scenario, the pre-fiscal income concept is Market income.<sup>34</sup> The alternative to the PGT scenario is the “Pensions as Deferred Income” (PDI) scenario, in which pensions are treated as private income that is saved. In this scenario, the pre-fiscal income concept is Market Income plus pensions (see results under PDI scenario in Annex A).\\n\\n**Finally, we consider different post-fiscal income concepts for calculating the impact of fiscal interventions on poverty and inequality.** For calculating total effect on _inequality_ reduction, the postfiscal income is the Final Income concept shown in Figure 3-1—that is, the total effect is the difference in - inequality levels between Market Income and Final Income. However, when calculating total _poverty_ reduction, the post-fiscal income is the Consumable Income concept shown in Figure 3-1 (that is, the total effect is the difference in the poverty rate between Market Income and Consumable Income). The reason for this is that in-kind health and education benefits are allocated at the average government cost of provision, which provides little information about how the transfers improve a household’s purchasing power or their poverty status. Education benefits, for example, will accrue in terms of future purchasing power but the CEQ Framework is limited to measuring distributional effects at a specific point-in-time (see Box 3.2 on CEQ caveats).\\n\\n> 33 These benefits are modeled as the average government cost of education and health services. The model does not include user fees nor subtract the incidence of user-fees from Final Income.\\n\\n> 34 As a result, we subtract pension income from Disposable income along with the direct transfers to get to Net Market Income, and we add pension contributions with the direct taxes to reach Market Income.\\n\\n23\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"# **Figure 3-1 Income concepts in the CEQ framework**\\n\\nCEQ Income Concepts in the Palestinian Territories; Pensions as a Government Transfer scenario <!-- Start of picture text -->\\n- Direct transfers<br>Government cash<br>and in-kind, and<br>pensions<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> + Indirect subsidies<br>Electricity and Fuel<br>subsidies<br><!-- End of picture text --> <!-- Start of picture text --> + In-kind transfers<br>Education & Health<br><!-- End of picture text --> <!-- Start of picture text --> Final income<br><!-- End of picture text --> <!-- Start of picture text --> Market income<br><!-- End of picture text --> <!-- Start of picture text --> Disposable income<br>Official<br>consumption<br>expenditure from<br>the PECS 2016/17<br><!-- End of picture text --> <!-- Start of picture text --> Net market income<br><!-- End of picture text --> <!-- Start of picture text --> Consumable<br>income<br><!-- End of picture text --> <!-- Start of picture text --> + Direct taxes<br>Personal Income Tax, Pension<br>Contributions, Property Tax<br><!-- End of picture text --> <!-- Start of picture text -->\\n- Indirect taxes<br>VAT, Customs, Excise<br><!-- End of picture text -->\\n\\nSource: Adapted from Lustig (2022)\\n\\n# **Box 3.2 A CEQ Assessment should be interpreted with several important caveats in mind**\\n\\nCEQ Assessment Caveats\\n\\n- a) **A CEQ assessment excludes important tax and public expenditure categories** , such as corporate income taxes (an important share of government revenues), and expenditure on infrastructure, defense, and other public goods because of data limitations and because there is no generally accepted methodology for assigning these benefits or burdens to any single individual or household.\\n\\n- b) **A CEQ assessment considers only the** **_redistributive_ effects of taxes and transfers,** which represent only one of many criteria to take into account when making public policy. A CEQ Assessment does not attempt to assess the sustainability of taxes and transfers from the macroeconomic, demographic, or natural capital perspectives, nor effects on future inequality, poverty, or vulnerability from fiscal policies that increase current economic growth. Results therefore should be weighed against other evidence before deciding whether a specific tax or transfer is desirable or is in need of reform.\\n\\n- c) **The fiscal incidence analysis used in CEQ assessments is an accounting framework and is point-in-time.** It is a first-order approximation and does not incorporate behavioral or general equilibrium effects. This limits the framework in important ways:\\n\\n- CEQ Assessment results cannot inform the trade-offs between public expenditure to alleviate present poverty against investments in physical and human capital (through health\\n\\n24\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"<mark>and education expenditures, for example) with large effects on future well-being through</mark> higher economic growth.\\n\\n- CEQ Assessments cannot measure the redistributive role of pensions in an intertemporal framework, which is important to accurately estimate the true redistributive effects of pensions.\\n\\n- The CEQ Assessment framework ignores behavioral responses that taxes and transfers trigger and which may imply important trade-offs in terms of efficiency, effectiveness, and sustainability of the fiscal redistribution compact.\\n\\n- d) **In-kind benefits from free government education and health services are valued at the average cost of provision.** This does not provide an estimate of the value of these services, nor of the change in purchasing power due to accessing these services. Given that in-kind services cannot be “eaten” or traded, a CEQ analysis does not typically attempt to treat in-kind benefits as additional income in cash terms.\\n\\n- e) **A CEQ Assessment is based on household survey data** , which typically underrepresent wealthier households.\\n\\n- Adapted from Lustig <u>(2022).</u>\\n\\n25\", \"datasets\": [{\"mention_name\": {\"text\": \"household survey data\", \"confidence\": 0.8328906297683716, \"start\": 1003, \"end\": 1024, \"clean_text\": \"household survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1003, \"end\": 1024, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1003, \"end\": 1024, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1003, \"end\": 1024}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1003, \"end\": 1024, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9457967281341553, \"start\": 1003, \"end\": 1024}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8328906297683716, \"start\": 1003, \"end\": 1024}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.86195969581604, \"start\": 1003, \"end\": 1024}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.86195969581604, \"start\": 1003, \"end\": 1024}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [27]}}, {\"page\": 28, \"chunk\": 28, \"input_text\": \"# **4 Methodologies for allocating taxes and transfers**\\n\\n### **This section describes in detail the allocation methodologies used to model each tax and transfer included in the analysis.**\\n\\n## **4.1 Pensions**\\n\\n**A CEQ Assessment has two scenarios for treating pensions.** The first scenario treats pensions as saved private income (in the form of contributions) spent later in the form of Pensions as Deferred Income (PDI). The second scenario treats Pensions as Government Transfer (PGT), a tool for redistribution from one segment of society to another by taxing individuals through pension contributions and transferring benefits through pension income.<sup>35</sup> Both scenarios present extremes and reality is always some combination of the two (see Annex A for comparison).\\n\\n**This analysis uses the (PGT) as the main scenario.** The reason is that pension expenditures are high and unsustainable in the West Bank and Gaza, with pension contributions funding roughly 45 percent of pension payouts, meaning that the Palestinian pension system redistributes from tax-paying individuals to pension recipients. Nonetheless, treating pensions solely as a government transfer means that those earning a comfortable pension will appear to have zero Market Income, thus overestimating the proportion of the poor (Lustig, 2022).\\n\\n**The pensions model has two components: (i) pension contributions from the current working population, and (ii) pension income received by the retired population.** For this study, the PECS survey directly identifies pension income recipients and amounts. The survey also directly identifies individuals that make pension contributions, and the amounts are inferred based on the region, public sector status and age. A key assumption of the CEQ Assessment is that employer contributions to pensions are borne by the employee in the form of lower wages—that is, the analysis allocates the total pension rate to employees, including both employee and employer’s contributions. The survey identifies roughly 140,000 pension contributors tax resident in the West Bank and Gaza - close to the numbers reported in the latest administrative data of 155,000 contributors.<sup>36</sup>\\n\\n**There are three major pension schemes in the West Bank and Gaza, and one in Israel.** In the West Bank and Gaza the unified pension scheme applies to individuals below the age of 55 as of 2016, while the Gaza civil servant pension scheme and the West Bank civil servant pension scheme variously apply to pension contributors that were above age 55 in 2016 (the former in Gaza and the latter in the West Bank). All pension contributors in the survey were below age 55 and so the unified pension scheme was applied to all directly identified pension contributors working in the West Bank and Gaza in the PECS 2015/16.<sup>37</sup> There are three components: (i) a defined benefit component (9 percent paid by employer and 7 percent\\n\\n> 35 Pensions can result in different types of redistribution. They can redistribute within the contributing population, for example by placing a floor and a cap on pension income, they can generate inter-distributional redistribution based on demographics and economic growth, they redistribute from workers who contribute but not enough to receive the benefits, and when the social security system is in a deficit, as is the case in the Palestinian Territories, they redistribute from tax-paying individuals to pension recipients.\\n\\n> 36 World Bank (2016). Public Expenditure Review of the Palestinian Authority: Towards Enhanced Public Finance Management and Improved Fiscal Sustainability. World Bank Report No: ACS18454. [http://hdl.handle.net/10986/25100]. Administrative data of pension contributors from the Public Expenditure Review (PER) for 2014 (the closest year available).\\n\\n> 37 According to the program rules, the scheme applies to anyone working in the public sector, a local authority, an International Organization (IO), the United Nations Relief and Works Agency (UNRWA), or a non-Governmental Organization (NGO).\\n\\n26\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS survey\", \"confidence\": 0.9042402505874634, \"start\": 1517, \"end\": 1528, \"clean_text\": \"PECS survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.32743245363235474, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1517, \"end\": 1528, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1517, \"end\": 1528}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1517, \"end\": 1528, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9998207688331604, \"start\": 763, \"end\": 774}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.9042402505874634, \"start\": 763, \"end\": 774}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.974339485168457, \"start\": 763, \"end\": 774}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.974339485168457, \"start\": 763, \"end\": 774}}, {\"mention_name\": {\"text\": \"administrative data\", \"confidence\": 0.5104694366455078, \"start\": 2151, \"end\": 2170, \"clean_text\": \"administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2151, \"end\": 2170, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2151, \"end\": 2170, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2151, \"end\": 2170}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2151, \"end\": 2170, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8755518794059753, \"start\": 572, \"end\": 591}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5104694366455078, \"start\": 572, \"end\": 591}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6411722898483276, \"start\": 572, \"end\": 591}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6411722898483276, \"start\": 572, \"end\": 591}}, {\"mention_name\": {\"text\": \"Public Expenditure Review\", \"confidence\": 0.535614013671875, \"start\": 3502, \"end\": 3527, \"clean_text\": \"Public Expenditure Review\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3502, \"end\": 3527, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3502, \"end\": 3527, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3502, \"end\": 3527}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3502, \"end\": 3527, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"report\", \"confidence\": 0.8052961230278015, \"start\": 385, \"end\": 410}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.535614013671875, \"start\": 385, \"end\": 410}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.740502119064331, \"start\": 385, \"end\": 410}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.740502119064331, \"start\": 385, \"end\": 410}}, {\"mention_name\": {\"text\": \"Administrative data of pension contributors\", \"confidence\": 0.6969385147094727, \"start\": 3706, \"end\": 3749, \"clean_text\": \"Administrative data of pension contributors\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3706, \"end\": 3749, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3706, \"end\": 3749, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3706, \"end\": 3749}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3706, \"end\": 3749, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9968717694282532, \"start\": 589, \"end\": 632}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6969385147094727, \"start\": 589, \"end\": 632}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5923044085502625, \"start\": 589, \"end\": 632}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5923044085502625, \"start\": 589, \"end\": 632}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [28]}}, {\"page\": 29, \"chunk\": 29, \"input_text\": \"paid by employee), (ii) a defined contribution component (3 percent paid by both the employer and employee), and (iii) a lump sum “welfare” payment of 700 NIS.<sup>38</sup> For pension contributors that are tax resident in Israel (5,750 pension contributors), a rate of 5.5 percent is applied for workers, and 12 percent for employers.\\n\\n# **4.2 Direct transfers**\\n\\n**The PECS 2016/2017 directly identifies cash transfers and near-cash**<sup>**39**</sup> **transfers from social protection.** The survey disaggregates transfers households received from the government (cash vs in-kind) but does not differentiate between types of cash transfer received. The unconditional Cash Transfer Program (CTP) was the main government-funded social protection program in 2016-17, covering households in both the West Bank and Gaza and representing about 63 percent of the Ministry of Social Development’s (MoSD) total budget.<sup>40</sup> According to MoSD’s 2017 Annual Statistical Report, CTP provided benefits to roughly 110,000 households in 2017 (72,000 in Gaza and 38,000 in the West Bank), constituting 687,000 individuals (493,000 in Gaza and 195,000 in the West Bank). The CTP identifies beneficiaries using a Proxy Means Test Formula (PMTF), which sorts households into categories relative to the extreme and national poverty lines. The exact PMT formula and threshold being used by the government to identify beneficiaries of the CTP was not available for inclusion in the model, and we approximated coverage through a combination of the location of the household and the average monthly transfer amounts to match the administrative data. For the social protection in-kind transfers, all households reported as receiving this transfer from the government have been directly identified and included.\\n\\n# **4.3 Direct taxes**\\n\\n## _4.3.1 Personal income tax (PIT)_\\n\\n**As with the pension contributions, rates for the West Bank and Gaza are applied to those working in the West Bank and Gaza and subject to tax by the PA (in Gaza this includes only PA employees), and Israeli income tax rates are applied to the West Bank and Gaza population working in East Jerusalem, Israel, or the settlements.** Personal Income Taxes paid are simulated by applying statutory rates to formal, tax-liable workers’ self-reported income). Formal workers are restricted to workers of working age, not an unregistered refugee, and that report making pension contributions on their behalf. They are treated as tax liable if they are tax resident in the West Bank and Gaza or Israel (i.e., not working abroad or for a foreign government) and exempt if they have a disability that has a difficulty level of 50 percent of more. The simulation is effective – 109,200 taxpayers are identified in the survey compared to 112,500 in the administrative data.\\n\\n**Taxable income:** For employees, income is assumed to be net of tax while for employers and selfemployed workers it is assumed to be gross of tax. We observe income in the survey from wages and\\n\\n> 38 The modeling proceeds under the assumption that the payment is annual.\\n\\n> 39 Note that while these are described in the PECS 2016/2017 as in-kind transfers, in this assessment they are referred to as near-cash transfers. This is in keeping with CEQ terminology which distinguishes between in-kind transfers such as public health and education services - which are valued at an average cost of provision - and near-cash transfers, such as school feeding or fertilizer - which are valued according to the change in a households’ purchasing power due to the receipt of a good or service.\\n\\n> 40 The total amount transferred to households through the CTP in 2017 was 518.4 billion NIS in 2017.\\n\\n27\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS 2016/2017\", \"confidence\": 0.4818365275859833, \"start\": 371, \"end\": 385, \"clean_text\": \"PECS 2016/2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 371, \"end\": 385, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 371, \"end\": 385, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 371, \"end\": 385}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 371, \"end\": 385, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.6657867431640625, \"start\": 371, \"end\": 385}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4818365275859833, \"start\": 371, \"end\": 385}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9166944026947021, \"start\": 371, \"end\": 385}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9166944026947021, \"start\": 371, \"end\": 385}}, {\"mention_name\": {\"text\": \"administrative data\", \"confidence\": 0.7136214971542358, \"start\": 1618, \"end\": 1637, \"clean_text\": \"administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1618, \"end\": 1637, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1618, \"end\": 1637, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1618, \"end\": 1637}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1618, \"end\": 1637, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9292411208152771, \"start\": 984, \"end\": 1003}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.30902257561683655, \"start\": 984, \"end\": 1003}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.7174077033996582, \"start\": 984, \"end\": 1003}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7174077033996582, \"start\": 984, \"end\": 1003}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [29]}}, {\"page\": 30, \"chunk\": 30, \"input_text\": \"employer income, self-employment income, retirement income, rental income, land, transport, and stocks.\\n\\n**A residential deduction is allocated to all West Bank and Gaza resident taxpayers, and additional deductions are allocated to those eligible due to expenditures on donations, education, housing, fixed transportation, and pension contributions, in that order.** We do not have information on the order in which these deductions are applied by law, and so they are allocated from largest to smallest (in terms of the average size amongst those that are eligible for the deduction, prior to scaling). Given other administrative and survey data gaps discussed in Annex B deductions cannot be precisely allocated and scaling is applied to compensate for these data gaps. This prevents a substantial over-allocation of these deductions, which would result in an under-allocation of PIT in the survey. Table 4-1 shows the difference in the allocated deduction amount in the survey prior to, and post, scaling, and compared with the administrative numbers. See Annex B for a discussion of the data gaps faced in allocating the deductions.\\n\\n**Table 4-1: Scaling is applied to deductions to prevent an under-allocation of PIT in the survey**\\n\\n|**Deduction**|**Survey**<br>**(pre-scaling)**|**Admin**|**Survey**<br>**(post-scaling)**|\\n|---|---|---|---|\\n|Residential(billion NIS)|7.1|1.1|1|\\n|Donations(million NIS)|46.4|n/a|46.4|\\n|Education(million NIS)|219|6.7|6.02|\\n|Housing (thousand NIS)|12,411.3|372|346.6|\\n|Fixed transport to work(thousand NIS)|79.8|329.3*|306.8|\\n|Pension contributions**|368860.5|34|31.8|\\n\\nSource: World Bank and Ministry of Finance staff.\\n\\n* Admin number from 2017\\n\\n** Based on the employee contributions component only.\\n\\n**Israeli tax policy allows deductions for disabilities, 52 percent of contributions to national health insurance, rental income, disabilities, and two tax credits: a tax credit**<sup>**41**</sup> **that varies by gender, and a tax credit of 35 percent of total contributions to pensions.** The size of the disability exemption varies based on duration; however the PECS 2016/17 does not indicate duration and so the full exemption of 72,720 NIS is applied to all individuals that report having a complete disability. A tax credit of 2.25 credit points (NIS 483 per month) was allocated to men and 2.75 credit points (NIS 591 per month) to women (a tax credit was worth NIS 215 per month in 2017). An additional 0.25 credit points were applied for travel to place of work for all men, and an additional 0.75 credit points for women. Due to data limitations, the following additional credits are not applied: credits for individuals who are supporting a spouse, have a working spouse, are a single parent supporting a child, have completed a degree within the last 2-3 years (depending on the degree), have applied for credit for individuals supporting a parent within an\\n\\n> 41 Note that a tax credit differs from an exemption in that it is subtracted from an individual’s PIT liability, rather than from their taxable income.\\n\\n28\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS 2016/17\", \"confidence\": 0.6297875642776489, \"start\": 2108, \"end\": 2120, \"clean_text\": \"PECS 2016/17\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2108, \"end\": 2120, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2108, \"end\": 2120, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2108, \"end\": 2120}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2108, \"end\": 2120, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.3632495701313019, \"start\": 619, \"end\": 631}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6297875642776489, \"start\": 619, \"end\": 631}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6202622056007385, \"start\": 619, \"end\": 631}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6202622056007385, \"start\": 619, \"end\": 631}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [30]}}, {\"page\": 31, \"chunk\": 31, \"input_text\": \"institution, or are caring for incapacitated persons. The details of the allocation of these deductions are reported in Annex B.\\n\\n## **_4.3.2. Property tax_**\\n\\n**The PECS 2016/17 includes self-reported annual rental property value used to estimate property tax, with simulation including the following steps:**\\n\\n- (i) Used administrative data to identify the cities from which property tax was collected.\\n\\n- (ii) Used a city identifier from the location module to identify the core set of municipalities in which property taxes are collected in the West Bank and Gaza.\\n\\n- (iii) Identified owners within the dataset using self-reported status (given that property tax is paid by owners only).\\n\\n- (iv) Used a self-reported estimated annual rental value variable to estimate the tax on buildings.\\n\\n- (v) For tax collection on land, we used administrative data on the valuation of land by applying an average per governorate.<sup>42</sup>\\n\\n- (vi) Estimated property tax on land is by multiplying the tax rate by the area of land and average land valuation (bifurcated by cities).\\n\\n- (vii) Scaled-down the survey-estimated property tax by applying a scaling factor such that the effective property tax rate in the survey (that is, the size of the property tax relative to households’ consumption) matches administrative data.\\n\\n# **4.4. Indirect taxes**\\n\\n**Indirect taxes in the model include customs duties, VAT and excises, which together represented about 92 percent of total tax collections in the West Bank and Gaza in 2017.** These taxes were simulated by applying statutory tax rates to self-reported household expenditure on each product in the PECS 2016/2017.\\n\\n## **4.4.1. Customs**\\n\\n**Simulation of customs duties followed these steps:**\\n\\n- (i) Mapped customs duty rates to PECS 2016/17 consumption items (average duty for those goods mapped to different HS codes).\\n\\n- (ii) Excluded non-taxable items from the PECS 2016/17 consumption dataset (self-production, gifts). Netted down to observed households’ expenditure.<sup>43</sup>\\n\\n- (iii) Applied statutory customs duty rates (differentiated by region) to all taxable products.\\n\\n**The PECS 2016/2017 does not have a variable to identify imported products, so the study assumes the “Law of One Price”—that is, the prices of domestic goods converge to prices of similar imported goods.** We also account for how tax systems vary by region by assuming that residents in the West Bank face Palestinian customs duty rates while residents in East Jerusalem face Israeli customs duty rates, and that Gaza residents pay no customs duties.\\n\\n## **4.4.2. VAT**\\n\\n## **VAT simulation followed these steps:**\\n\\n> 42 The property valuation data provided by the Government for the year 2017 had significant variation across cities for the valuation amounts. As a result, for the analysis, an average property value per governorate is applied.\\n\\n> 43 Netting-down for customs duties means that self-reported household expenditure in the PECS needs to be divided by (1+customs duty rate)*(1+VAT rate). This allows the analysis to recover the households’ pre-fiscal expenditure before any taxes.\\n\\n29\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS 2016/17\", \"confidence\": 0.49471643567085266, \"start\": 166, \"end\": 178, \"clean_text\": \"PECS 2016/17\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 166, \"end\": 178, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 166, \"end\": 178, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 166, \"end\": 178}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 166, \"end\": 178, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.937188982963562, \"start\": 166, \"end\": 178}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3022560477256775, \"start\": 166, \"end\": 178}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6354859471321106, \"start\": 166, \"end\": 178}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6354859471321106, \"start\": 166, \"end\": 178}}, {\"mention_name\": {\"text\": \"administrative data\", \"confidence\": 0.85777747631073, \"start\": 323, \"end\": 342, \"clean_text\": \"administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 323, \"end\": 342, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 323, \"end\": 342, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 323, \"end\": 342}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 323, \"end\": 342, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.993536651134491, \"start\": 323, \"end\": 342}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.85777747631073, \"start\": 323, \"end\": 342}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5783967971801758, \"start\": 323, \"end\": 342}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5783967971801758, \"start\": 323, \"end\": 342}}, {\"mention_name\": {\"text\": \"administrative data on the valuation of land\", \"confidence\": 0.47922584414482117, \"start\": 837, \"end\": 881, \"clean_text\": \"administrative data on the valuation of land\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 837, \"end\": 881, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 837, \"end\": 881, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 837, \"end\": 881}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 837, \"end\": 881, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9888525009155273, \"start\": 837, \"end\": 881}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.47922584414482117, \"start\": 837, \"end\": 881}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5769621133804321, \"start\": 837, \"end\": 881}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5769621133804321, \"start\": 837, \"end\": 881}}, {\"mention_name\": {\"text\": \"PECS 2016/2017\", \"confidence\": 0.7795620560646057, \"start\": 1647, \"end\": 1661, \"clean_text\": \"PECS 2016/2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.4046705961227417, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1647, \"end\": 1661, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"2016/17\", \"confidence\": 0.37342968583106995, \"start\": 1783, \"end\": 1790}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1647, \"end\": 1661, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.48682284355163574, \"start\": 174, \"end\": 188}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7795620560646057, \"start\": 174, \"end\": 188}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6111668944358826, \"start\": 174, \"end\": 188}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6111668944358826, \"start\": 174, \"end\": 188}}, {\"mention_name\": {\"text\": \"PECS 2016/17 consumption dataset\", \"confidence\": 0.6715909242630005, \"start\": 1914, \"end\": 1946, \"clean_text\": \"PECS 2016/17 consumption dataset\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1914, \"end\": 1946, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1914, \"end\": 1946, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1914, \"end\": 1946}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1914, \"end\": 1946, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.5873233079910278, \"start\": 441, \"end\": 473}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6715909242630005, \"start\": 441, \"end\": 473}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6081895232200623, \"start\": 441, \"end\": 473}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6081895232200623, \"start\": 441, \"end\": 473}}, {\"mention_name\": {\"text\": \"property valuation data\", \"confidence\": 0.5822007656097412, \"start\": 2660, \"end\": 2683, \"clean_text\": \"property valuation data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2660, \"end\": 2683, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2660, \"end\": 2683, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"2017\", \"confidence\": 0.3900883197784424, \"start\": 2150, \"end\": 2154}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2660, \"end\": 2683, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.26386967301368713, \"start\": 511, \"end\": 534}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5822007656097412, \"start\": 511, \"end\": 534}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.3959006369113922, \"start\": 511, \"end\": 534}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.3959006369113922, \"start\": 511, \"end\": 534}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [31]}}, {\"page\": 32, \"chunk\": 32, \"input_text\": \"- (i) Mapped VAT rates to PECS 2016/17 consumption items.\\n\\n- (ii) Prepared the Augmented IO Matrix<sup>44</sup> for the West Bank and mapped the Augmented IO sectors to VAT rates and to consumption items (COICOP codes) in the PECS 2016/17.\\n\\n- (iii) Calculated indirect effects based on the Augmented IO Matrix of the West Bank and the CostPush Model (Inchauste and Jellema 2018). The model applies the price shock of VAT rates to taxable IO sectors; output sectors that have used inputs that paid VAT will carry indirect effects.\\n\\n- (iv) Excluded non-taxable categories such as self-production and gifts from the PECS 2016/17 consumption dataset. Netted down self-reported households’ expenditure.<sup>45</sup>\\n\\n- (v) Calculated direct effects based on VAT statutory rates for the formal purchase share of all taxable items in the PECS 2016/17. The estimation of households’ formal purchase share (by decile) was imputed based on average consumption informal shares (by decile) estimated in similar countries available in Bachas et al. 2020.<sup>46</sup>\\n\\n- (vi) Applied indirect effects<sup>47</sup> (calculated for the West Bank) for exempt formal items and for the informal purchase share<sup>48</sup> in the PECS 2016/17.\\n\\n- (vii) Scaled-up the VAT simulation in the survey to match the VAT effective rate in the administrative accounts (total VAT relative to consumption).\\n\\n**The VAT model accounts for the different tax systems applied by region:** We assume residents in the West Bank face Palestinian VAT rates, that residents in East Jerusalem face Israeli VAT rates, and that Gaza residents do not pay VAT.\\n\\n**We apply the latest methodologies by accounting for informality in households’ purchases based on Bachas et al. (2020), and including VAT indirect effects (e.g., increases in prices of goods and services due to unclaimed VAT paid on production inputs)**<sup>**49**</sup> **following Inchauste et al. (forthcoming).** Accounting for consumption informality in the West Bank and Gaza is important because: (i) the size of the informal economy is large, estimated above 50 percent of GDP (IMF 2022)<sup>50</sup> ; and (ii) recent cross-country evidence shows that households’ consumption informality tends to be higher among poorer households, hence\\n\\n> 44 The IO Matrix was augmented for sectors that have a mix of exempt and non-exempt products, based on a 50-50 share. These sectors were: agricultural products, water, accommodation services, transport, financial services, business services, education, and health.\\n\\n> 45 The netting-down in the case of VAT consists in dividing self-reported gross expenditure of each product by (1+ VAT statutory rate) that corresponds to each product.\\n\\n> 46 Consumption informality based on average from 10 comparator countries with similar WB income classifications available in Bachas et al. (2020): Bolivia; Cameroon; Comoros; Congo Rep.; Eswatini; Morocco; Papua New Guinea; Sao Tome; Senegal; Tunisia.\\n\\n> 47 VAT indirect effects or cascading effects refer to the increase in final prices that comes from VAT paid over inputs. It is assumed that sellers that cannot reclaim input VAT credit (for example, those selling exempt or informal items) will pass-through their unrecovered VAT in the form of higher final prices to be paid by consumers. Source: CEQ Handbook, Chapter 7 (Jellema & Inchauste, 2018).\\n\\n> 48 To estimate the informal purchase share by decile in the West Bank and Gaza, the estimations of “self-consumption” (nonmarket informal consumption available in the PECS 2016/17) was deducted from “total informal consumption” (average from 10 similar countries in Bachas et al. 2020).\\n\\n> 49 In an ideal VAT system (without exemptions and informality), there would not be VAT indirect effects or cascading effects since all producers and sellers along the distribution chain are able to receive a VAT credit for the VAT they paid on their inputs. However, in the presence of exemptions and informality, the VAT input credit mechanism is broken. In the Cost-Push Model used in the current study it is assumed that sellers of exempt and informal goods cannot claim their VAT input credit, hence they passthrough their unrecovered input VAT through higher final prices paid by consumers.\\n\\n> 50 IMF (2022). West Bank and Gaza: Selected Issues.\\n\\n30\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS 2016/17 consumption dataset\", \"confidence\": 0.6479360461235046, \"start\": 613, \"end\": 645, \"clean_text\": \"PECS 2016/17 consumption dataset\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 613, \"end\": 645, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 613, \"end\": 645, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 613, \"end\": 645}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 613, \"end\": 645, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.6837378740310669, \"start\": 613, \"end\": 645}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6479360461235046, \"start\": 613, \"end\": 645}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5791117548942566, \"start\": 613, \"end\": 645}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5791117548942566, \"start\": 613, \"end\": 645}}, {\"mention_name\": {\"text\": \"PECS 2016/17\", \"confidence\": 0.5467906594276428, \"start\": 831, \"end\": 843, \"clean_text\": \"PECS 2016/17\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 831, \"end\": 843, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 831, \"end\": 843, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 831, \"end\": 843}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 831, \"end\": 843, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9794119000434875, \"start\": 831, \"end\": 843}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5467906594276428, \"start\": 831, \"end\": 843}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7430120706558228, \"start\": 831, \"end\": 843}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7430120706558228, \"start\": 831, \"end\": 843}}, {\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.5615620017051697, \"start\": 3535, \"end\": 3539, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.38349437713623047, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3535, \"end\": 3539, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3535, \"end\": 3539}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3535, \"end\": 3539, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9604102373123169, \"start\": 681, \"end\": 685}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5615620017051697, \"start\": 681, \"end\": 685}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7279625535011292, \"start\": 681, \"end\": 685}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7279625535011292, \"start\": 681, \"end\": 685}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [32]}}, {\"page\": 33, \"chunk\": 33, \"input_text\": \"increasing the progressivity of the VAT (Bachas et al., 2020). While the PECS 2016/17 does not have a place of purchase variable to proxy households’ consumption informality, this study imputed average consumption informality by decile based on estimates from similar countries available in Bachas et al. (2020). We calculated the indirect effects of the VAT for the West Bank region using the 2017 Input-Output table and the Cost-Push Model (Inchauste and Jellema 2018, Inchauste et al. forthcoming).<sup>51</sup>\\n\\n## **4.4.3. Excises**\\n\\n**We model the fuel excises for gasoline, diesel, and kerosene** . Consistent with previous indirect taxes’ assumptions. We only simulate fuel excises in the West Bank, whereas we assume individuals in Gaza pay zero excises. The model of excises does not account for households’ consumption informality given that fuel products are typically not prone to tax evasion.\\n\\n## **The simulation of fuel excises followed these steps:**\\n\\n- (i) Mapped excise rates to relevant fuel products in the PECS 2016/17. Given data limitations, the model uses 2017 fuel excise rates from the Israeli Tax Authority<sup>52</sup> ; this is a plausible assumption since the excise rates from the Palestinian Territories and Israel are similar to avoid arbitrage between both countries.\\n\\n- (ii) Prepared the Augmented IO Matrix for the West Bank<sup>53</sup> and mapped IO sectors (fuel products) to relevant excise rates. Also, all IO sectors were mapped to the associated final goods (COICOP codes) from the PECS.\\n\\n- (iii) Excluded non-taxable categories such as self-production and gifts from the PECS 2016/17 consumption dataset.\\n\\n- (iv) Calculated indirect effects of fuel excises based on the Augmented IO Matrix for the West Bank and the Cost-Push Model (Inchauste and Jellema, 2018). The model applies the price shock of fuel excises (differentiated for gasoline, diesel, and kerosene) to the petroleum input sectors; output sectors that have used petroleum products as an input will carry indirect effects from fuel excises.\\n\\n- (v) Calculated direct effects based on (estimated) quantities of gasoline, diesel and kerosene consumed by households in the West Bank and the fuel excise rates for each product. Quantities were estimated based on households’ self-reported expenditure in gasoline, diesel and kerosene divided by each products’ prices in 2017 (differentiate by region).<sup>54</sup> As explained above, Israeli fuel excise rates (2017) were applied for gasoline, diesel and kerosene quantities.\\n\\n- (vi) Applied indirect effects of fuel excises in the PECS 2016/17 to all households’ products that have used petroleum as a production input.\\n\\n# **4.5. Indirect subsidies**\\n\\n## **4.5.1. Domestic electricity**\\n\\n**Estimation of the electricity subsidies to the domestic consumers is based on the applicable tariff structure (Table 4-2) and estimated cost of supply.** There was no data available on the cost of electricity\\n\\n> 51 See Inchauste and Jellema (2018), for an explanation of the methodology on indirect effects and the Cost-Push Model.\\n\\n> 52 The 2017 (Israeli) excise rates used in the model were: gasoline (2.89 NIS/L), diesel (3.02 NIS/L) and kerosene (2.89 NIS/L). 53 The IO sector of “Petroleum products” was disaggregated between “gasoline”, “diesel” and “other oils” based on consumption shares calculated using the PECS 2016/17.\\n\\n> 54 Fuel prices (2017) are differentiated for residents between East Jerusalem and West Bank.\\n\\n31\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS 2016/17\", \"confidence\": 0.8021583557128906, \"start\": 73, \"end\": 85, \"clean_text\": \"PECS 2016/17\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 73, \"end\": 85, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 73, \"end\": 85, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 73, \"end\": 85}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 73, \"end\": 85, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9700726270675659, \"start\": 73, \"end\": 85}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5364424586296082, \"start\": 73, \"end\": 85}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.8207919001579285, \"start\": 73, \"end\": 85}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8207919001579285, \"start\": 73, \"end\": 85}}, {\"mention_name\": {\"text\": \"2017 Input-Output table\", \"confidence\": 0.5727798938751221, \"start\": 394, \"end\": 417, \"clean_text\": \"2017 Input-Output table\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 394, \"end\": 417, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 394, \"end\": 417, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 394, \"end\": 417}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 394, \"end\": 417, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8922945261001587, \"start\": 394, \"end\": 417}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5727798938751221, \"start\": 394, \"end\": 417}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9263402223587036, \"start\": 394, \"end\": 417}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9263402223587036, \"start\": 394, \"end\": 417}}, {\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.4453890025615692, \"start\": 1526, \"end\": 1530, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1526, \"end\": 1530, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1526, \"end\": 1530, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1526, \"end\": 1530}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1526, \"end\": 1530, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.9133032560348511, \"start\": 820, \"end\": 824}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4453890025615692, \"start\": 820, \"end\": 824}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7684682011604309, \"start\": 820, \"end\": 824}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7684682011604309, \"start\": 820, \"end\": 824}}, {\"mention_name\": {\"text\": \"PECS 2016/17 consumption dataset\", \"confidence\": 0.770155668258667, \"start\": 1616, \"end\": 1648, \"clean_text\": \"PECS 2016/17 consumption dataset\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.5442022085189819, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1616, \"end\": 1648, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1616, \"end\": 1648}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1616, \"end\": 1648, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.6731627583503723, \"start\": 910, \"end\": 942}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.770155668258667, \"start\": 910, \"end\": 942}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.595515251159668, \"start\": 910, \"end\": 942}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.595515251159668, \"start\": 910, \"end\": 942}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [33]}}, {\"page\": 34, \"chunk\": 34, \"input_text\": \"supply inclusive of the “distributional cost”. We estimated cost of supply of electricity to domestic consumers by taking the lower bound of the price range DISCOs paid while importing/purchasing the electricity (NIS 0.53) and adding a 20 percent margin to cover distributional costs, surcharges, and recoveries.<sup>55</sup> This results in an estimated average cost of supply of 0.636 NIS per kWh.<sup>56</sup> We then calculated the subsidy for each consumption slab by taking the difference between the cost of supply (0.636 NIS per kWh) and the tariff in that consumption slab. Table 4-2 describes the subsidies.\\n\\n**Table 4-2 The estimated electricity subsidy per kWh ranges from 0.178 to -0.0371 in the West Bank**\\n\\n**and Gaza**\\n\\nElectricity subsidy <u>per kWh by consumption slab, West Bank and Gaza</u>\\n\\n||**Quantity**<br>|**Subsidy**<br>|\\n|---|---|---|\\n|**Slab**|(kWh per<br>month)|(NIS per<br>kWh)|\\n|1|1-160|0.18|\\n|2|161-250|0.14|\\n|3|251-400|0.07|\\n|4|401-600|0.03|\\n|5|More than 600|-0.04|\\n\\nSource: Authors’ estimates based on Prime Minister’s Office (2017)\\n\\n# **We performed the following steps to simulate direct effects from the domestic electricity subsidy:**\\n\\n- (i) Identified the amount of households’ expenditure on electricity based on the self-reported “electricity charges” in the PECS.\\n\\n- (ii) Estimated the quantity of electricity consumed by dividing total expenditure on electricity by the tariff (which varies by consumption slab).\\n\\n- (iii) Calculated the direct effect as the subsidy amount per kWh (between 0.178 and -0.0371) times the (estimated) electricity quantity (kWh) used domestically consumed per the PECS.<sup>57</sup>\\n\\n- The analysis focuses on direct effects and assumes no indirect effects since the domestic electricity subsidy is only provided to households and not to electricity producers/distributors.\\n\\n# **4.5.2. Fuel subsidies**\\n\\n# **Simulation of gasoline and diesel fuel subsidies is based on self-reported PECS households’ expenditure**\\n\\n**data and subsidy calculated from administrative data (0.0686 and 0.1962 NIS per liter for gasoline and diesel, respectively).**<sup>58</sup> The simulation calculates the direct effects of fuel subsidies for residents in the West Bank and Gaza, and indirect effects for residents in the West Bank only. The direct effects of subsidies on gasoline and diesel were calculated by multiplying the subsidy per liter of fuel by the\\n\\n> 55 Estimation of distribution margin (inclusive of all surcharges and costs) of 20 percent in the electricity sector is provided by the World Bank Energy Team based on regional average distribution margins.\\n\\n> 56 Cost of supply after 20 percent margin calculated as NIS 0.53/ KwH * 1.2.\\n\\n> 57 Note that due to the data limitation, the subsidy estimation is based on assumption of “volumetric consumption” instead of Increasing block tariff (IBT). Furthermore, the estimation only includes “Tariff Differential subsidy” and not” cross subsidy” within the tariff structure.\\n\\n> 58 The per liter subsidy for gasoline and diesel was calculated by taking the difference between the retail price and GDoP price for the year of analysis.\\n\\n32\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.31284764409065247, \"start\": 1300, \"end\": 1304, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.6145274639129639, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1300, \"end\": 1304, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1300, \"end\": 1304}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1300, \"end\": 1304, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.6136753559112549, \"start\": 240, \"end\": 244}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.31284764409065247, \"start\": 240, \"end\": 244}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7234784364700317, \"start\": 240, \"end\": 244}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7234784364700317, \"start\": 240, \"end\": 244}}, {\"mention_name\": {\"text\": \"administrative data\", \"confidence\": 0.8639732003211975, \"start\": 2022, \"end\": 2041, \"clean_text\": \"administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2022, \"end\": 2041, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2022, \"end\": 2041, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2022, \"end\": 2041}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2022, \"end\": 2041, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9996757507324219, \"start\": 212, \"end\": 231}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8639732003211975, \"start\": 212, \"end\": 231}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4486117362976074, \"start\": 212, \"end\": 231}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4486117362976074, \"start\": 212, \"end\": 231}}, {\"mention_name\": {\"text\": \"regional average distribution margins\", \"confidence\": 0.33216047286987305, \"start\": 2586, \"end\": 2623, \"clean_text\": \"regional average distribution margins\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2586, \"end\": 2623, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2586, \"end\": 2623, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2586, \"end\": 2623}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2586, \"end\": 2623, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9985029697418213, \"start\": 776, \"end\": 813}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.33216047286987305, \"start\": 776, \"end\": 813}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6566819548606873, \"start\": 776, \"end\": 813}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6566819548606873, \"start\": 776, \"end\": 813}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [34]}}, {\"page\": 35, \"chunk\": 35, \"input_text\": \"estimated quantities of gasoline and diesel in the PECS 2016/2017. The quantity of household consumption of gasoline and diesel was estimated by dividing households’ self-reported expenditure in the survey by the 2017 final consumer price. The indirect effects of gasoline and diesel subsidies (used as inputs in the production process) were calculated based on the West Bank IO table and the “Cost-Push Model\\\" (Inchauste and Jellema, 2018). We applied the indirect effects from gasoline and diesel subsidies to all applicable household residents’ expenditures in the West Bank<sup>59</sup> .\\n\\n- **4.6 In-kind benefits**\\n\\n**The assessment includes in-kind benefits for public education and public health.** Both allocation methods impute “average benefits” to users and individuals eligible for these public services, respectively, based on information from the PECS 2016/17 and government administrative data.\\n\\n# **Education**\\n\\n**This study follows the “average-cost approach” for users of public education services to allocate in-kind public education benefits.** First, the in-kind benefits for public education services (per student) were calculated at the national level based on administrative data of (executed) public expenditure on education divided by the number of total students enrolled in public schools from administrative accounts; different benefits were calculated for the basic and secondary levels of education. Second, the study infers potential beneficiaries from public in-kind education benefits in the PECS survey based on the following reported variables: (i) individuals currently enrolled in public schools; and (ii) self-reported information on “highest grade of education achieved” plus one year.<sup>60</sup> The admin numbers indicated that for 2017 there were 585,821 students enrolled in (public) basic education and 214,663 students enrolled in (public) secondary education. In the case of (public) basic education, the inference approach from the PECS resulted in a larger number of students than the administrative target<sup>61</sup> ; hence, a random simulation within the eligible pool was performed to sub-select the number of (public) basic students in the PECS that matched the administrative number.<sup>62</sup> Third for the students identified in the PECS as those enrolled in (public) basic and secondary education (calculated in the second step), the study imputes the average in-kind benefits differentiated for each education level (calculated in the first step).\\n\\n# **Health**\\n\\n**The analysis uses the “insurance approach” to allocate in-kind public health benefits.** First, the in-kind benefits from public health services (per individual) were calculated based on executed health expenditure data from administrative accounts (primary care versus hospitals) divided by the number of individuals insured under government insurance self-reported in the PECS. Second, the analysis tries to\\n\\n> 59 All households’ expenditure items that have used gasoline or diesel as a production input will carry indirect effects from fuel subsidies.\\n\\n> 60 The PECS does not have information on the current grade of education. We estimated this variable based on the highest grade of education achieved, plus one year for individuals currently enrolled in public schools. It is only possible to estimate the current grade for primary and secondary schooling.\\n\\n> 61 Originally, the inference approach from the PECS resulted in 716,822 students enrolled in (public) basic education in the survey, 122 percent times higher than the admin target. After the random allocation, the number of (public) basic education students simulated in the PECS came down to 580,558 (99 percent of the admin target).\\n\\n> 62 The random allocation approach was not needed for secondary education since the inference approach from the PECS resulted in 168,783 students enrolled in (public) secondary education according to the survey, which was 79 percent of the admin target. Hence, no random allocation was needed to sub-select students since there was no overestimation stemming from the PECS.\\n\\n33\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS 2016/2017\", \"confidence\": 0.7614545822143555, \"start\": 51, \"end\": 65, \"clean_text\": \"PECS 2016/2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 51, \"end\": 65, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 51, \"end\": 65, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"2016/2017\", \"confidence\": 0.5084565877914429, \"start\": 56, \"end\": 65}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 51, \"end\": 65, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9993738532066345, \"start\": 51, \"end\": 65}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7614545822143555, \"start\": 51, \"end\": 65}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5235509276390076, \"start\": 51, \"end\": 65}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5235509276390076, \"start\": 51, \"end\": 65}}, {\"mention_name\": {\"text\": \"West Bank IO table\", \"confidence\": 0.7476013898849487, \"start\": 366, \"end\": 384, \"clean_text\": \"West Bank IO table\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 366, \"end\": 384, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 366, \"end\": 384, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 366, \"end\": 384}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 366, \"end\": 384, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.5933272838592529, \"start\": 366, \"end\": 384}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7476013898849487, \"start\": 366, \"end\": 384}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.80515056848526, \"start\": 366, \"end\": 384}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.80515056848526, \"start\": 366, \"end\": 384}}, {\"mention_name\": {\"text\": \"PECS 2016/17\", \"confidence\": 0.7246806025505066, \"start\": 862, \"end\": 874, \"clean_text\": \"PECS 2016/17\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 862, \"end\": 874, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 862, \"end\": 874, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 862, \"end\": 874}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 862, \"end\": 874, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.8312130570411682, \"start\": 862, \"end\": 874}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7246806025505066, \"start\": 862, \"end\": 874}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8123822808265686, \"start\": 862, \"end\": 874}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8123822808265686, \"start\": 862, \"end\": 874}}, {\"mention_name\": {\"text\": \"government administrative data\", \"confidence\": 0.6856887936592102, \"start\": 879, \"end\": 909, \"clean_text\": \"government administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 879, \"end\": 909, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 879, \"end\": 909, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 879, \"end\": 909}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 879, \"end\": 909, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.8312130570411682, \"start\": 879, \"end\": 909}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6856887936592102, \"start\": 879, \"end\": 909}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8123822808265686, \"start\": 879, \"end\": 909}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8123822808265686, \"start\": 879, \"end\": 909}}, {\"mention_name\": {\"text\": \"administrative data\", \"confidence\": 0.7316536903381348, \"start\": 1185, \"end\": 1204, \"clean_text\": \"administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1185, \"end\": 1204, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1185, \"end\": 1204, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1185, \"end\": 1204}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1185, \"end\": 1204, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9794619679450989, \"start\": 390, \"end\": 409}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7316536903381348, \"start\": 390, \"end\": 409}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7622935771942139, \"start\": 390, \"end\": 409}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7622935771942139, \"start\": 390, \"end\": 409}}, {\"mention_name\": {\"text\": \"PECS survey\", \"confidence\": 0.7465879321098328, \"start\": 1527, \"end\": 1538, \"clean_text\": \"PECS survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.359173446893692, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1527, \"end\": 1538, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1527, \"end\": 1538}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1527, \"end\": 1538, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9945105314254761, \"start\": 732, \"end\": 743}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7465879321098328, \"start\": 732, \"end\": 743}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6669612526893616, \"start\": 732, \"end\": 743}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6669612526893616, \"start\": 732, \"end\": 743}}, {\"mention_name\": {\"text\": \"executed health expenditure data from administrative accounts\", \"confidence\": 0.5392659306526184, \"start\": 2719, \"end\": 2780, \"clean_text\": \"executed health expenditure data from administrative accounts\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2719, \"end\": 2780, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2719, \"end\": 2780, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2719, \"end\": 2780}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2719, \"end\": 2780, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.8588375449180603, \"start\": 330, \"end\": 391}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.5392659306526184, \"start\": 330, \"end\": 391}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7667125463485718, \"start\": 330, \"end\": 391}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7667125463485718, \"start\": 330, \"end\": 391}}, {\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.5180257558822632, \"start\": 2906, \"end\": 2910, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.40609991550445557, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2906, \"end\": 2910, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2906, \"end\": 2910}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2906, \"end\": 2910, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.8588375449180603, \"start\": 517, \"end\": 521}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5180257558822632, \"start\": 517, \"end\": 521}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7667125463485718, \"start\": 517, \"end\": 521}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7667125463485718, \"start\": 517, \"end\": 521}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [35]}}, {\"page\": 36, \"chunk\": 36, \"input_text\": \"infer potential beneficiaries of public health benefits in the PECS. Since the PECS does not have information on access to health services, the analysis focuses on the eligible population of public health services—that is, those individuals that self-report having “government health insurance” in the PECS. Third, we imputed the average health benefit (primary care versus hospital services) equally to all individuals benefitting from government health insurance.\\n\\n34\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [36]}}, {\"page\": 37, \"chunk\": 37, \"input_text\": \"# **5 Headline results**\\n\\n## **5.1 Poverty and inequality**\\n\\n**This section assesses the joint distributional impacts of the main taxes and transfers modeled in the West Bank and Gaza.** Results include changes in national poverty (headcount and gap) and inequality (measured by the Gini coefficient).\\n\\n**To assess the distributional impacts of taxes and transfers on poverty, we calculated the national poverty headcount and the national poverty gap at the different CEQ income concepts.** Since the CEQ income concepts starts from Disposable Income equated to Consumption, all CEQ income concepts are consumption-based (in real, per capita terms), hence consistent with the official poverty line. The results shown in Figure 5-1 show that the combination of taxes and transfers modeled increased the national poverty headcount by 8.4 percentage points from Market to Consumable Income (from 30.6 to 39.0 percent), while the national poverty gap increased by 1.0 percentage point from Market Income to Consumable Income (from 9.8 to 10.8 percent). Most of the increase in the national poverty headcount and poverty gap is the result of the change between Disposable Income and Consumable Income; this suggests that indirect taxes drive the poverty-increase.\\n\\n**When pension is measured as a deferred income, the total poverty increase due to the portion of the fiscal system measured here** **_decreases_ to 7.6 percentage points** (Figure A.1). This suggests that the poverty-increasing impact of the pension contributions in the PGT scenario outweighs the povertyreducing impact of the pension income. This is likely because the numbers of individuals in deciles 1 and 2 that receive pension income are lower than the numbers that pay pension contributions, even though the total amount of pension income received in deciles 1 and 2 is higher than the total amount of contributions paid. See Annex A for a more detailed discussion of this result.\\n\\n35\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [37]}}, {\"page\": 38, \"chunk\": 38, \"input_text\": \"<!-- Start of picture text --> Poverty by income concept (PGT scenario)<br>40 39.0<br>33.4<br>=a)DQ 30.6 29.3<br>£ 30<br>°<br>a<br>oa<br>D<br>8<br>=<br>o<br>2 320<br>o<br>2<br>><br>o4 11.63<br>& 98 10.8<br>10 79<br>0<br>Market Net market Disposable Consumable<br>G9 Headcount HS Gap<br>Slslotes: Authors’ estimates based on PECS 2016/2017; LFS 2017.<br>1] The official national poverty headcount matches poverty at Disposable Income.<br>2] We do not calculate poverty at Final income because we value the health and education transfers at the average cost of provision.<br><!-- End of picture text -->\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.4472680389881134, \"start\": 324, \"end\": 328, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 324, \"end\": 328, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 324, \"end\": 328, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 324, \"end\": 328}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 324, \"end\": 328, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.7690966725349426, \"start\": 324, \"end\": 328}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4472680389881134, \"start\": 324, \"end\": 328}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.682553768157959, \"start\": 324, \"end\": 328}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.682553768157959, \"start\": 324, \"end\": 328}}, {\"mention_name\": {\"text\": \"LFS 2017\", \"confidence\": 0.35530760884284973, \"start\": 340, \"end\": 348, \"clean_text\": \"LFS 2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 340, \"end\": 348, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 340, \"end\": 348, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 340, \"end\": 348}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 340, \"end\": 348, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.7690966725349426, \"start\": 340, \"end\": 348}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.35530760884284973, \"start\": 340, \"end\": 348}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.682553768157959, \"start\": 340, \"end\": 348}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.682553768157959, \"start\": 340, \"end\": 348}}, {\"mention_name\": {\"text\": \"national poverty headcount\", \"confidence\": 0.4993620812892914, \"start\": 369, \"end\": 395, \"clean_text\": \"national poverty headcount\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 369, \"end\": 395, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 369, \"end\": 395, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 369, \"end\": 395}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 369, \"end\": 395, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.6692121624946594, \"start\": 369, \"end\": 395}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4993620812892914, \"start\": 369, \"end\": 395}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.758123517036438, \"start\": 369, \"end\": 395}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.758123517036438, \"start\": 369, \"end\": 395}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [38]}}, {\"page\": 39, \"chunk\": 39, \"input_text\": \"~~-~~ 1(y Jalyy mn(y ,z)-—mn(y\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [39]}}, {\"page\": 40, \"chunk\": 40, \"input_text\": \"**Almost 50 percent of the Consumable Income poor in the West Bank and Gaza experience some degree of FI due to the combination of taxes and transfers modeled, with an average amount of FI of NIS 41.7 per Consumable income poor person, per month, while 71 percent of the pre-fiscal poor population experience FGP, with an average amount of FGP of NIS 38.4 per Market income poor person, per month.** Figure 5-3a shows the proportion of the poor population that, relative to their pre-fiscal income levels, experience FI (depicted by a negative blue bar) and the proportion that experience FGP (depicted by a positive orange bar). Figure 5.3b shows the average monthly amount of FI (negative blue bar) and FGP (positive orange bar), per pre-fiscal and post-fiscal poor person respectively, in LCU. Note that the shares of the poor population that experience FI and FGP at Consumable Income, in Figure 5.3, sum to a value greater than one because the FI is calculated as a share of the Consumable Income poor (i.e. 950,727 individuals divided by 1,929,374 Consumable Income poor individuals = 49.3 percent), whereas the FGP is calculated as a share of the Market Income poor (i.e. 1,067,067 individuals divided by 1,511,930 Market Income poor individuals = 70.6 percent).\\n\\n**Of the net market income poor population, 30.1 percent experiences fiscal impoverishment due to the direct taxes on income and property and contributions to pensions, with an average amount of NIS 26.7 (USD 6.9) per month.** When including cash and near-cash direct transfers and pension income into the system, the proportion of the disposable income poor that experience FI is 20.9 percent, by an average amount of NIS 4.4 (USD 1.1) per month. The proportion of the market income poor that experience FGP due to the combination of direct taxes and transfers is 25.8 percent of the population, with an average gain of NIS 39.3 (10.2 USD) per month. Finally, when the indirect taxes and subsidies are included in the system (VAT, customs tax, and electricity and fuel subsidies), the share of the Consumable income poor population that experience FI is 49.3 percent, and the average value of FI increases to NIS 41.7 (10.8 USD) per month, while the proportion of the market income poor population that experiences FGP rises to 70.6 percent, with an average value of FGP of NIS 38.4 (USD 10.0) per month.\\n\\n38\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [40]}}, {\"page\": 41, \"chunk\": 41, \"input_text\": \"<!-- Start of picture text --> Share of poor population Average amount per poor population, monthly<br>100 40 39.3 384<br>70.6 ><br>=<br>£i< 50 5= 6<br>3s<br>a258 3<br>3 0.0 3 0.0<br>BaSw 8a 0<br>ze 2 44<br>&r<br>8os<br>oOo Sot -20.9 5E<br>5 -50 neo<br>a -49.3 2<br>g<br>< 30.0<br>-100 -40<br>———— ett.7<br>Net market Disposable Consumable Net market Disposable Consumable<br>@ Fiscal Impoverishment MH Fiscal Gains to the Poor ‘Total Fl “Total FGP<br>erelotes: Authors' estimates based on PECS 2016/2017; LFS 2017.<br>i2 ] Fiscal impoverishment:gains to the poor:PercentPercent ofofMeerapopulationi)i) pop o orr at postat pre -fiscal income and i i) madeexperience poorergains by thein fiscalincome systdu e m.to the fiscal system.<br>[3] Pre-fiscal income is excluded from the graph because no fiscal impoverishment or fiscal gains can occur, by definition.<br>[4] Final income is excluded from the graph because we don’t measure changes in poverty at Final income.<br>[5] Sum of absolute value of bars doesn't add to 100 because percentages are calculated over poor populations at different income concepts.<br><!-- End of picture text -->\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.5205776691436768, \"start\": 490, \"end\": 494, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 490, \"end\": 494, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 490, \"end\": 494, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 490, \"end\": 494}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 490, \"end\": 494, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.913944661617279, \"start\": 183, \"end\": 187}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5205776691436768, \"start\": 183, \"end\": 187}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6450962424278259, \"start\": 183, \"end\": 187}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6450962424278259, \"start\": 183, \"end\": 187}}, {\"mention_name\": {\"text\": \"LFS 2017\", \"confidence\": 0.3646624684333801, \"start\": 506, \"end\": 514, \"clean_text\": \"LFS 2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 506, \"end\": 514, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 506, \"end\": 514, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 506, \"end\": 514}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 506, \"end\": 514, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8510153293609619, \"start\": 506, \"end\": 514}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3646624684333801, \"start\": 506, \"end\": 514}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6083838939666748, \"start\": 506, \"end\": 514}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6083838939666748, \"start\": 506, \"end\": 514}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [41]}}, {\"page\": 42, \"chunk\": 42, \"input_text\": \"(comprised of the poorest 10 percent of households) received a net cash benefit equivalent to an increase of 53.4 percent of their Market Income; these households in the poorest decile benefitted more from direct transfers and indirect subsidies than total direct and indirect taxes that they had to pay. In contrast, the rest of the deciles were net cash payers into the fiscal system, with the cash loss after taxes and cashable transfers (reduction in Market Income) ranging from –1.4 percent in decile 2 to –19.7 percent in decile 10 (the richest 10 percent of households). The fact that decile 2 (which is below the official poverty line) experienced a cash loss is consistent with the national poverty increase estimated after the combination of taxes and cashable transfers modeled. While there was a slight poverty gap increase at the national level, the effect was very small and likely because of the net cash gain in decile 1. Also, the fact that the cash loss was higher for richer deciles relative to poorer deciles is consistent with the estimated inequality reduction after the combination of taxes and transfers modeled.\\n\\n**Figure 5-4 Households in decile 1 are net cash beneficiaries of fiscal policy in the West Bank and Gaza**\\n\\nNet cash position of households after taxes and transfers, by deciles <!-- Start of picture text --> Net cash beneficiary<br>100<br>Contributions<br>Direct taxes<br>50<br>Indirect taxes<br>Education<br>Health<br>Indirect subsidies<br>Direct transfers<br>0<br>Net total benefit<br>Net cash benefit<br>-50<br>Poorest 2 3 4 5 6 7 8 9 Richest<br>Decile of Market income<br>Percent of Market income<br><!-- End of picture text --> Source: Authors' estimates based on PECS 2016/2017; LFS 2017. Notes:\\n\\n[1] Net cash benefit refers to the system including all direct and indirect taxes, transfers and subsidies, and excluding in-kind health and education transfers. [2] Net total benefit refers to the system including all elements, including in-kind health and education transfers.\\n\\n40\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.5789967179298401, \"start\": 1709, \"end\": 1713, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1709, \"end\": 1713, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1709, \"end\": 1713, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1709, \"end\": 1713}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1709, \"end\": 1713, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9407825469970703, \"start\": 229, \"end\": 233}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5789967179298401, \"start\": 229, \"end\": 233}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5324223637580872, \"start\": 229, \"end\": 233}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5324223637580872, \"start\": 229, \"end\": 233}}, {\"mention_name\": {\"text\": \"LFS 2017\", \"confidence\": 0.3170207142829895, \"start\": 1725, \"end\": 1733, \"clean_text\": \"LFS 2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1725, \"end\": 1733, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1725, \"end\": 1733, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1725, \"end\": 1733}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1725, \"end\": 1733, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9407825469970703, \"start\": 245, \"end\": 253}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3170207142829895, \"start\": 245, \"end\": 253}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5324223637580872, \"start\": 245, \"end\": 253}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5324223637580872, \"start\": 245, \"end\": 253}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [42]}}, {\"page\": 43, \"chunk\": 43, \"input_text\": \"- **6 Marginal contributions and progressivity**\\n\\n# **6.1 Marginal contributions**\\n\\n**Marginal contributions are defined as the contribution a particular fiscal intervention makes to the measured outcome of interest (Lustig, 2022)** . Marginal contributions consider both the progressivity and the size of an instrument or group of instruments. They are calculated by measuring the change in a poverty or inequality indicator with and without the tax or transfer of interest. A positive marginal contribution in Figure 6-1 and Figure 6-2 below signifies that the tax or transfer is inequality- or povertyreducing while a negative marginal contribution means that the tax or transfer is inequality- or povertyincreasing. In this section we examine the marginal contributions to inequality and then poverty.\\n\\n**Marginal contributions are measured here once all other** **_cash_ taxes and transfers are included in the system (at Consumable Income) given that this best represents the fiscal system for poverty estimates.** This can be more useful for explaining the final result of including or excluding a particular instrument, because the interaction of one instrument with the other instruments already included in the fiscal system can, at times, result in a starkly different marginal contribution estimate than if included at Market Income (i.e., with no other fiscal instruments included in the model). However, for a better understanding of the impact of a particular instrument it can be informative to examine the contributions of direct taxes and transfers at market income, indirect taxes and subsidies at disposable income, and in-kind benefits at consumable income (when measuring inequality). We therefore show the marginal contributions at Market, Disposable and Consumable income in Annex C. For example, the marginal contribution of the Pension income to poverty reduction at Market Income is 0.7 percentage points vs. 2.0 percentage points at Consumable Income (once the direct taxes and transfers, and indirect taxes and subsidies are included).\\n\\n## **6.1.1 Inequality**\\n\\n**Except for VAT (indirect effect), all instruments in the West Bank and Gaza are either inequalityreducing or have no effect.** Instruments do not affect outcomes in some cases because they are small (see, for instance, Figure 7-1 on the incidence of taxes and transfers).\\n\\n## **The three instruments with the largest marginal contribution to inequality-reduction are all transfers:**\\n\\n- In-kind benefits from public basic education have the largest marginal contribution to inequality reduction at 1.4 Gini points due to their large size and despite being less progressive than direct transfers (Figure 6-1). This is expected since poor households in the West Bank and Gaza have more people and more children attending public schools.\\n\\n- The second largest marginal contribution to inequality reduction stems from in-kind benefits from public hospitals services at 1.3 Gini points. While they have the same level of progressivity as the public primary healthcare services, the expenditure size of the former is larger and so the marginal contribution is greater.\\n\\n- These are followed by the CTP (direct transfer) at 1.2 Gini points. The large marginal contribution to inequality reduction from the CTP is expected given that it is targeted to the poor using a PMTF.\\n\\n**The VAT, the largest of all taxes, has the next largest marginal contribution to inequality reduction at 0.8 Gini points.** This is due to a combination of two characteristics: its progressivity and its substantial size relative to a households’ consumption. Most of the marginal contribution from VAT stems from the direct effects (tax levied on final goods and services), while the indirect effects (price increase on informal and\\n\\n41\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [43]}}, {\"page\": 44, \"chunk\": 44, \"input_text\": \"exempt items due to taxed production inputs) have a small (negative) effect on inequality reduction. While a VAT without exemptions, thresholds, or tax avoidance would have no impact on inequality (since the concentration of VAT would be similar to the concentration of consumption), in reality the tax is progressive given that the poor source a larger share of total consumption at informal places of purchase (Bachas et al 2020).\\n\\n**Public secondary education benefits, public primary healthcare benefits, fuel excise (direct effect), nearcash social protection transfers, electricity subsidies (direct effects), pension income and pension contributions also make positive marginal contributions to inequality reduction (ranging between 0.6 and 0.1 Gini points) while the PIT has a neutral effect on inequality reduction.** Unusually, in the West Bank and Gaza, the PIT has a negligeable effect on inequality reduction (0.0 Gini points), while the VAT decreases inequality by 0.8 points. This can be attributed to the small size of the PIT and the comparatively flat structure of the tax system.\\n\\n**VAT (indirect effects), Fuel excises and fuel subsidies have negative marginal contribution to the inequality reduction (Figure 6-4).** Finally, our estimates show that the indirect effect of VAT is the only fiscal component that increases inequality (with a marginal contribution of -0.1 Gini points) (Figure 6-4). As for fuel excises, only the direct effects decrease inequality (by 0.4 Gini points) as fuel is consumed more by wealthier households, while the indirect effect of the fuel excise has a negligeable effect on inequality reduction.<sup>69</sup>\\n\\n> 69 Indirect effects of fuel excises represent about 75 percent of total fuel excises. The results show that for deciles 1-8, the largest source of indirect effects stems from “Taxi fares (public transport)” whereas for deciles 9-10, the largest source of indirect effects stems from “Electricity charges”.\\n\\n42\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [44]}}, {\"page\": 45, \"chunk\": 45, \"input_text\": \"# **Figure 6-1 Almost all fiscal instruments in the West Bank and Gaza are either inequality-reducing or have no effect**\\n\\nMarginal contribution to inequality reduction of each fiscal instrument, in Gini points, measured at Consumable Income <!-- Start of picture text --> Inequality reduction<br>VAT (indirect effect)-0.1<br>Fuel excise (indirect effect) -0.0<br>Customs (direct effect) -0.0<br>Fuel subs. (direct effect) -0.0<br>Fuel subs. (total) -0.0<br>Property -0.0<br>Fuel subs. (indirect effect) 0.0<br>PIT 0.0<br>Pension contributions 0.1<br>Pension income 0.1<br>Elec. subs. (direct effect) 0.1<br>Near-cash transfers 0.2<br>Fuel excise (total) 0.4<br>Fuel excise (direct effect) 0.4<br>Primary healthc. 0.4<br>Secondary educ. 0.6<br>VAT (direct + indirect effect) 0.8<br>VAT (direct effect) 0.9<br>Cash Transfer Program (CTP) 1.2<br>Hospital healthc. 1.3<br>Basic educ. 1.4<br>0 .5 1 1.5<br>Marginal contribution at Consumable income (Gini points)<br>Source: Authors' estimates based on PECS 2016/2017; LFS 2017.<br>Notes:<br>[1] Marginal contributions are calculated as the change in a poverty or inequality indicator that results with the inclusion of a tax or transfer of interest.<br>[2] A positive bar represents a reduction in poverty or inequality, while a negative bar represents an increase.<br><!-- End of picture text -->\\n\\n# **6.1.2 Poverty**\\n\\n**All transfers are either poverty-reducing or neutral.** The CTP (largest targeted social protection program) and pension income have the largest poverty reduction effects in the West Bank and Gaza, reducing the national poverty headcount by 2.2 and 2.0 percentage points, respectively. While wealthier households capture a larger share of pension income than their share of Market income, a portion of the transfer still goes to poorer households, and the transfer increases income in Decile 1 by 1.6 percent, and by 3.9 and 0.7 percent in deciles 2 and 3 respectively. Electricity subsidies (direct effects), fuel subsidies (direct and indirect effects), and near-cash transfers have a much smaller marginal contribution to poverty reduction at only 0.6 and 0.3 percentage points, respectively.\\n\\n**All taxes either increase poverty or are neutral. Customs duties and VAT have the largest (negative) contribution to poverty reduction as they increase the national poverty headcount by 5.3 and 4.8 percentage points, respectively.** Although nationally the VAT places a larger tax burden on households relative to customs duties (at 7.3 and 6.3 percent of Market Income, respectively), the VAT increases\\n\\n43\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.57591712474823, \"start\": 998, \"end\": 1002, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 998, \"end\": 1002, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 998, \"end\": 1002, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 998, \"end\": 1002}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 998, \"end\": 1002, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8888963460922241, \"start\": 476, \"end\": 480}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.57591712474823, \"start\": 476, \"end\": 480}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6580405831336975, \"start\": 476, \"end\": 480}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6580405831336975, \"start\": 476, \"end\": 480}}, {\"mention_name\": {\"text\": \"LFS 2017\", \"confidence\": 0.3394412100315094, \"start\": 1014, \"end\": 1022, \"clean_text\": \"LFS 2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1014, \"end\": 1022, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1014, \"end\": 1022, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1014, \"end\": 1022}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1014, \"end\": 1022, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8888963460922241, \"start\": 492, \"end\": 500}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3394412100315094, \"start\": 492, \"end\": 500}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6580405831336975, \"start\": 492, \"end\": 500}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6580405831336975, \"start\": 492, \"end\": 500}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [45]}}, {\"page\": 46, \"chunk\": 46, \"input_text\": \"poverty by less than the customs duties because it is a more progressive tax<sup>70</sup> (Figure 6-4). The total PIT contribution in the combined West Bank and Gaza only increases the national poverty headcount by 0.4 percentage points. This is because the PIT is a smaller and more targeted tax than indirect taxes, and because labor informality is high in the West Bank and Gaza and reducing with income (the share of formal workers ranges from 4.3 percent of workers in decile 1 to 11.8 percent in decile 10). Fuel excises are poverty-increasing, and their indirect effects are more poverty-increasing with respect to their direct effects (2.3 vs. 0.5 percentage points, respectively). The indirect effects of fuel excises have a larger marginal contribution to the poverty increase relative to the direct effects because they are larger in size (75 percent of total fuel excises simulated); and for households in deciles 1-8 the indirect effects from fuel excises stem from their use of public transport (taxi fares). **We do not show the marginal contributions of the in-kind benefits to poverty-reduction (Figure 6.2).** A typical CEQ Assessment does not measure effects from in-kind benefits on household purchasing power, as outlined in the section on methodology.\\n\\n# **Figure 6-2 In the West Bank and Gaza, direct transfers and the direct effects of electricity subsidies are the only poverty-reducing instruments**\\n\\nMarginal contribution to poverty reduction of each fiscal instrument, in percentage points, measured at Consumable income <!-- Start of picture text --> Poverty reduction<br>Customs (direct effect) -5.3<br>VAT (direct + indirect effect) -4.8<br>VAT (indirect effect) -3.0<br>Fuel excise (total) -2.7<br>VAT (direct effect) -2.5<br>Fuel excise (indirect effect) -2.3<br>Pension contributions -1.6<br>Fuel excise (direct effect) -0.5<br>PIT Poverty increasing -0.4<br>Property -0.0<br>Fuel subs. (direct effect) 0.0<br>Near-cash transfers 0.2<br>Fuel subs. (total) 0.3<br>Fuel subs. (indirect effect) 0.3<br>Elec. subs. (direct effect) 0.6<br>Pension income 2.0<br>Cash Transfer Program (CTP) 2.2<br>-6 -4 -2 0 2<br>Marginal contribution at Consumable income (percentage points)<br><!-- End of picture text --> Source: Authors' estimates based on PECS 2016/2017; LFS 2017.\\n\\nNotes:\\n\\n- [1] Marginal contributions are calculated as the change in a poverty or inequality indicator that results with the inclusion of a tax or transfer of interest. [2] A positive bar represents a reduction in poverty or inequality, while a negative bar represents an increase. [4] We don't calculate contributions to poverty for in-kind subsidies.\\n\\n70 VAT Progressivity in this fiscal incidence study is driven by the fact that the model accounts for households’ consumption informality (which is higher among poorer households), which improves VAT progressivity (Bacha et al. 2020). In contrast, for the case of customs duties, the model assumes the Law of One Price.\\n\\n44\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.5441809892654419, \"start\": 2272, \"end\": 2276, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2272, \"end\": 2276, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2272, \"end\": 2276, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2272, \"end\": 2276}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2272, \"end\": 2276, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9407824277877808, \"start\": 296, \"end\": 300}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5441809892654419, \"start\": 296, \"end\": 300}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5324221849441528, \"start\": 296, \"end\": 300}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5324221849441528, \"start\": 296, \"end\": 300}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [46]}}, {\"page\": 47, \"chunk\": 47, \"input_text\": \"# **6.2 Progressivity**\\n\\n**The** **_Kakwani_ Index is a summary statistic of progressivity.** For taxes, it is calculated by subtracting the concentration coefficient of the tax from the Gini coefficient of a reference income (in this case Market Income);<sup>71</sup> for transfers, the index is calculated by subtracting the Gini coefficient from the concentration coefficient of the transfer. A value greater than zero represents a progressive tax or transfer, while a value below zero represents a regressive tax or transfer. The minimum and maximum values of the Kakwani Index depend on the size of the Gini coefficient. With a Gini coefficient at Market income of 35.1, the maximum value of the Kakwani is 64.9 for taxes (100-35.1), and 135.1 for transfers (35.1+100). The values of the Kakwani Index for transfers (Figure 6-3) are therefore not comparable with those of taxes (Figure 6-4).\\n\\n**Most fiscal instruments in the West Bank and Gaza are progressive at Market income except for fuel subsidies (direct and indirect effects) and pension income. The direct transfers and in-kind benefits from health and education are the most progressive instruments in the West Bank and Gaza.** The CTP and near-cash transfers score the highest on the Kakwani index (at 100.3 and 73.0, respectively). The inkind health and education benefits are substantially less progressive than the direct transfers – which is unsurprising given that the latter directly target the poor, particularly through the CTP program.<sup>72</sup> Hospital and primary healthcare both score 47.9 on the Kakwani Index, while basic and secondary education score 45.3 and 39.0 respectively.\\n\\n**Fuel subsidies and pension income are regressive (they have a negative score on the Kakwani Index).** This is not unusual in either case since fuel subsidies generally benefit the wealthy who spend a larger share of their consumption on fuel, while Pension Income is mostly awarded to public sector employees.<sup>73</sup> This means that the rich are capturing a larger share of these transfers relative to their income shares than the poor; hence, these instruments represent an inefficient means of supporting the poor. While Pension Income is of course not designed solely to support the poor, it is nonetheless important to note that the inclusion of the pension system results in a large redistribution of income from pension contributors to pension recipients.\\n\\n> 71 The concentration coefficient is a summary statistic of the concentration shares and is thus a measure of the absolute distribution of a tax or transfer. It is measured similarly to the Gini coefficient, and a positive concentration coefficient indicates that the rich pay a greater share of a tax (receive a greater share of a benefit) than their share of the population, while a negative concentration coefficient indicates the opposite.\\n\\n> 72 Even though the CTP program could not be identified in the PECS, it is likely that the program contributes largely to the aggregate results found in “direct transfers”, since the CTP represents about 63 percent of Ministry of Social Development budget.\\n\\n> 73 According to the PER (World Bank, 2016), as of September 2014, only 2.2 percent of contributors were from the private sector (3,350 out of 154,986 total contributors).\\n\\n45\", \"datasets\": [{\"mention_name\": {\"text\": \"Kakwani Index\", \"confidence\": 0.6945486068725586, \"start\": 1579, \"end\": 1592, \"clean_text\": \"Kakwani Index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1579, \"end\": 1592, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1579, \"end\": 1592, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1579, \"end\": 1592}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1579, \"end\": 1592, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.8499511480331421, \"start\": 858, \"end\": 871}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6945486068725586, \"start\": 858, \"end\": 871}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8803054690361023, \"start\": 858, \"end\": 871}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8803054690361023, \"start\": 858, \"end\": 871}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [47]}}, {\"page\": 48, \"chunk\": 48, \"input_text\": \"# **Figure 6-3 All transfers are progressive, except for fuel subsidies and pension income**\\n\\nProgressivity of transfers, measured by the Kakwani Index at Market Income <!-- Start of picture text --> a. Progressivity of transfers<br>Fuel subs. (direct effect) -23.0<br>Pension income -14.0<br>Fuel subs. (total) -8.3<br>Fuel subs. (indirect effect) -4.7<br>Elec. subs. (direct effect) 33.4<br>Secondary educ. 39.0<br>Basic educ. 45.2<br>Primary healthc. 47.9<br>Hospital healthc. 47.9<br>Near-cash transfers 73.0<br>Cash Transfer Program (CTP) 100.3<br>-50 0 50 100<br>Kakwani Index w.r.t. Market Income<br><!-- End of picture text --> Source: Authors' estimates based on PECS 2016/2017; LFS 2017. Notes: [1] The Kakwani Index is a summary statistic of progressivity [>0 = progressive]. [2] It is calculated for transfers by subtracting the concentration coefficient from the Gini coefficient.\\n\\n**All taxes are progressive relative to Market income. The PIT is the most progressive, followed by the total VAT, the fuel excises, and the customs tax (with a score of 14.1, 12.6, 11.9 and 3.7 on the Kakwani Index respectively).** The PIT is only slightly more progressive than the total VAT (direct and indirect effects combined). This is because of the flat design of the PIT in the West Bank and Gaza with relatively few income brackets.<sup>74</sup>\\n\\n**The progressivity of the VAT is primarily driven by its direct effects, with a Kakwani index of 19.5, higher than both the PIT and the indirect effects of the VAT (3.2 on the Kakwani index).** The direct effects are more targeted than the indirect effects due to exemptions and zero-ratings on certain strategic goods. However, the progressivity of the VAT is also partly due to our methodological approach which accounts for differential informal consumption across deciles. The poor have a higher share of informal consumption, thus making the VAT more progressive (Bachas et al. 2020).\\n\\n**The direct effects of the fuel excise are the most progressive component of the fiscal system.** Although it is not designed for progressivity, the consumption patterns of wealthier households render it particularly progressive.\\n\\n> 74 The PIT paid according to the West Bank and Gaza tax schedule has a Kakwani Index of only 13.1 compared to the PIT paid according to Israeli tax schedule which has a Kakwani Index of 57.6.\\n\\n46\", \"datasets\": [{\"mention_name\": {\"text\": \"Kakwani Index\", \"confidence\": 0.8426668643951416, \"start\": 138, \"end\": 151, \"clean_text\": \"Kakwani Index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.434627890586853, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 138, \"end\": 151, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 138, \"end\": 151}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 138, \"end\": 151, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9613550305366516, \"start\": 138, \"end\": 151}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7687200307846069, \"start\": 138, \"end\": 151}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8196393847465515, \"start\": 138, \"end\": 151}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8196393847465515, \"start\": 138, \"end\": 151}}, {\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.41549360752105713, \"start\": 672, \"end\": 676, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 672, \"end\": 676, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 672, \"end\": 676, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 672, \"end\": 676}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 672, \"end\": 676, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9407825469970703, \"start\": 184, \"end\": 188}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.41549360752105713, \"start\": 184, \"end\": 188}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5324223637580872, \"start\": 184, \"end\": 188}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5324223637580872, \"start\": 184, \"end\": 188}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [48]}}, {\"page\": 49, \"chunk\": 49, \"input_text\": \"# **Figure 6-4 All taxes are progressive. The PIT is the most progressive tax, however the direct effects of**\\n\\n**fuel excises are the most progressive component of the fiscal system.**\\n\\nProgressivity of taxes, measured by the Kakwani Index at Market Income <!-- Start of picture text --> b. Progressivity of taxes<br>VAT (indirect effect) 3.2<br>Customs (direct effect) 3.7<br>Fuel excise (indirect effect) 4.9<br>Property 5.9<br>Pension contributions 10.4<br>Fuel excise (total) 11.9<br>VAT (direct + indirect effect) 12.6<br>PIT 14.1<br>VAT (direct effect) 19.5<br>Fuel excise (direct effect) 26.0<br>0 5 10 15 20 25<br>Kakwani Index w.r.t. Market Income<br><!-- End of picture text --> Source: Authors' estimates based on PECS 2016/2017; LFS 2017. Notes:\\n\\n[1] The Kakwani Index is a summary statistic of progressivity [>0 = progressive].\\n\\n[2] It is calculated for taxes by subtracting the Gini coefficient from the concentration coefficient.\\n\\n47\", \"datasets\": [{\"mention_name\": {\"text\": \"Kakwani Index\", \"confidence\": 0.8181499242782593, \"start\": 227, \"end\": 240, \"clean_text\": \"Kakwani Index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 227, \"end\": 240, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 227, \"end\": 240, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 227, \"end\": 240}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 227, \"end\": 240, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9138190746307373, \"start\": 227, \"end\": 240}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8181499242782593, \"start\": 227, \"end\": 240}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8402416706085205, \"start\": 227, \"end\": 240}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8402416706085205, \"start\": 227, \"end\": 240}}, {\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.4674564003944397, \"start\": 726, \"end\": 730, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 726, \"end\": 730, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 726, \"end\": 730, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 726, \"end\": 730}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 726, \"end\": 730, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9407824277877808, \"start\": 189, \"end\": 193}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4674564003944397, \"start\": 189, \"end\": 193}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5324221849441528, \"start\": 189, \"end\": 193}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5324221849441528, \"start\": 189, \"end\": 193}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [49]}}, {\"page\": 50, \"chunk\": 50, \"input_text\": \"# **7 Incidence and concentration shares by decile**\\n\\n**This section shows the distribution of each tax and transfer both in absolute terms (concentration shares) and relative to Market Income (incidence).** Households are sorted by real, per capita market income and grouped into national deciles of equal population. Decile 1 represents the poorest households, while decile 10 represents the richest households. Measuring the incidence of taxes and transfers by decile shows the total size of a tax or transfer relative to the total size of income in that decile. Incidence therefore measures size and distribution relative to a reference income (in this case Market Income).\\n\\n**Concentration shares, in contrast, measure the proportion of each tax or transfer paid or received by each decile.** The sum of bars in each decile, for each instrument, must sum to 1 or 100 percent, and so the height of the bars provides no indication of size, only of distribution. To determine whether the instrument is progressive, we need to analyze the distribution compared to the distribution of a chosen reference income, therefore we show Market Income (red dashed line in graphs). A negative bar signals a tax, and a positive bar signals a transfer.\\n\\n## **7.1. Taxes**\\n\\n**The incidence of the direct taxes (PIT, property taxes) is small across all the deciles, despite increasing more sharply than the indirect taxes (Figure 7-1a).**\\n\\nWhile they are most concentrated among richer households (Figure 7-1b) they are (unusually) not much more progressive than the indirect taxes in the West Bank and Gaza, given the poor targeting of the PIT and the relatively small size of the property taxes (Figure 7-1).\\n\\n**The incidence of the Indirect taxes is large relative to the direct taxes (Figure 7-1a).** The direct effects of customs duties (brown bar) have the largest incidence of all tax instruments across deciles 1-6 of Market Income; they range from 4.1 percent of Market Income in decile 1 (the poorest) to 5.9 percent in decile 10 (the richest). In contrast, in deciles 7-10, the incidence of the VAT is largest. The direct effects of the VAT (red bar) are larger than the indirect effects in the upper half of the population and increase more sharply with income. The direct effects range from 1.3 percent of Market Income in decile 1 to 6.5 percent in decile 10. The indirect effects of VAT (purple bar) range from 1.9 percent of Market Income in decile 1 to 2.8 percent in decile 10. The incidence of fuel excise ranges from 2.0 percent of Market Income in decile 1 to 4.9 percent in decile 10. The direct effects (pink bar) are smaller (ranging from 0.3 in decile 1 to 2.3 in decile 10), while the indirect effects (grey bar) are larger (ranging from 1.7 in decile 1 to 2.6 in decile 10). The incidence of both the direct and indirect effects of the fuel excise is larger for richer deciles, reflecting that fuel is a luxury good and that even the consumption of goods into which fuel is an input, are more concentrated among richer households than income.\\n\\n**The concentration of indirect taxes in deciles 1-3 is similar to the direct taxes, at 6.3 and 4.6 respectively of the total tax collected (Figure 7-1b) and increases only slightly less steeply than the direct taxes.** The share of the direct taxes concentrated in decile 10 is 28.8 percent of income - just higher than the share of the indirect taxes at 28.3 percent of income.\\n\\n48\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [50]}}, {\"page\": 51, \"chunk\": 51, \"input_text\": \"**Pension contributions are fairly flat.** Their incidence ranges from 0.7 (decile 1) and 2.6 (decile 8) (Figure 7-1a). Their concentration is lower than the indirect and direct taxes in decile 10, with a concentration share of 24 percent, versus 28-29 percent for the indirect and direct taxes, respectively (Figure 7-1b).\\n\\n**Figure 7-1 Of all taxes, the Customs Tax and VAT have the largest incidence across the distribution**\\n\\nTaxes and contributions: Relative incidence and concentration shares <!-- Start of picture text --> a. Incidence: taxes & contributions<br>0<br>-5<br>-10<br>-15<br>-20<br>-25<br>Poorest 2 3 4 5 6 7 8 9 Richest<br>PIT Property tax Pension contributions<br>VAT (direct effect) VAT (indirect effect) Customs (direct effect)<br>Fuel excise (direct effect) Fuel excise (indirect effect)<br>Source: Authors' estimates based on PECS 2016/2017; LFS 2017.<br>Notes:<br>[1] Incidence is a measure of size and distribution relative to a reference income.<br>[2] Transfers are shown as a positive bar and taxes are shown as a negative bar.<br>[3] Households are ranked over Market income<br>Incidence (% of Market income)<br><!-- End of picture text --> 49\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.5927929282188416, \"start\": 851, \"end\": 855, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 851, \"end\": 855, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 851, \"end\": 855, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 851, \"end\": 855}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 851, \"end\": 855, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.900557279586792, \"start\": 247, \"end\": 251}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5927929282188416, \"start\": 247, \"end\": 251}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7084320783615112, \"start\": 247, \"end\": 251}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7084320783615112, \"start\": 247, \"end\": 251}}, {\"mention_name\": {\"text\": \"LFS 2017\", \"confidence\": 0.3266867697238922, \"start\": 867, \"end\": 875, \"clean_text\": \"LFS 2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 867, \"end\": 875, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 867, \"end\": 875, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 867, \"end\": 875}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 867, \"end\": 875, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.900557279586792, \"start\": 263, \"end\": 271}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3266867697238922, \"start\": 263, \"end\": 271}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7084320783615112, \"start\": 263, \"end\": 271}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7084320783615112, \"start\": 263, \"end\": 271}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [51]}}, {\"page\": 52, \"chunk\": 52, \"input_text\": \"<!-- Start of picture text --> b. Concentration shares: taxes & contributions<br>30<br>20<br>10<br>0<br>Poorest 2 3 4 5 6 7 8 9 Richest<br>Decile of Market income<br>Direct taxes Contributions Indirect taxes Market income<br>Source: Authors' estimates based on PECS 2016/2017; LFS 2017.<br>Notes:<br>[1] Concentration shares show the proportion of each tax/transfer paid by / benefitting each decile. They sum to 100%.<br>[2] Households are ranked over Market income<br>Concentration shares (%)<br><!-- End of picture text -->\\n\\n# **7.2. Transfers**\\n\\n**The incidence of the social protection direct transfers is higher in poorer deciles; and the largest direct transfer is the CTP.** The incidence of the CTP is largest in the first decile and decreases steeply with income (Figure 7-2a) from 56.3 percent of Market Income in decile 1 to 0.04 in decile 10; this is consistent with the fact that the CTP is targeted to the poor through a PMT. The incidence of the social protection nearcash transfers is much smaller than the CTP in all but the top decile, ranging from 4.4 in decile 1 to 0.06 in decile 10.\\n\\n**Poor households receive a larger share of direct transfers.** Almost 60 percent of the direct transfers are received by decile 1 and only 1.2 percent in decile 10.\\n\\n**The incidence of the indirect subsidies is small.** The electricity subsidies are visible in deciles 1 to 5 only, they start at 1.8 percent of Market Income in decile 1, decrease to 0.6 percent in decile 5, and further to 0.2 percent in decile 10. The fuel subsidies are not visible at all – they start at 0.1 percent of Market income in decile 1 and increase to 0.2 percent in decile 10.<sup>75</sup>\\n\\n**The share of indirect subsidies received by decile is increasing monotonically with income.** While decile 1 receives 6.8 percent of the total indirect subsidies, decile 10 receives 16.1 percent.\\n\\n**In-kind benefits from public hospital healthcare constitute a substantial benefit for households.** They range from 16.8 percent of Market Income in decile 1 to 0.7 percent in decile 10, followed by in-kind\\n\\n> 75 We do not separate out the direct and indirect effects in the graph as they would be too small to be visible.\\n\\n50\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.6087288856506348, \"start\": 261, \"end\": 265, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 261, \"end\": 265, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 261, \"end\": 265, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 261, \"end\": 265}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 261, \"end\": 265, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8695189952850342, \"start\": 261, \"end\": 265}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6087288856506348, \"start\": 261, \"end\": 265}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6497018337249756, \"start\": 261, \"end\": 265}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6497018337249756, \"start\": 261, \"end\": 265}}, {\"mention_name\": {\"text\": \"LFS 2017\", \"confidence\": 0.3572482764720917, \"start\": 277, \"end\": 285, \"clean_text\": \"LFS 2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 277, \"end\": 285, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 277, \"end\": 285, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 277, \"end\": 285}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 277, \"end\": 285, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.8695189952850342, \"start\": 277, \"end\": 285}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3572482764720917, \"start\": 277, \"end\": 285}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6497018337249756, \"start\": 277, \"end\": 285}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6497018337249756, \"start\": 277, \"end\": 285}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [52]}}, {\"page\": 53, \"chunk\": 53, \"input_text\": \"<!-- Start of picture text --> a. Incidence: transfers<br>100<br>_f-)<br>=<br>8<br>£ 80<br>0)<br>PE”4<br>o<br>=> 60<br>-fo}<br>&<br>a 40<br>°<br>®<br>3<br>2 20<br>0<br>Poorest 2 3 4 5 6 7 8 9 Richest<br>3 Pensions &% Cash Transfer Program (CTP) 19 Near-cash transfers<br>G8 Elec. subs. (direct eff.) | MEG) Fuel subs. (direct + indirect eff.) | HE Primary healthc.<br>[5 Hosp. healthe. 9 Basic educ. [3 Seco. educ.<br>Sonejotes: Authors' estimates based on PECS 2016/2017; LFS 2017.<br>[1] Incidence is a measure of size and distribution relative to a reference income.<br>(3{3] TransfersHouseholds areare shownranked as aover positiveMarket barincome and taxes are shown as a negative bar.<br><!-- End of picture text -->\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS 2016/2017\", \"confidence\": 0.6960286498069763, \"start\": 457, \"end\": 471, \"clean_text\": \"PECS 2016/2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 457, \"end\": 471, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 457, \"end\": 471, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 457, \"end\": 471}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 457, \"end\": 471, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.945521891117096, \"start\": 457, \"end\": 471}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6960286498069763, \"start\": 457, \"end\": 471}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7509920001029968, \"start\": 457, \"end\": 471}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7509920001029968, \"start\": 457, \"end\": 471}}, {\"mention_name\": {\"text\": \"LFS 2017\", \"confidence\": 0.41883811354637146, \"start\": 473, \"end\": 481, \"clean_text\": \"LFS 2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 473, \"end\": 481, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 473, \"end\": 481, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 473, \"end\": 481}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 473, \"end\": 481, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.945521891117096, \"start\": 473, \"end\": 481}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.41883811354637146, \"start\": 473, \"end\": 481}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7509920001029968, \"start\": 473, \"end\": 481}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7509920001029968, \"start\": 473, \"end\": 481}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [53]}}, {\"page\": 54, \"chunk\": 54, \"input_text\": \"<!-- Start of picture text --> b. Concentration shares: transfers<br>60<br>=<br>&— 40<br>oO<br>=<br>7)<br>(=<br>2<br>g<br>Is a<br>8 20 om<br>e) ae<br>1S) _-_-—-o ia<br>0<br>, fc a yg « : @g vo ion<br>Poorest 7 3 4 5 6 7 8 9 Richest<br>Decile of Market income<br>13 Direct transfers 4 Indirect subsidies MG Education ME Health -e- Market income<br><!-- End of picture text --> reliajotes: Authors' estimates based on PECS 2016/2017; LFS 2017. [1] Concentration shares show the proportion of each tax/transfer paid by / benefitting each decile. They sum to 100%. [2] Households are ranked over Market income\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.6314646005630493, \"start\": 416, \"end\": 420, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 416, \"end\": 420, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 416, \"end\": 420, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 416, \"end\": 420}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 416, \"end\": 420, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9231957197189331, \"start\": 416, \"end\": 420}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6314646005630493, \"start\": 416, \"end\": 420}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4988376796245575, \"start\": 416, \"end\": 420}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4988376796245575, \"start\": 416, \"end\": 420}}, {\"mention_name\": {\"text\": \"LFS 2017\", \"confidence\": 0.43111366033554077, \"start\": 432, \"end\": 440, \"clean_text\": \"LFS 2017\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 432, \"end\": 440, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 432, \"end\": 440, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 432, \"end\": 440}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 432, \"end\": 440, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9503487944602966, \"start\": 60, \"end\": 68}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.43111366033554077, \"start\": 60, \"end\": 68}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7314682602882385, \"start\": 60, \"end\": 68}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7314682602882385, \"start\": 60, \"end\": 68}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [54]}}, {\"page\": 55, \"chunk\": 55, \"input_text\": \"# **8 International comparison**\\n\\n## **8.1 Inequality**\\n\\n**This section compares fiscal policy effects on inequality across 63 countries assessed in CEQ standard indicators.** We measure this by calculating the reduction in the Gini coefficient from Market Income (prefiscal income) to Final Income (Market Income plus all taxes, transfers, and in-kind benefits from health and education). We refer to the Gini reduction as the “total redistributive impact”. If a country has been assessed multiple times, we include only the most recent datapoint.\\n\\n**Globally, the total redistributive impact of the combination of taxes and transfers modelled ranges from 0.2 to 20.4 Gini points reduction).** The West Bank and Gaza achieves 6.5 Gini points of redistribution—a moderate result relative to other countries.\\n\\n### **Figure 8-1 Total inequality reduction in the West Bank and Gaza is moderate relative to countries in other CEQ assessments**\\n\\nInequality reduction, Gini coefficient (change in Gini from Market to Final Income) <!-- Start of picture text --> 0<br>-5<br>West Bank and Gaza:<br>-10<br>6.5 Gini point reduction in inequality<br>-15<br>-20<br>-25<br>High incomeHigh income Upper middle incomeUpper middle income Lower middle incomeLower middle income Low inLow in c omeome<br>Source: World Bank estimations for the West Bank and Gaza based on PECS 2016-17. CEQ Data Center for the rest of the<br>countries.<br>Spain Uruguay Panama United States Croatia Mauritius Romania South Africa Argentina Brazil Mexico Namibia Georgia Venezuela Costa Rica Botswana Dominican Republic Colombia China Ecuador Thailand Peru Iraq Iran Turkey West Bank and Gaza West Bank and Gaza Belarus Malaysia Jordan Russia Albania Guatemala Paraguay Indonesia eSwatini Lesotho Zambia Tunisia Kenya Ukraine Honduras El Salvador Mongolia Bolivia India Nicaragua Moldova Egypt Tanzania Ghana Sri Lanka Comoros Cambodia Ivory Coast Uganda Burkina Faso Togo Mali Ethiopia Niger Gambia Tajikistan Guinea<br><!-- End of picture text --> Note: Pensions treated as government transfers in China, Georgia, Guinea, Mali, Mongolia, Niger, and West Bank and Gaza.\\n\\n## **8.2 Poverty**\\n\\n**We compare fiscal policy effects on poverty using the 2011 Purchasing Power Parity (PPP) international poverty lines.**<sup>76</sup> Estimates for the economies in each income group use the poverty line appropriate to that income level: US$5.50 per day for high income and upper middle income, US$3.20 per day for lower middle income, or US$1.90 per day for low income. We measure poverty impact by calculating the change\\n\\n> 76 The International poverty lines have been updated to 2017 PPP but the results are not yet available in the CEQ database. The $5.50 in 2011 (PPP) line is the closest of the international poverty lines to the official national poverty line in the West Bank and Gaza.\\n\\n53\", \"datasets\": [{\"mention_name\": {\"text\": \"CEQ standard indicators\", \"confidence\": 0.8615860939025879, \"start\": 149, \"end\": 172, \"clean_text\": \"CEQ standard indicators\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 149, \"end\": 172, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 149, \"end\": 172, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 149, \"end\": 172}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 149, \"end\": 172, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9778624773025513, \"start\": 149, \"end\": 172}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8615860939025879, \"start\": 149, \"end\": 172}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9221449494361877, \"start\": 149, \"end\": 172}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9221449494361877, \"start\": 149, \"end\": 172}}, {\"mention_name\": {\"text\": \"PECS 2016-17\", \"confidence\": 0.692303478717804, \"start\": 1353, \"end\": 1365, \"clean_text\": \"PECS 2016-17\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1353, \"end\": 1365, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1353, \"end\": 1365, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1353, \"end\": 1365}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1353, \"end\": 1365, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9587975740432739, \"start\": 622, \"end\": 634}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.692303478717804, \"start\": 622, \"end\": 634}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6040455102920532, \"start\": 622, \"end\": 634}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6040455102920532, \"start\": 622, \"end\": 634}}, {\"mention_name\": {\"text\": \"2011 Purchasing Power Parity\", \"confidence\": 0.38457006216049194, \"start\": 2211, \"end\": 2239, \"clean_text\": \"2011 Purchasing Power Parity\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2211, \"end\": 2239, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2211, \"end\": 2239, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2211, \"end\": 2239}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2211, \"end\": 2239, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.8279205560684204, \"start\": 846, \"end\": 874}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.38457006216049194, \"start\": 846, \"end\": 874}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9485101699829102, \"start\": 846, \"end\": 874}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9485101699829102, \"start\": 846, \"end\": 874}}, {\"mention_name\": {\"text\": \"international poverty lines\", \"confidence\": 0.6163579821586609, \"start\": 2246, \"end\": 2273, \"clean_text\": \"international poverty lines\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2246, \"end\": 2273, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2246, \"end\": 2273, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2246, \"end\": 2273}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2246, \"end\": 2273, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.7985886931419373, \"start\": 71, \"end\": 98}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6163579821586609, \"start\": 71, \"end\": 98}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9361097812652588, \"start\": 71, \"end\": 98}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9361097812652588, \"start\": 71, \"end\": 98}}, {\"mention_name\": {\"text\": \"CEQ database\", \"confidence\": 0.5035363435745239, \"start\": 2692, \"end\": 2704, \"clean_text\": \"CEQ database\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2692, \"end\": 2704, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2692, \"end\": 2704, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2692, \"end\": 2704}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2692, \"end\": 2704, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.3249514102935791, \"start\": 517, \"end\": 529}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5035363435745239, \"start\": 517, \"end\": 529}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4310993552207947, \"start\": 517, \"end\": 529}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4310993552207947, \"start\": 517, \"end\": 529}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [55]}}, {\"page\": 56, \"chunk\": 56, \"input_text\": \"in the poverty headcount from Market Income (pre-fiscal income) to Consumable Income (Market Income after taxes and transfers, excluding in-kind benefits from health and education). We include 58 countries in total.\\n\\n**Globally, total poverty change due to the combination of taxes and transfers modeled ranges from the worst case of an 8.9 percentage point increase in Armenia to the best case of an 8.4 percentage point decrease in Spain.** The fiscal system in the West Bank and Gaza increases the poverty rate at the $5.5 2011 PPP poverty line by 6.0 percentage points. Note, however, that this result can vary substantially depending on the poverty line used. At the $3.20 and $1.90 2011 PPP per day international poverty lines, the combination of taxes and transfers in the West Bank and Gaza is poverty- _reducing_ by 1.2 and 0.8 percentage points respectively.\\n\\nFigure 8.2 Total poverty impact at the $5.50 line is severe relative to other CEQ assessments Percentage point change in poverty rate, 2011PPP international poverty lines (changes from Market to Consumable Income)\\n\\n# **Figure 8-2 Total poverty impact at the $5.50 line is severe relative to countries in other CEQ assessments**\\n\\nPercentage point change in poverty rate, 2011 PPP international poverty line (changes from Market to Consumable Income) <!-- Start of picture text --> 10<br>8<br>West Bank and Gaza: 6.0 percentage point increase in poverty rate<br>6<br>4<br>2<br>0<br>-2<br>-4<br>-6<br>-8<br>-10<br>High incomeHigh income Upper middle incomeUpper middle income Lower mLower m idd le incomele income Low Low incom ee<br>Spain Mauritius United States Panama Chile Uruguay Romania Argentina Venezuela Iraq Ecuador Jordan Thailand Paraguay Russia Iran Malaysia Colombia Russia Indonesia South Africa Costa Rica Dominican Republic China Botswana Mexico Peru Namibia Guatemala Brazil Turkey West Bank and Gaza West Bank and Gaza Albania Armenia Ukraine Bolivia Sri Lanka Vietnam Nicaragua Moldova Comoros India Tunisia El Salvador Honduras Zambia eSwatini Lesotho Ghana Ivory Coast Kenya Tajikistan Ethiopia Uganda Guinea Burkina Faso Togo Tanzania<br><!-- End of picture text --> Source: World Bank estimations for the West Bank and Gaza based on PECS 2016-17. CEQ Data Center for the rest of the countries.\\n\\nNote: Pensions treated as government transfers in China, Georgia, Guinea, Mali, Mongolia, Niger, and West Bank and Gaza.\\n\\n54\", \"datasets\": [{\"mention_name\": {\"text\": \"poverty headcount\", \"confidence\": 0.5350432991981506, \"start\": 7, \"end\": 24, \"clean_text\": \"poverty headcount\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 7, \"end\": 24, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 7, \"end\": 24, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"2011\", \"confidence\": 0.3552081286907196, \"start\": 526, \"end\": 530}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 7, \"end\": 24, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.3177725076675415, \"start\": 7, \"end\": 24}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5350432991981506, \"start\": 7, \"end\": 24}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7252963185310364, \"start\": 7, \"end\": 24}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7252963185310364, \"start\": 7, \"end\": 24}}, {\"mention_name\": {\"text\": \"PECS 2016-17\", \"confidence\": 0.6882439851760864, \"start\": 2224, \"end\": 2236, \"clean_text\": \"PECS 2016-17\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2224, \"end\": 2236, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2224, \"end\": 2236, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2224, \"end\": 2236}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2224, \"end\": 2236, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9989263415336609, \"start\": 797, \"end\": 809}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6882439851760864, \"start\": 797, \"end\": 809}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6089823842048645, \"start\": 797, \"end\": 809}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6089823842048645, \"start\": 797, \"end\": 809}}, {\"mention_name\": {\"text\": \"CEQ Data Center\", \"confidence\": 0.303229957818985, \"start\": 2238, \"end\": 2253, \"clean_text\": \"CEQ Data Center\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2238, \"end\": 2253, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2238, \"end\": 2253, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2238, \"end\": 2253}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2238, \"end\": 2253, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9164347052574158, \"start\": 811, \"end\": 826}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.303229957818985, \"start\": 811, \"end\": 826}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5204246044158936, \"start\": 811, \"end\": 826}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5204246044158936, \"start\": 811, \"end\": 826}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [56]}}, {\"page\": 57, \"chunk\": 57, \"input_text\": \"- **9 Macro-validation and limitations of the analysis**\\n\\n- **9.1 Macro-validation**\\n\\n**This section presents an overview of how well each instrument is modeled by comparing survey estimations with official fiscal administrative data.** We call this process “macro-validation” because it allows us to validate our methodology by comparing the results in the survey at the national level with the results in the national accounts (Table 9-1). Note that East Jerusalem is excluded for all the instruments shown in the table as well as for consumption.\\n\\n- **Total consumption:** As outlined earlier, Disposable Income in this study is equivalent to the household consumption aggregate in the PECS 2016/17. The survey has good coverage of households’ consumption as total consumption in the PECS represented about 81 percent of the official aggregate in national accounts.<sup>77</sup> We expected a slight gap between the survey and macro estimations of household consumption due to methodological differences between the survey and national accounts and both non-response rates of wealthier households and underestimation of income levels of wealthier households (see the section on limitations).\\n\\n- **Taxes:** All direct taxes simulated in the PECS have good coverage of their respective administrative absolute values; that is, the amounts allocated in the survey are close in size to the amounts reported in the executed budget. The PIT covers 85 percent of the official administrative value of tax collections; pension contributions covered 108 percent and the property tax covered 84 percent. Indirect taxes—customs duties, VAT and fuel excise—covered 66, 95 and 73 percent, respectively, of total administrative value of tax collections. The Property Tax, and the VAT were scaled to match the effective rate relative to consumption (0.03 and 6.6 percent, respectively).\\n\\n- **Transfers:** The total of CTP and in-kind transfers in the survey covered 65 percent of the administrative data, where income from pensions covered 107 percent (similar to the pension contributions). The value of indirect subsidies from domestic electricity and fuel (jointly) compared to the official administrative value cannot be estimated because the survey data includes both implicit and explicit subsidies, unlike the administrative data. The value of in-kind benefits from primary and hospital public services both represented 85 percent of the administrative value. Lastly, for in-kind benefits from public education, the survey estimates covered 99 percent of the administrative value for basic education and 79 percent for secondary education. Originally basic education benefits were overestimated suggesting that the PECS 2016/17 overestimates total students enrolled in basic education at public schools. Recipients of the basic education subsidy are therefore scaled down as discussed in the methodology section. 78\\n\\n> 77 It is not clear whether the national accounts amount includes East Jerusalem or not. If it does, then it is more appropriate to compare it with consumption including East Jerusalem also, which would result in a ratio of 95 percent.\\n\\n> 78 As highlighted in the methodology, one limitation of the PECS is that the education model did not allow to identify ‘current grade where the student is enrolled’. Hence the current grade was estimated as ‘highest grade achieved’ + 1 year if the student was currently attending a public school.\\n\\n55\", \"datasets\": [{\"mention_name\": {\"text\": \"fiscal administrative data\", \"confidence\": 0.40849819779396057, \"start\": 207, \"end\": 233, \"clean_text\": \"fiscal administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 207, \"end\": 233, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 207, \"end\": 233, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 207, \"end\": 233}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 207, \"end\": 233, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.5943387746810913, \"start\": 207, \"end\": 233}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.40849819779396057, \"start\": 207, \"end\": 233}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6638318300247192, \"start\": 207, \"end\": 233}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6638318300247192, \"start\": 207, \"end\": 233}}, {\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.5955619215965271, \"start\": 787, \"end\": 791, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 787, \"end\": 791, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 787, \"end\": 791, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 787, \"end\": 791}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 787, \"end\": 791, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9906226992607117, \"start\": 787, \"end\": 791}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5955619215965271, \"start\": 787, \"end\": 791}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6818451881408691, \"start\": 787, \"end\": 791}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6818451881408691, \"start\": 787, \"end\": 791}}, {\"mention_name\": {\"text\": \"national accounts\", \"confidence\": 0.34605154395103455, \"start\": 1030, \"end\": 1047, \"clean_text\": \"national accounts\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1030, \"end\": 1047, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1030, \"end\": 1047, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1030, \"end\": 1047}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1030, \"end\": 1047, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.6296972036361694, \"start\": 243, \"end\": 260}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.34605154395103455, \"start\": 243, \"end\": 260}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5881487131118774, \"start\": 243, \"end\": 260}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5881487131118774, \"start\": 243, \"end\": 260}}, {\"mention_name\": {\"text\": \"survey data\", \"confidence\": 0.45941829681396484, \"start\": 2234, \"end\": 2245, \"clean_text\": \"survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2234, \"end\": 2245, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2234, \"end\": 2245, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2234, \"end\": 2245}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2234, \"end\": 2245, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8334231376647949, \"start\": 594, \"end\": 605}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.45941829681396484, \"start\": 594, \"end\": 605}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.38931190967559814, \"start\": 594, \"end\": 605}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.38931190967559814, \"start\": 594, \"end\": 605}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [57]}}, {\"page\": 58, \"chunk\": 58, \"input_text\": \"# **Table 9-1 The ratio of the amount allocated in the survey to the amount recorded in the budget ranges from 65% (government transfers) to 108% (for pension contributions)**\\n\\nMacro-validation of the fiscal incidence analysis in the West Bank and Gaza, 2017\\n\\n|**M**<br>**Scenario : Pensio**|**ACROVALIDATI**<br>**ns as Governme**<br>|**ON**<br>**nt Transfers(**<br>|**PGT)**|**Ratio**|**w.r.t.**|\\n|---|---|---|---|---|---|\\n||**Value in NI**|**S Millions**|**Ratio**|||\\n|||||**consu**|**mption**|\\n|**Fiscal instrument**|||**PECS:**|||\\n||**PECS**<br>**2016/17**|**Admin**<br>**2017**|**Admin**|**PECS**<br>**2016/17**|**Admin**<br>**2017**|\\n|**Consumption**|44 742.3|55 087.2|81.2|100.0|100.0|\\n|**Direct Taxes and contributions**|1 146.7|1 123.9|102.0|2.6|2.0|\\n|Personal Income Tax (PSE)|220.3|260.0|84.7|0.5|0.5|\\n|PropertyTax|12.0|14.3|84.2|0.0|0.0|\\n|Pension contributions|914.4|879.2|104.0|2.0|1.6|\\n|**Direct Transfers and contributory pensions**|2 285.7|2 803.7|81.5|5.1|5.1|\\n|Income from contributory pensions|1 173.1|1 093.2|107.3|2.6|2.0|\\n|Govt assistance (cash and in-kind)|1 112.6|1 710.5|65.0|2.5|3.1|\\n|Cash Transfer Program(CTP)|982.9|||||\\n|In-kind transfers|129.7|||||\\n|**Indirect Taxes**|8 505.7|10 979.5|77.5|19.0|19.9|\\n|VAT|3 468.2|3 646.1|95.1|7.8|6.6|\\n|Direct effects|1 959.9|||||\\n|Indirect effects|1 669.6|||||\\n|Customs Duties|2 926.3|4 429.9|66.1|6.5|8.0|\\n|Fuel excise|2 111.3|2 903.5|72.7|4.7|5.3|\\n|Direct effects|598.4|||||\\n|Indirect effects|1 512.8|||||\\n|**Indirect Subsidies**|333.3|*, **|n.a.|0.7|n.a.|\\n|Fuel|110.4|*|n.a.|0.2|n.a.|\\n|Direct effects|21.3|||||\\n|Indirect effects|89.1|||||\\n|Electricity (direct effects)|223.0|**|n.a.|0.5|n.a.|\\n|**In-Kind Benefits**|3 941.9|4 496.3|87.7|8.8|8.2|\\n|**Health**|**1 723.7**|**2 032.2**|**84.8**|**3.9**|**3.7**|\\n|Primary|424.1|500.0|84.8|0.9|0.9|\\n|Hospital|1 299.6|1 532.2|84.8|2.9|2.8|\\n|**Education**|**2 218.2**|**2 464.1**|**90.0**|**5.0**|**4.5**|\\n|Basic|1 362.2|1 377.8|98.9|3.0|2.5|\\n\\n56\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [58]}}, {\"page\": 59, \"chunk\": 59, \"input_text\": \"Secondary 856.0 1 086.3 78.8 1.9 2.0 Note: Own elaboration. Survey estimates correspond to authors’ simulation of the fiscal incidence model; for indirect taxes, survey estimates correspond to tax collected from West Bank since the PA only collects taxes in this region. For transfers, survey estimates include both West Bank and Gaza since the PA distributes public expenditure in both regions. Fuel excise numbers are sourced from Ministry of Finance monthly reports, Personal Income Tax numbers are sourced from the Income Tax Department of the Ministry of Finance, while pension contribution estimates are from 2014 and imputed based on information on number of each scheme’s number of active contributors, average wage, and rate, sourced from the World Bank (2016).\\n\\n* Bifurcated amount of government spending/allocation for fuel subsidy is not available. IMF estimates (2017) for the fuel subsidies indicates an amount of NIS 251 million which includes both implicit and explicit subsidies and is therefore larger than the estimate of total indirect subsidies discussed elsewhere in the report.\\n\\n*** Bifurcated amount of government spending/allocation for domestic electricity subsidy is not available. The closest estimate of the size of the implicit and explicit subsidy is from 2015. This includes direct and indirect effects, however, and so our estimate is expected to be substantially smaller given that it only includes domestic subsidies.\\n\\n# **9.2 Limitations**\\n\\n## **9.2.1 Direct taxes and pension contributions**\\n\\n## **Personal Income Tax (PIT)**\\n\\n**The main challenge in modeling the PIT was the complexity involved.** Two different tax systems were modeled depending on region and worker status, and each with numerous tax exemptions and deductions. In both systems, information on the order and size of the deductions and exemptions was lacking.\\n\\n**The PECS 2016/17 also does not specify whether income data is net or gross.** It was assumed that employees would provide their income net of tax, while employers or self-employed workers would be more likely to report their gross income.\\n\\nA description of the detailed challenges faced for each deduction can be found in Annex B.\\n\\n## **Pension contributions**\\n\\n**The latest administrative numbers available on pension income and contributions are a few years out of date, from December 2013 and September 2014.** There were also a number of different schemes which made it complex to understand and implement.\\n\\n## **Property tax**\\n\\n**The main limitation for the estimation of the property tax is that there were several pieces of information that were not available in the PECS 2016/17 dataset, including:**\\n\\n- (i) The area of the building and number of floors. This would be used to estimate the annual rental value of the building.\\n\\n- (ii) The lease contract information. Having this information would be useful because the value of the tax varies with occupancy status (rented; prepared for rent; inhabited by the owner). To overcome this limitation, the modeling has proceeded with the assumption that all the properties are inhabited by the owners themselves.\\n\\n- (iii) For Land evaluation, in lieu of the missing survey data, administrative data (by city) on evaluation was used and it was assumed that the valuation value is “average value per property” and not “per square meter”. Due to the lack of update in the evaluation, the evaluation amounts for some\\n\\n57\", \"datasets\": [{\"mention_name\": {\"text\": \"income data\", \"confidence\": 0.5413410663604736, \"start\": 1915, \"end\": 1926, \"clean_text\": \"income data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1915, \"end\": 1926, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1915, \"end\": 1926, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1915, \"end\": 1926}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1915, \"end\": 1926, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.25216034054756165, \"start\": 351, \"end\": 362}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5413410663604736, \"start\": 351, \"end\": 362}, \"usage_context\": {\"text\": \"background\", \"confidence\": 0.46476203203201294, \"start\": 351, \"end\": 362}, \"is_used\": {\"text\": \"False\", \"confidence\": 0.46476203203201294, \"start\": 351, \"end\": 362}}, {\"mention_name\": {\"text\": \"PECS 2016/17 dataset\", \"confidence\": 0.9128821492195129, \"start\": 2642, \"end\": 2662, \"clean_text\": \"PECS 2016/17 dataset\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.36137863993644714, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2642, \"end\": 2662, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2642, \"end\": 2662}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2642, \"end\": 2662, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9019227027893066, \"start\": 301, \"end\": 321}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.9128821492195129, \"start\": 301, \"end\": 321}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5820990800857544, \"start\": 301, \"end\": 321}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5820990800857544, \"start\": 301, \"end\": 321}}, {\"mention_name\": {\"text\": \"administrative data\", \"confidence\": 0.8404825925827026, \"start\": 3200, \"end\": 3219, \"clean_text\": \"administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3200, \"end\": 3219, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3200, \"end\": 3219, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3200, \"end\": 3219}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3200, \"end\": 3219, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9894399046897888, \"start\": 121, \"end\": 140}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8404825925827026, \"start\": 121, \"end\": 140}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.7106826305389404, \"start\": 121, \"end\": 140}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7106826305389404, \"start\": 121, \"end\": 140}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [59]}}, {\"page\": 60, \"chunk\": 60, \"input_text\": \"municipalities seems very high and varies a lot for several reasons (commercial area, industrial etc.).\\n\\n- (iv) Land area information was not available to estimate the annual rental value of the property.\\n\\n- (v) Informality in the payment of property tax. To overcome this limitation, the model applied proportional downscaling to match the effective property tax rate in the administrative data (relative to consumption).\\n\\n# **9.2.2 Indirect taxes**\\n\\n# **The main limitations for the VAT model are:**\\n\\n- (i) We assume households purchase goods and services in the same region where they live.\\n\\n- (ii) There is no place of purchase variable nor a secondary study on consumption informality for the West Bank and Gaza.\\n\\n- (iii) We only applied indirect effects from VAT to West Bank residents.\\n\\nTo address the second limitation, we used estimations of consumption informality by decile based on an average from 10 similar countries (available in Bachas et al. 2020).\\n\\n**In the case of customs duties, the major challenge is that we do not have an “imported” dummy in the PECS (a limitation of most household budget surveys).** To address this limitation, we assume the “Law of One Price”—that is, that prices of domestic goods converge to prices of similar imported goods. In applying the Law of One Price for simulating customs duties, we could overestimate the incidence of this tax, particularly if there are price differences between similar domestic and imported products and if poor or rich households have different consumption patterns between domestic and imported products. However, sensitivity analysis conducted for other fiscal incidence studies shows that alternative methods are likely to yield similar results.<sup>79</sup>\\n\\n**Lastly, we assumed that Gaza residents are not paying indirect taxes.** We could be underestimating the tax burden for Gaza residents if they do pay informal taxes to the _de facto_ authorities, or if indirect effects of taxation from using taxed production inputs from the West Bank affect Gaza residents.\\n\\n# **9.2.3 Indirect subsidies**\\n\\n**The main limitation to the estimation of the electricity subsidy is lack of administrative data on “average cost of supply” of electricity by DISCOs to the domestic consumers.** Data is also not available on distribution margin of the DISCOs. Additionally, the PECS only provides the information about the “electricity charges” paid by the households. It does not provide the information about the “type of meter” households have. The application of “incremental block Tariff” is based on the type of meter. In the absence of this information, it is assumed that electricity tariff is volumetric.\\n\\n> 79 In a fiscal incidence analysis of Benin in 2018 the authors model customs using two different assumptions: in the baseline they assume that domestic and imported goods are differentiated enough (in terms of quality and type of good within IO sector) to prevent price convergence, while as a sensitivity analysis they assume that domestic and imported goods belong to the same IO subsector and are of similar levels of quality, such that the Law of One Price (LOOP) applies. (Goldman et al., 2022). They find little difference in the results of the two methods, with a slight increase in the Kakwani Index from 5.7 to 5.8 with the LOOP assumption, no change in the marginal contribution to inequality of 0.2, and a slightly greater marginal poverty increase decrease in the marginal contribution to poverty reduction from –1.9 to –2.0 (i.e., a larger poverty increase).\\n\\n58\", \"datasets\": [{\"mention_name\": {\"text\": \"administrative data\", \"confidence\": 0.6897786855697632, \"start\": 376, \"end\": 395, \"clean_text\": \"administrative data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 376, \"end\": 395, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 376, \"end\": 395, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 376, \"end\": 395}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 376, \"end\": 395, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.9941872358322144, \"start\": 376, \"end\": 395}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6897786855697632, \"start\": 376, \"end\": 395}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6057443022727966, \"start\": 376, \"end\": 395}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6057443022727966, \"start\": 376, \"end\": 395}}, {\"mention_name\": {\"text\": \"household budget surveys\", \"confidence\": 0.7597997188568115, \"start\": 1097, \"end\": 1121, \"clean_text\": \"household budget surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1097, \"end\": 1121, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1097, \"end\": 1121, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1097, \"end\": 1121}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1097, \"end\": 1121, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9303484559059143, \"start\": 376, \"end\": 400}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7597997188568115, \"start\": 376, \"end\": 400}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6690645217895508, \"start\": 376, \"end\": 400}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6690645217895508, \"start\": 376, \"end\": 400}}, {\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.41089531779289246, \"start\": 2345, \"end\": 2349, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.5164163112640381, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2345, \"end\": 2349, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2345, \"end\": 2349}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2345, \"end\": 2349, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"report\", \"confidence\": 0.32431307435035706, \"start\": 865, \"end\": 869}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.41089531779289246, \"start\": 865, \"end\": 869}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5015591382980347, \"start\": 865, \"end\": 869}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5015591382980347, \"start\": 865, \"end\": 869}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [60]}}, {\"page\": 61, \"chunk\": 61, \"input_text\": \"# **9.3 Direct transfers**\\n\\n**The PECS only provides aggregate data of households’ income from social protection (e.g., government assistance).** This includes one variable for “total cash” and another variable for “total in kind”, the latter labeled as near-cash transfers in our model. Hence, it was not possible to disaggregate transfers households received for each of the four main specific social protection (cash transfer) programs in the country. Currently, the model infers that the cash-transfers reported in the PECS are from the CTP program based on cross-validation with other administrative data. However, if the team had access to the PMT formula of the CTP and threshold or the allocation methods for other programs, it would have been possible to do a more detailed simulation of social protection programs in the West Bank and Gaza.\\n\\n## **9.3.1 Education**\\n\\n**Data limitations in the PECS 2016/17 required that we made assumptions about beneficiaries of public education benefits.** The PECS 2016/17 survey does not have information on “current grade of education”; instead, this study inferred beneficiaries combining self-reported information on “highest grade of education achieved (plus one year)” and “currently enrolled in public education”. The analysis also imputes an average education benefit calculated at the national level; implicitly, this assumes that students in the West Bank and Gaza receive the same benefits across regions at each education level. Information on public expenditure by education level, disaggregated by region, would improve the calculation of education benefits. Lastly, due to PECS data limitations, it was not possible to analyze distributional effects of public expenditure from other levels of education (pre-primary and tertiary).\\n\\n## **9.3.2 Health**\\n\\n**The PECS does not have information on access to public health services.** By using the health-insurance approach (focusing on individuals reporting coverage of government health insurance in the PECS), the model implicitly assumes that eligible individuals do not have significant barriers to access public health services. However, in practice, usage of health services by the eligible population could be lower, particularly in Gaza where most residents use UNRWA-provided health services. Hence, using the healthinsurance approach could overestimate the in-kind health benefits among Gaza residents. We also calculated public health benefits at the national level, assuming that benefits for each health level are the same for residents in the West Bank and Gaza. However, having public health expenditure data disaggregated by health service and region would improve calculation of health benefits.\\n\\n59\", \"datasets\": [{\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.3848762810230255, \"start\": 34, \"end\": 38, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.3438805341720581, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 34, \"end\": 38, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 34, \"end\": 38}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 34, \"end\": 38, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.43282654881477356, \"start\": 34, \"end\": 38}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3848762810230255, \"start\": 34, \"end\": 38}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7388270497322083, \"start\": 34, \"end\": 38}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7388270497322083, \"start\": 34, \"end\": 38}}, {\"mention_name\": {\"text\": \"PECS 2016/17\", \"confidence\": 0.6368517875671387, \"start\": 902, \"end\": 914, \"clean_text\": \"PECS 2016/17\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 902, \"end\": 914, \"clean_text\": \"\"}, \"producer\": {\"text\": \"PECS\", \"confidence\": 0.3166857957839966, \"start\": 902, \"end\": 906, \"clean_text\": \"PECS\"}, \"reference_year\": {\"text\": \"2016/17\", \"confidence\": 0.36592644453048706, \"start\": 1010, \"end\": 1017}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 902, \"end\": 914, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.4791812300682068, \"start\": 138, \"end\": 150}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6368517875671387, \"start\": 138, \"end\": 150}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.540048360824585, \"start\": 138, \"end\": 150}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.540048360824585, \"start\": 138, \"end\": 150}}, {\"mention_name\": {\"text\": \"PECS 2016/17 survey\", \"confidence\": 0.6750066876411438, \"start\": 1005, \"end\": 1024, \"clean_text\": \"PECS 2016/17 survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1005, \"end\": 1024, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1005, \"end\": 1024, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1005, \"end\": 1024}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1005, \"end\": 1024, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9942770600318909, \"start\": 241, \"end\": 260}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6750066876411438, \"start\": 241, \"end\": 260}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.6590123176574707, \"start\": 241, \"end\": 260}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6590123176574707, \"start\": 241, \"end\": 260}}, {\"mention_name\": {\"text\": \"public health expenditure data\", \"confidence\": 0.6866605877876282, \"start\": 2598, \"end\": 2628, \"clean_text\": \"public health expenditure data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2598, \"end\": 2628, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2598, \"end\": 2628, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2598, \"end\": 2628}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2598, \"end\": 2628, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.1949291229248047, \"start\": 1033, \"end\": 1063}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6866605877876282, \"start\": 1033, \"end\": 1063}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.500217080116272, \"start\": 1033, \"end\": 1063}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.500217080116272, \"start\": 1033, \"end\": 1063}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [61]}}, {\"page\": 62, \"chunk\": 62, \"input_text\": \"# **10 Conclusion and next steps**\\n\\n## **10.1 Conclusion**\\n\\n**The fiscal incidence analysis for the West Bank and Gaza in 2017 shows that the combination of taxes and transfers modeled reduced inequality but increased poverty.** Inequality decreased by 6.5 Gini points, from 35.0 at Market Income to 28.5 at Final Income), while the national poverty headcount increased by 8.4 percentage points, from 30.6 percent at Market Income to 39.0 percent at Consumable Income. This pattern is common in developing countries where the tax collection is mostly comprised of indirect taxes and social protection coverage is limited.\\n\\n**The large increase in poverty due to combined taxes and transfers suggests that fiscal redistribution in the West Bank and Gaza could be improved.** Only the poorest decile is a net-receiver of the combination of taxes and transfers modeled in this study, all other deciles were net payers. Almost 50 percent of the poor population at Consumable Income (roughly 19 percent of the total population) experienced some degree of fiscal impoverishment—that is, a reduction of their Market Income after they paid taxes and received transfers.\\n\\n**Most fiscal interventions were progressive relative to Market Income.** We measured this using the Kakwani Index. The regressive interventions, looking at Market Income, were fuel subsidies, pension income, and property tax.\\n\\n**The marginal benefit of including in-kind benefits from health and education into the system accounts for 40 percent of the reduction in inequality, while the marginal benefit of including the CTP into the system accounts for 18 percent of the inequality reduction.**<sup>80</sup> We measured the marginal contributions of each instrument to either reducing inequality or reducing poverty **.** The large marginal contributions to inequality reduction stemming from public basic education is expected since poor households have more children that access basic public schools. The model could, however, overestimate the inequality reduction stemming from in-kind public health benefits considering that benefits were imputed based on access to public-health insurance rather than actual usage of public health services (not available in the PECS 2016/17). Yet human capital outcomes as measured by the SES-Disaggregated Human Capital Index are relatively even across socioeconomic quintiles compared to data for other countries,<sup>81</sup> suggesting that public spending on education and healthcare has been effective at reducing gaps in quality.\\n\\n**The VAT and customs duties are important tools for domestic revenue mobilization, but they represent the main drivers of poverty increase in the West Bank and Gaza.** While VAT and customs duties were progressive overall, the large burden they cause explains why they represent the largest marginal contributions to the national increase in poverty. Pension contributions and the PIT also increase poverty, although to a lesser extent due to their smaller size.\\n\\n**Strengthening the social protection transfer system could improve poverty and inequality reduction.** Cash transfers from social protection (CTP) and pension income contribute most to poverty reduction in\\n\\n> 80 The marginal contribution of the in-kind benefits at Consumable income is 2.9 Gini points, while the CTP has a marginal contribution at Consumable income of 1.2 Gini points. 81 Avitabile et al (2020).\\n\\n60\", \"datasets\": [{\"mention_name\": {\"text\": \"Kakwani Index\", \"confidence\": 0.8445246815681458, \"start\": 1264, \"end\": 1277, \"clean_text\": \"Kakwani Index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1264, \"end\": 1277, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1264, \"end\": 1277, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1264, \"end\": 1277}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1264, \"end\": 1277, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.998860239982605, \"start\": 525, \"end\": 538}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8445246815681458, \"start\": 525, \"end\": 538}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9950154423713684, \"start\": 525, \"end\": 538}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9950154423713684, \"start\": 525, \"end\": 538}}, {\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.31000927090644836, \"start\": 2233, \"end\": 2237, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2233, \"end\": 2237, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2233, \"end\": 2237, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2233, \"end\": 2237}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2233, \"end\": 2237, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.5792490839958191, \"start\": 664, \"end\": 668}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.31000927090644836, \"start\": 664, \"end\": 668}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7726215720176697, \"start\": 664, \"end\": 668}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7726215720176697, \"start\": 664, \"end\": 668}}, {\"mention_name\": {\"text\": \"SES-Disaggregated Human Capital Index\", \"confidence\": 0.7427325248718262, \"start\": 2294, \"end\": 2331, \"clean_text\": \"SES-Disaggregated Human Capital Index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2294, \"end\": 2331, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2294, \"end\": 2331, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2294, \"end\": 2331}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2294, \"end\": 2331, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.6797555685043335, \"start\": 725, \"end\": 762}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7427325248718262, \"start\": 725, \"end\": 762}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.662284791469574, \"start\": 725, \"end\": 762}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.662284791469574, \"start\": 725, \"end\": 762}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [62]}}, {\"page\": 63, \"chunk\": 63, \"input_text\": \"the West Bank and Gaza, while contributing to inequality reduction. However, their coverage is still limited in deciles 2 and 3. Improving social protection coverage among deciles 2 and 3 to compensate for taxes they pay could reduce inequality and poverty in the West Bank and Gaza. Using targeted programs such as the flagship CTP (targeted through the PMT) could represent the most effective way to redistribute resources in the country.\\n\\n# **10.2 Next steps**\\n\\n**Future extensions of the current study should aim to reassess the incidence of public health benefits based on actual access to health services.** This could be done using secondary household surveys such as the Demographic and Health Survey or the next PECS, for which data are being collected in 2023.\\n\\n**Future study should also aim to include the simulation of the flagship and targeted CTP.** In the PECS 2016/2017, social protection transfers are aggregated as government cash assistance versus government near-cash assistance, making it impossible to disaggregate the impacts of individual social programs. The current analysis infers that the CTP is the reported cash assistance program in the PECS. The PECS 2023 will have more disaggregated data of social protection programs to refine this analysis. On the other hand, having access to the PMT targeting formula of the CTP would allow to perform other policy-reform simulations, such as expansions of coverage of the CTP beyond that reported in the PECS. This could be useful to analyze the impact of social protection compensation measures.\\n\\n**Finally, the fiscal incidence model we developed for the West Bank and Gaza could be used to inform likely distributional effects for new PA policies and reforms.** Examples of these reforms include adjusting the electricity tariff structure to increase progressivity and decrease costs to the poor, reducing the fuel subsidy to increase fiscal space, updating PIT tax brackets, and expanding the number of municipalities applying property tax. Furthermore, the PECS 2023 survey has been updated to address some of the limitations highlighted in this analysis. For example, the PECS 2023 is better able to disaggregate social protection programs’ coverage and payments to, and use of, different healthcare facilities.\\n\\n61\", \"datasets\": [{\"mention_name\": {\"text\": \"Demographic and Health Survey\", \"confidence\": 0.918981671333313, \"start\": 679, \"end\": 708, \"clean_text\": \"Demographic and Health Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 679, \"end\": 708, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 679, \"end\": 708, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 679, \"end\": 708}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 679, \"end\": 708, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9900029897689819, \"start\": 679, \"end\": 708}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.918981671333313, \"start\": 679, \"end\": 708}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.7821693420410156, \"start\": 679, \"end\": 708}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7821693420410156, \"start\": 679, \"end\": 708}}, {\"mention_name\": {\"text\": \"PECS\", \"confidence\": 0.745543360710144, \"start\": 721, \"end\": 725, \"clean_text\": \"PECS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 721, \"end\": 725, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 721, \"end\": 725, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 721, \"end\": 725}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 721, \"end\": 725, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9900029897689819, \"start\": 721, \"end\": 725}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.745543360710144, \"start\": 721, \"end\": 725}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.7821693420410156, \"start\": 721, \"end\": 725}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7821693420410156, \"start\": 721, \"end\": 725}}, {\"mention_name\": {\"text\": \"disaggregated data of social protection programs\", \"confidence\": 0.41766268014907837, \"start\": 1204, \"end\": 1252, \"clean_text\": \"disaggregated data of social protection programs\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1204, \"end\": 1252, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1204, \"end\": 1252, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1204, \"end\": 1252}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1204, \"end\": 1252, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"microdata\", \"confidence\": 0.45327308773994446, \"start\": 417, \"end\": 465}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.41766268014907837, \"start\": 417, \"end\": 465}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.7210237383842468, \"start\": 417, \"end\": 465}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7210237383842468, \"start\": 417, \"end\": 465}}, {\"mention_name\": {\"text\": \"PECS 2023 survey\", \"confidence\": 0.9529137015342712, \"start\": 2035, \"end\": 2051, \"clean_text\": \"PECS 2023 survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2035, \"end\": 2051, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2035, \"end\": 2051, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2035, \"end\": 2051}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2035, \"end\": 2051, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.999921441078186, \"start\": 391, \"end\": 407}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.9529137015342712, \"start\": 391, \"end\": 407}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.578127920627594, \"start\": 391, \"end\": 407}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.578127920627594, \"start\": 391, \"end\": 407}}, {\"mention_name\": {\"text\": \"PECS 2023\", \"confidence\": 0.5489239692687988, \"start\": 2151, \"end\": 2160, \"clean_text\": \"PECS 2023\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2151, \"end\": 2160, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2151, \"end\": 2160, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2151, \"end\": 2160}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2151, \"end\": 2160, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"administrative\", \"confidence\": 0.5197623372077942, \"start\": 507, \"end\": 516}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5489239692687988, \"start\": 507, \"end\": 516}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4746275842189789, \"start\": 507, \"end\": 516}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4746275842189789, \"start\": 507, \"end\": 516}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099646401302432923/pdf/IDU11fba1aab1fb111482e1a1441122a743dfde3.pdf\", \"pages\": [63]}}]}"}}},{"rowIdx":92,"cells":{"doc_id":{"kind":"string","value":"34249073"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34249073\", \"last_modified_date\": \"2024-04-14T00:00:00Z\", \"authors\": {\"0\": {\"author\": \"Bou Habib,Chadi\"}}, \"count\": \"Lebanon\", \"count_key\": \"82571\", \"docna\": {\"0\": {\"docna\": \"Globalization, Dutch Disease, and\\n Vulnerability to External Shocks in a Small Open Economy :\\n The Case of Lebanon in 1916 and 2019\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"EFI-AFR1-MTI-MacroFiscal-1 (EAEM1)\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34249073\"}, \"repnb\": \"WPS10688\", \"docdt\": \"2024-02-01T00:00:00Z\", \"datestored\": \"2024-01-31T01:24:37Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"This paper investigates the similarities\\n between the economy of 1912 Mount Lebanon on the eve of the\\n famine of 1916 and the economy of 2004 Lebanon that set the\\n stage for the major economic and social crisis of 2019. A\\n simple general equilibrium simulation shows that, as long as\\n the Lebanese economy remains reliant on foreign inflows,\\n crises will persist, with different manifestations.\\n Regardless of the period considered, foreign inflows\\n increase domestic prices and induce real appreciation. Low\\n productive capacities and insufficient job creation lead to\\n high emigration. Emigration increases the reliance on\\n foreign inflows, which in turn increase domestic prices and\\n reduce competitiveness, hence triggering further emigration\\n and further reliance on foreign inflows. Income and prices\\n increase, but exports decline, and growth remains volatile.\\n The interruption of the flows of capital and goods and the\\n impossibility to migrate due to the First World War drove\\n Lebanon into starvation in 1916. The interruption of inflows\\n of capital in 2019 led to a major crisis and massive\\n outmigration, as predicted through the simulations based on\\n the structure of the Lebanese economy in 2004. The\\n simulations effectively capture the impact of external\\n shocks on the Lebanese economy and closely align with the\\n actual changes in economic variables during 2005 to 2020.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10688; PROSPERITY\", \"display_title\": \"Globalization, Dutch Disease, and\\n Vulnerability to External Shocks in a Small Open Economy :\\n The Case of Lebanon in 1916 and 2019\", \"disclosure_date\": \"2024-01-31T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099324301312410845/text/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.txt\", \"listing_relative_url\": \"/research/2024/02/34249073/globalization-dutch-disease-vulnerability-external-shocks-small-open-economy-case-lebanon-1916-2019\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099324301312410845/IDU16cf9281c16b28145b51b4fa17b1af694b7a9\", \"new_url\": \"2024/02/34249073/IDU16cf9281c16b28145b51b4fa17b1af694b7a9\", \"ext_pub_date\": \"2024-01-31T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"International Remittances m1326998 355,Exchange Rate Policy m1327143 1124,Jobs and Globalization m1503937 2682,Emigration m1504169 2844,Computable General Equilibrium Modeling\\n m1504533 3110\", \"docm_id\": \"IDU-6cf9281c-6b28-45b5-b4fa-7b1af694b7a9\", \"chronical_docm_id\": \"IDU-6cf9281c-6b28-45b5-b4fa-7b1af694b7a9\", \"publishtoextweb_dt\": \"2024-01-31T00:00:00Z\", \"dois\": \"10.1596/41007\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"guid\": \"099324301312410845\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"18\", \"author_keys\": {\"0\": {\"author_key\": \"668562\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099324301312410845/IDU16cf9281c16b28145b51b4fa17b1af694b7a9\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10688 Globalization, Dutch Disease, and Vulnerability to External Shocks in a Small Open Economy The Case of Lebanon in 1916 and 2019 _Chadi Bou Habib_ Macroeconomics, Trade and Investment Global Practice January 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10688\\n\\n# **Abstract**\\n\\nThis paper investigates the similarities between the economy of 1912 Mount Lebanon on the eve of the famine of 1916 and the economy of 2004 Lebanon that set the stage for the major economic and social crisis of 2019. A simple general equilibrium simulation shows that, as long as the Lebanese economy remains reliant on foreign inflows, crises will persist, with different manifestations. Regardless of the period considered, foreign inflows increase domestic prices and induce real appreciation. Low productive capacities and insufficient job creation lead to high emigration. Emigration increases the reliance on foreign inflows, which in turn increase domestic prices and reduce competitiveness, hence triggering further emigration and further reliance on foreign inflows. Income and prices increase, but exports decline, and growth remains volatile. The interruption of the flows of capital and goods and the impossibility to migrate due to the First World War drove Lebanon into starvation in 1916. The interruption of inflows of capital in 2019 led to a major crisis and massive outmigration, as predicted through the simulations based on the structure of the Lebanese economy in 2004. The simulations effectively capture the impact of external shocks on the Lebanese economy and closely align with the actual changes in economic variables during 2005 to 2020.\\n\\nThis paper is a product of the Macroeconomics, Trade and Investment Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The author may be contacted at cbouhabib@worldbank.org.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# **Globalization, Dutch Disease, and Vulnerability to External Shocks in a Small Open Economy: The Case of Lebanon in 1916 and 2019**\\n\\nChadi Bou Habib<sup>1</sup> JEL Classification: D58, F6, F22, F24, G01.\\n\\nKeywords: General Equilibrium, Globalization, Migration, Remittances, Real Exchange Rate.\\n\\n1 Lead Economist at the World Bank. The paper carries the name of the author and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. This paper reproduces elements from an initial work for a thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts to the Department of Economics of the Faculty of Arts and Sciences at the American University of Beirut on September 24, 2007. See: _The effects of capital inflows on a small open economy growth or Dutch disease? a study of the Lebanese case_ , Bou Habib, Chadi Adib. American University of Beirut, 2007. <u>https://scholarworks.aub.edu.lb/handle/10938/7549 I am grateful for Dr Samir Makdisi (Advisor), Dr Simon Neaime</u> (Member of Committee), and Dr Markus Marktanner (Member of Committee) for their guidance during the preparation of the thesis in 2007. The current paper written in 2023 reinserts, revises, expands, and updates the work from 2007. This version benefited from the valuable advice from Emilia Skrok, and Eric Le Borgne, Practice Managers at the World Bank, and from insights by John François, Economist at the World Bank.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **Introduction**\\n\\nFrom the end of the 19th century to the beginning of the 21st century, Lebanon has managed to attract large amounts of resources that are disconnected from the country’s productive system and its internal accumulation of capital. This was possible thanks to three major characteristics: a large Lebanese diaspora, a comparative educational advantage of human resources, and an efficient banking sector. Transfers of wealth from the diaspora were constantly high, but what changed is their nature. These transfers in the early years were mainly remittances sent to relatives or invested in housing or in different sectors of the economy. They were first amplified by capital inflows from the region and gradually took the form of deposits in the banking sector.\\n\\nSeeking to attract capital that fled Palestine in 1948 and other Arab countries that experienced waves of nationalization, Lebanon implemented a law on bank secrecy in the 1950s and established the legal foundations for a modern banking system. These inflows helped to create the basic infrastructure on which Lebanon built its productive capacities in agriculture, industry, tourism, and services in the 1950s and 1960s. At the end of that period, the oil boom reached its height with multiple increases in oil prices starting 1973. In parallel, and thanks to skills acquired through an efficient education system set by missionaries in the 19th century, Lebanese labor massively moved to the Gulf countries where they were employed as engineers, specialists, and skilled workers. These emigrants have naturally transferred an important share of their wealth accumulated in the Gulf into Lebanon and its banking sector.\\n\\n# **A. Lebanon: A History of Foreign Inflows**\\n\\n“The wealth and luxury you see in this Mountain do not come from its land, but from the money saved by the sons of Lebanon in the emigration.” This sentence from the study ordered by the Turkish governor of Mount Lebanon in 1918, ( _Lubnan, mabaheth ilmiah wa ijtimaya – Lebanon, scientific and social studies – 1918_ ) remains true more than a century later.\\n\\nThis part of the Middle East was prematurely exposed to the international flows of people, goods, and capital during the first globalization. Statistics from 1912 show that 25 to 30 percent of the population of Mount Lebanon was living and working abroad. The reasons for this emigration, according to contemporaneous sources, are the repeated crisis in an economy that was heavily relying on the production and export of raw natural silk. Another reason was the high concentration of the population in Mount Lebanon with a density of 110 individuals per square kilometer, which was high compared to the international standards of that period.\\n\\n# **_1. The Silk Curse: Monoculture, Deindustrialization and Emigration_**\\n\\nSince the first quarter of the 19th century and the implementation of the first factories for the treatment of raw silk, Lebanese agriculture has gradually been monopolized by mulberry culture to provide mulberry leaves to silkworm breeding plants. The low cost of a large labor force concentrated in regions close to the coast and ports was the main reason that motivated foreign investors to develop silk industry in Mount Lebanon.<sup>2</sup> This exporting activity raised the revenues of Mount Lebanon and triggered an enhancement of the standard of living together with substantial\\n\\n> 2 Mount-Lebanon, or the Mount Lebanon Mutessarifat (1861-1915) covered one-third of the area of modern Lebanon and was a semi-autonomous district of the Ottoman Empire created in 1861 as a consequence of the civil war of 1860 between Christian Maronites and Muslim Druses.\\n\\n2\", \"datasets\": [{\"mention_name\": {\"text\": \"Statistics from 1912\", \"confidence\": 0.7530025839805603, \"start\": 2255, \"end\": 2275, \"clean_text\": \"Statistics from 1912\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2255, \"end\": 2275, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2255, \"end\": 2275, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2255, \"end\": 2275}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2255, \"end\": 2275, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.9919923543930054, \"start\": 537, \"end\": 557}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.7530025839805603, \"start\": 537, \"end\": 557}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6969510912895203, \"start\": 537, \"end\": 557}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6969510912895203, \"start\": 537, \"end\": 557}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"|Year|Price|Year|Price|Year|Price|\\n|---|---|---|---|---|---|\\n|1861|73.5|1882|55.5|1901|40.5|\\n|1862|71|1880|45|1903|46.6|\\n|1864|73.5|1890|52|1904|40.5|\\n|1870|93|1892|44|1908|40|\\n|1872|105|1894|37|1909|40|\\n|1875|57|1896|34|1910|40.5|\\n|1881|58|1900|38|||\\n\\n<!-- Start of picture text --> 25,000- \\\\<br>20,000<br>|<br>15,000 |<br>ak V<br>5,000 fo<br>5 5 S 8 i z & 8 : z z<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> N<br>q “hy<br>7eZ |<br>ec<br>185 1861 187( TM 1890 “ “ | 1905 1910 1914<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"<!-- Start of picture text --> Other<br>Remittances<br>Silk |<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"of GDP. Despite the lack of accuracy of these statistics, they hint to the size transfers had reached in the small economy of Mount Lebanon. Such a level of transfers had deep structural and permanent impact on the economy.\\n\\n# **_3. The Impact of Transfers on the Economy of Mount Lebanon_**\\n\\nScholars from the beginning of the 20th century acknowledged the positive impact of transfers on household income and welfare. But they complained about the misuse of this money, pointing to an uncontrolled real estate development with an “excessive inflation of the prices of land.” This statement is simply indicating the occurrence of a price increase in the non-tradable real-estate sector due to a strong spending effect caused by large transfers into the domestic economy. With land being an input to the productive activity, the large increase in its prices had a harmful impact on agriculture and on many industrial activities. Also, the booming real-estate sector has initiated further adjustments with resources moving into construction activities. Scholars talk about some technology transfers to agriculture and industry by returning migrants, but they present their effect as being marginal and limited. In sum, transfers have reinforced the deindustrialization already experienced by Mount Lebanon due to the mono-industry of raw silk.\\n\\nTable 3. Imports, exports, and transfers of Mount Lebanon\\n\\n|Item – in Ottoman Pound|Imports & outflows|Exports & inflows|Balance|\\n|---|---|---|---|\\n|Food and livestock|760,000|78,000|-682,000|\\n|Tobacco,beverage & olive oil|80,000|198,000|118,000|\\n|Textile and silk|63,000|687,000|624,000|\\n|Manufacturedgoods|650,000|85,000|-565,000|\\n|Total|1,553,000|1,048,000|-505,000|\\n|Tourism Income/Expenditures|0|15,000|15,000|\\n|Transfers from/to America|300,000|1,200,000|900,000|\\n|Current Account Balance|||410,000|\\n\\n_Sources: Lubnan, mabaheth ilmiah wa ijtimaya - Lebanon, scientific and social studies - 1918._ The eviction of tradable sectors became quasi total by 1910-1914 and while silk reached 61 percent of total exports, Mount Lebanon became a net importer of tradable goods financed by Net Transfers from the Americas. Table 3 summarizes the imports and exports of Mount Lebanon for an unspecified year between 1910 and 1914, most likely 1912. It shows a trade deficit equivalent to 30 percent of Mount Lebanon’s GDP and a ratio of Current Account surplus to GDP of 24.5 percent. Transfers and silk exports are clearly the main sources of revenues fueling the growing demand in Mount Lebanon that is met through increase in the imports of goods. For instance, statistics show that imports of food and beverage were three times larger than the exports of these same items. Also, imports of manufactured goods from Europe are almost eight times higher than exports of manufactured items other than textile and silk. This dependency on mono-exports and on transfers to finance imports of food and manufactured goods would have disastrous consequences for Mount Lebanon in times of economic or political crisis.\\n\\n5\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"# **B. Confronting History and Theory: A General Equilibrium Analysis of Dutch Disease in Mount Lebanon of 1912**\\n\\nWith the mono-export of silk and net transfers from the Americas accounting respectively for 41 and 54 percent of GDP, the economy of Mount Lebanon was both particularly globalized and highly vulnerable to external shocks. The data set available for the year 1912 is suitable for a simple general equilibrium analysis.\\n\\n# **_1. Dependence and Vulnerability to External Shocks: The Case of Mount Lebanon through a Simple General Equilibrium Framework_**\\n\\nIn the following, I use the simple general equilibrium framework of the 1 2 3 Model developed by Devarajan et al ( _Simple general equilibrium modeling – 1997 & The simplest general equilibrium model of an open economy - 1998_ ) to assess the effect of external shocks on the economy of Mount Lebanon.\\n\\nThe 1 2 3 Model builds on the Harrod-Samuelson-Balassa two-sector (tradable and non-tradable sectors) model to examine the effects of remittances, inflows, and external shocks. The model involves one country, two producing sectors and three goods. The country has a small open economy and is a price taker on the international market. The two producing sectors are (i) one sector that produces goods for export only and (ii) another sector producing goods for the domestic market only. The three goods are (a) export goods, (b) goods produced and consumed domestically (mainly non-tradable goods), and (c) imported goods. Remittances, export prices and import prices are set as exogenous variables. The model does not take factor markets into consideration. In this model, the RER is weighted by the shares of imports and exports in the total foreign trade of the economy, _PD M PD E_ according to the following equation: _RER_ = + where PD is the price of domestic _PM FT PE FT_\\n\\n> goods, PM the price of imported goods, PE the price of exported goods, FT foreign trade, M\\n\\n> imports and E exports _._ In this equation, real appreciation is equivalent to an increase.\\n\\nThe model fits the case of Mount Lebanon because: (i) Mount Lebanon was a small open economy; (ii) it had a large industry producing a single good dedicated for exports (silk) and (iii) relied on imports for almost all other types of goods, including food. The simulation aims to: (i) assess the impact of external shocks on GDP, income, and prices and, (ii) test the relation between remittances and exports. While the model assesses the spending effect of foreign inflows, it cannot assess the impact in terms of resource movement since it neglects factor markets.\\n\\n# **_2. A Simple General Equilibrium Analysis for Mount Lebanon: Simulation and Results_**\\n\\nThe analysis below is done through four scenarios: a drop by 90 percent in remittances, an increase by 10 percent in remittances, a decline by 10 percent in the prices of exports, and a compensation for the effect of that decline on consumption through a rise in remittances. The four scenarios are designed in a way both to confront theory with history and to understand the latter through the former. Table 4 summarizes the results of all four scenarios. Prior to the simulations, I have brought the current account surplus to zero by increasing imports in resources and raising consumption and investment in uses. This is equivalent to assuming that Mount Lebanon did not build up reserves and used all the funds available. This modification neutralizes the effects, however limited, of a large 6\", \"datasets\": [{\"mention_name\": {\"text\": \"data set available for the year 1912\", \"confidence\": 0.6025062203407288, \"start\": 342, \"end\": 378, \"clean_text\": \"data set available for the year 1912\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 342, \"end\": 378, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 342, \"end\": 378, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 342, \"end\": 378}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 342, \"end\": 378, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6495075225830078, \"start\": 342, \"end\": 378}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6025062203407288, \"start\": 342, \"end\": 378}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7248286008834839, \"start\": 342, \"end\": 378}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7248286008834839, \"start\": 342, \"end\": 378}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"current account surplus (large negative foreign savings) on the results of the simulation. Also, this modification is not far from the reality since Mount Lebanon back then did not have a monetary authority.\\n\\nTable 4. Simulation of external shocks on the economy of Mount Lebanon in 1912 (change, in\\n\\n|Percentage change compared to base year|percent)<br>Remitt.<br>-90%|Remitt.<br>+10%|Prices of<br>exports -<br>10%|Prices of exports -<br>10% and<br>consumption per<br>capita unchanged|\\n|---|---|---|---|---|\\n|Remittances|-90.0|10.0|0.0|16.1|\\n|Total Income|-39.7|4.5|-6.4|0.9|\\n|Private Consumption|-33.7|3.6|-4.3|1.6|\\n|Private Consumption/capita|-34.7|2.0|-5.7|0.0|\\n|Investment|-31.3|3.3|-4.0|1.5|\\n|Exports|7.9|-0.9|-0.1|-1.7|\\n|Imports|-37.0|4.1|-5.5|1.2|\\n|Demand of domesticgoods|-17.6|1.6|0.3|2.9|\\n|Price of domesticgood|-36.1|4.2|-9.4|-2.9|\\n|Real exchange rate|-36.1|4.2|-5.9|0.9|\\n|Real GDP at marketprice|-2.6|0.0|-0.1|0.0|\\n\\nN.B: Per capita consumption is calculated taking into consideration an estimated 3 percent yearly natural increase in population. In the last scenario, I add an emigration rate of 1.4 percent. These assumptions are set according to tendencies observed during 1900-12. Note: Remitt. Stands for remittances. Source: author simulations.\\n\\nThe first column summarizes the effect of a decline in remittances by 90 percent. The impact on income is strong and generates a sharp decline in consumption and investment. This is consistent with historical developments: when remittances inflow to Mount Lebanon halted in 1914-18 because of WWI, the country fell into a disastrous famine. Imports, including food imports, collapsed. The RER decreases strongly, due to a collapse in the demand and prices of domestic goods. The decline in the RER boosts exports by 8 percent. The increase in exports keeps the decline in real GDP moderate at a negative 2.6 percent. In reality, history shows that exports stopped in 1914-18 because WWI interrupted all trade flows.\\n\\nThe second column simulates the effect of a 10 percent increase in remittances. Income, consumption, and imports increase while real GDP remains unchanged. The country loses in competitiveness since RER appreciates due to higher demand and prices of domestic goods, and exports decline. This result meets the observations on the increase in the prices of land, a key nontradable good, with the rise in remittances to Mount Lebanon.\\n\\nThe third column assesses the impact of a decline by 10 percent in the prices of exports. Income declines and triggers a decrease in consumption. Imports also decrease since there is less foreign financing available due to the drop in exports. Real GDP regresses slowly while RER declines due to lower prices of domestic goods. The decline in the RER is however less marked than the drop in the prices of domestic goods since the relative price of domestic non-traded goods to traded goods declines moderately. Indeed, in the denominator, part of the prices of traded goods (exports) has also declined. Also in this case, history reports on the negative impact of a decline in the international prices of silk on income and consumption.\\n\\n7\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"The last column simulates the increase needed in remittances to compensate for the impact of a 10 percent decline in the price of exports on private per capita consumption. The assumption here is that in the face of a drop in income and consumption, part of the population would emigrate while the diaspora would increase transfers to the motherland. Interestingly, to compensate for the impact of a 10 percent drop in the prices of exports on private per capita consumption, Mount Lebanon needed to receive remittances higher by 16 percent. The combined effect of the decline in exports and rise in remittances is higher income, higher imports, and higher demand for domestic goods while real GDP remains unchanged. The prices of domestic goods remain on the decline due to a combination of a high deflationary effect stemming from the decline in exports prices and of a more moderate inflationary impact related to higher remittances. The RER increases despite the decline in domestic prices since, in the denominator, the decline in the prices of a part of the traded goods (exports) is strong, and the relative price of domestic non-traded goods to traded goods increases. In line with the appreciation of the RER, exports decline; pointing to a loss in competitiveness. In short, an important decline in the price of exports on the international market would reduce the income of the population and increase the reliance on remittances which in turn has a negative impact on competitiveness and depresses exports.\\n\\n# **C. Back to the Future: A Simple General Equilibrium Analysis of Dutch Disease in Modern Lebanon**\\n\\nAfter the First World War, Lebanon was extended beyond Mount Lebanon to its current borders to include larger farming lands. The famine had claimed the lives of one-third of the population and, while emigration continued, it was much lower than the pre-war levels. Between 1920 and 1950, the country developed a more diversified economy and abandoned the mono-industry model of the silk era. Remittances continued to play an important role, but less than in 1880-1914. In the late 1950s, social and economic transformations drove growing numbers of the rural population out of their areas. Some of them emigrated, but the largest part settled in cities. The second generation of this new urban population had access to better education and was attracted by the booming economies of the oil rich countries of the Gulf, starting in the early 1970s. Since that period, remittances and foreign inflows have become once again a major element of GNDI. The war of 1975-1990 triggered a large outmigration and increased the reliance on foreign inflows, since large segments of the economy were severely damaged due to the direct and indirect consequences of the war. In the following paragraphs, I will be considering foreign inflows as a whole, since the development of the banking sector reduced the boundaries between remittances and capital inflows, making it possible to examine the economic impact of the two variables combined.\\n\\n# **_1. Reliance on Inflows and Transfers in Modern Lebanon: A View through National Accounts_**\\n\\nThe National Accounts of Lebanon reflect the strong impact of the war on the Lebanese economy in terms of reduction in productive capacities and structural reliance on foreign inflows. Through extensive destruction of capital, the war reduced the productive capacities of the economy and increased the reliance on imports financed by substantial flows of income, transfers, and capital. Table 5 compares the structures of GDP and external accounts between the pre-war period (1972), the aftermath of the war (1994) and 2004. I chose 2004 because it is the year when the post-war reconstruction era ended. The champion of that era, former PM Rafik Hariri was assassinated in February 2005. Since then, no other economic and political project was proposed to the Lebanese society, and the rulers focused on extending the life of the post-reconstruction system they inherited from Mr. Hariri, until the system collapsed by end 2019.\\n\\n8\", \"datasets\": [{\"mention_name\": {\"text\": \"National Accounts of Lebanon\", \"confidence\": 0.3614228367805481, \"start\": 3153, \"end\": 3181, \"clean_text\": \"National Accounts of Lebanon\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3153, \"end\": 3181, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3153, \"end\": 3181, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3153, \"end\": 3181}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3153, \"end\": 3181, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.375042587518692, \"start\": 664, \"end\": 692}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3614228367805481, \"start\": 664, \"end\": 692}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6906523704528809, \"start\": 664, \"end\": 692}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6906523704528809, \"start\": 664, \"end\": 692}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"Table 5. GDP, GNDI and external accounts in Lebanon (in percent of GDP)\\n\\n||**Share to GDP (%)**|**1972**|**1994**|**2004**|\\n|---|---|---|---|---|\\n||Agriculture|9.9|12.1|5.2|\\n||Transport and communication|7.5|2.8|7.4|\\n|**ces**|Industry|15.9|17.7|11.9|\\n|**our**|Trade|31.5|28.8|24.0|\\n|**Res**|Construction|4.6|9.5|7.4|\\n||Public sector|7.5|8.3|11.2|\\n||Services and housing|23.1|20.9|32.9|\\n||**GDP (uses = resources)**|**100.0**|**100.0**|**100.0**|\\n||Consumption-public|9.0|10.6|15.0|\\n||Consumption- private|78.7|110.9|85.5|\\n|**s**|Investment-public|3.1|4.1|3.1|\\n|**Use**|Investment-private|17.3|32.3|18.8|\\n||Exports of goods and services|38.0|8.4|19.9|\\n||Imports of goods and services|46.0|66.3|42.1|\\n||**Net exports (trade balance)**|**-8.0**|**-57.8**|**-22.3**|\\n||Net transfers and factor income|6.3|32.6|11.9|\\n||Current account deficit|-1.7|-25.2|-10.4|\\n||Net foreign savings|13.8|37.7|11.2|\\n||Net<br>change<br>in<br>foreign<br>assets<br>(negative is an increase)|-12.0|-12.4|-0.7|\\n||**GNDI = GDP + net transfers and**<br>**factor income**|**106.3**|**132.6**|**111.9**|\\n\\nSources: National Accounts for 1973, 1994 and 2004. Economic dislocation and recovery in Lebanon - IMF, 1995. Back to the future, postwar reconstruction and stabilization in Lebanon – IMF, 1999.\\n\\nIn 1972, Lebanon remained open to international flows of goods, services, and capital. The foreign inflows of income, transfers, and capital amounted altogether to 20 percent of GDP, which is high but much lower than the 54 percent observed 60 years earlier. Interestingly, only part of these inflows poured into the local economy to finance imports for local consumption and investment. The largest part of inflows helped to build foreign currency reserves and purchase gold. The biggest source of foreign currencies was a dynamic exporting sector with 38 percent of GDP that helped to cover the largest part of imports. Consumption remained at 79 percent of GDP.\\n\\nAfter the destruction of the war and the large resource movements in the form of strong outmigration between 1975 and 1990, a radically different picture emerged in 1994 compared to 1972. The second globalization was in full motion and, in 1994, the economy was receiving up to 70 percent of GDP in income, transfers, and capital flows. These inflows financed imports reaching 66 percent of GDP and boosted private consumption up to 111 percent of GDP. Exports shrunk to less than 9 percent of GDP. Investments reached a high of 32 percent of GDP, but large parts of these investments were in housing and real estate, with little impact on the productive capacity. If we look at the distribution of GDP by sector, we notice an increase in the shares of industry and agriculture, two sectors that benefited from the devaluations of 1985-88. Transport and communication declined due to severe damages inflicted to infrastructures and to the redirection of regional transport away from Lebanon during the war. Trade and services also suffered, especially with the decline in tourism.\\n\\n9\", \"datasets\": [{\"mention_name\": {\"text\": \"National Accounts for 1973, 1994 and 2004\", \"confidence\": 0.5914897918701172, \"start\": 1083, \"end\": 1124, \"clean_text\": \"National Accounts for 1973, 1994 and 2004\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1083, \"end\": 1124, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1083, \"end\": 1124, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1083, \"end\": 1124}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1083, \"end\": 1124, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"report\", \"confidence\": 0.42960745096206665, \"start\": 103, \"end\": 144}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.5914897918701172, \"start\": 103, \"end\": 144}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5569033026695251, \"start\": 103, \"end\": 144}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5569033026695251, \"start\": 103, \"end\": 144}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"In sum, the real GDP in 1994 was at 60 percent of its 1972 value and at 42 percent in per capita terms, while the country relied on foreign inflows to import goods and services for investment and consumption.\\n\\nIn 2004, Lebanon recovered much of the characteristics of 1972, but with an inflated public sector since public investment and consumption reached 18 percent of GDP against 12 percent in 1972. Much of the increase was in public consumption that reached 15 percent of GDP against 9.0 percent in 1972. The export sector recovered from 1994 but was still half its 1972 level (19.9 percent of GDP in 2004 against 38 percent in 1972). More importantly, the sector composition of GDP shows the impact of a decade of foreign inflow on the structure of the economy between 1994 and 2004. Sectors producing tradable goods in agriculture and industry saw their share drop significantly, while sectors producing non-tradable goods in services and public sector developed strongly. These are all symptoms of deindustrialization with resource diversion to non-tradable sectors that developed due to a strong spending effect. Noticeably, in real terms, the GDP in 2004 was at 93 percent of its 1972 value, and per capita GDP at 56 percent.\\n\\n# **_2. A Simple General Equilibrium Analysis for Modern Lebanon: Simulation and Results_**\\n\\nThe analysis below simulates the impact of a 10 percent change in foreign inflows on the Lebanese economy for the years 1972, 1994, and 2004. For the base case, foreign reserves are excluded, and only the amount of net foreign savings that closes the current account deficit is maintained. This modification assumes that any increase in foreign savings is directly and fully channeled to the economy with incidence on demand and prices. The simulation includes both an increase and a decrease of foreign inflows by 10 percent. There is no simulation of any shock on the prices of exports, due to the diversification of industry, exports, and agriculture in modern Lebanon. With this diversification in the tradable sectors, a shock in any single sector would not have an effect comparable to the significant effect of a change in the prices of silk in 1912. The exercise aims to assess the vulnerability of Lebanon to external shocks and their impact on the economy. Table 6 summarizes the results of the analysis.\\n\\nTable 6. Simulation of external shocks on the economy of modern Lebanon (change, in percent)\\n\\n|Change in remittances & foreign savings|197|2|199|4|200|4|\\n|---|---|---|---|---|---|---|\\n|Scenarios|10.0|-10.0|10.0|-10.0|10.0|-10.0|\\n|Total Income|1.5|-1.5|10.4|-9.9|4.9|-4.8|\\n|Private Consumption|0.7|-0.7|3.5|-3.6|1.4|-1.4|\\n|Private Consumptionper capita|-1.7|-3.1|1.5|-5.5|0.1|-2.7|\\n|Investment|1.4|-1.4|4.4|-5.0|4.5|-4.8|\\n|Exports|-0.5|0.5|-6.2|6.9|-2.6|2.7|\\n|Imports|1.3|-1.3|7.9|-7.8|4.1|-4.0|\\n|Demand of domesticgoods|0.4|-0.4|0.6|-0.7|0.7|-0.7|\\n|Price of domesticgood|1.6|-1.5|12.4|-11.6|5.7|-5.5|\\n|Real exchange rate|1.6|-1.5|12.4|-11.6|5.7|-5.5|\\n|Real GDP at marketprice|0.1|-0.1|0.0|-0.4|0.1|-0.2|\\n\\nSource: author simulations.\\n\\nThe simulation for 1972 shows an economy that is relatively sheltered against the changes in foreign inflows. In terms of uses, domestic consumption and investment were largely covered by the supply of domestic goods and services, while exports provided a substantial coverage of imports. Foreign 10\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"inflows helped to close a small gap and did not have a strong impact on income, and subsequently on demand. The 10 percent rise in remittances and foreign savings in 1972 boosts income, consumption, and investment while GDP rises slightly. The losses of the country in competitiveness are moderate; and RER appreciates due to both higher demand and higher prices of domestic goods. Exports decline slightly while the increase in imports is relatively high. The decline by 10 percent in foreign inflows would have almost exactly the symmetric effect of the rise in foreign inflows.\\n\\nLebanon’s economy in 1994 is vulnerable and exposed, with low productive capacities in all sectors. Four years after the end of the war, Lebanon is pouring foreign flows in the economy to boost disposable income and finance large imports supplying a growing aggregate domestic demand. In 1994, a 10 percent increase in foreign inflows has an impact of the same magnitude on income and imports; consumption and investment rise. Competitiveness drops sharply with a strong rise in RER and prices of domestic goods, while exports drop dramatically. However, a rise in inflows remains neutral in terms of impact on GDP growth. Interestingly, GDP declines by 0.4 percent if foreign inflows decrease by 10 percent. Here, the impact shows increased sensitivity of the economy towards the spending effect of foreign inflows when their share to GDP reaches high levels, while the share of exports remains low. Another consequence of the high share of inflows to GDP is increased sensitivity of the prices of domestic goods towards a rise in foreign inflows.\\n\\nIn 2004, the Lebanese economy concluded a decade of high magnitude foreign inflows that diverted the economy towards higher shares of non-tradable sectors at the expense of tradable activities. Exports recovered some of the pre-war dynamic, especially in services and tourism. However, the reliance on foreign inflows to finance imports for local consumption and investment remained high. The effects of these inflows on RER and prices of domestic goods are comparable to the effects observed in 1912. Simulation predicted that a 10 percent decline in foreign inflows would reduce private consumption per capita by 2.7 percent. The reversals in foreign inflows starting 2011 and the aggravation of the tendency in the months leading to the crisis that unfolded in October 2019 have indeed exerted a downward pressure on consumption and welfare and led to large outmigration.\\n\\n# **_3. Confronting Predictions with Reality: Simulations and Actual Observations for the Period 20052020_**\\n\\nThe analysis below compares the simulations of the impact of foreign inflow shocks on economic variables to the actual changes observed through the National Accounts and the Balance of Payments. Table 7 shows that those simulations match the actual observations from 2005 to 2020.\\n\\nIndeed, while a simulated positive shock with a 10 percent increase in foreign inflows raises income by 4.9 percent, the actual yearly average change in those inflows by 14.1 percent between 2005 and 2010 increased income by 6.8 percent on average and the actual change by 7.0 percent in 2011-2016 increased income by 2.1 percent on average. Actual private consumption and investment also increased as predicted by the simulation, as well as imports and domestic demand of domestic goods. However, while consumption per capita increased in 2005-2010, it declined in 2011-2016, due to the rapid increase in the resident population (Figure 4) with the inflow of refugees from the Syrian Arab Republic, a development that the simulations could not predict.\\n\\nThe simulated negative shock with a 10 percent decline in foreign inflows reduces income by 4.8 percent, and the actual decline in inflows by 10.8 percent on average in 2017-2019 reduced income by 2.8 percent. Actual private consumption, investments, imports, and domestic demand for 11\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"domestic goods declined as predicted. Interestingly, the predicted and observed declines in consumption are extremely close (1.4 and 1.3 percent, respectively). Yet, the actual consumption per capita merely changed, hence pointing to the accelerated outmigration and to the reversal in refugees’ trends (Figure 4).\\n\\nTable 7. Comparing simulations and observations for the period 2005-2020 (change, in percent)\\n\\n|Change in|2004|2005-2010|2011-2016|2004|2017-2019|1912|2020|\\n|---|---|---|---|---|---|---|---|\\n|remittances &|Simulated|Observed|Observed|Simulated|Observed|Simulated|Observed|\\n|foreign savings|10.0|14.1|7.0|-10.0|-10.8|-90.0|-50.3|\\n|Total Income|4.9|6.8|2.1|-4.8|-2.8|-39.7|-16.5|\\n|Private Consumption|1.4|6.3|2.7|-1.4|-1.3|-33.7|-27.4|\\n|Private Consumption<br>per capita|0.1|5.2|-1.8|-2.7|-0.1|-34.7|-24.4|\\n|Investment|4.5|10.8|3.4|-4.8|-16.9|-31.3|-65.9|\\n|Exports|-2.6|6.7|-1.5|2.7|-1.0|7.9|-34.2|\\n|Imports|4.1|6.9|3.6|-4.0|-2.2|-37.0|-33.4|\\n|Demand of domestic<br>goods|0.7|6.9|3.4|-0.7|-4.2|-17.6|-23.0|\\n|Price of domestic<br>good|5.7|3.6|4.7|-5.5|6.2|-36.1|40.5|\\n|Real exchange rate|5.7|0.1|8.1|-5.5|5.5|-36.1|-43.7|\\n|Real GDP at market<br>price|0.1|6.7|2.0|-0.2|-2.7|-2.6|-25.9|\\n\\nSource: author simulations and observations from Lebanon National Accounts 2004-2020 (CAS, 2022) and WDI, 2023.\\n\\nFigure 4. Population in Lebanon 2004-2022 (thousands, right hand axis) and share of ages 15-64 (percent, left hand axis) <!-- Start of picture text --> 70% 6,500<br>69% 6,250<br>68% 6,000<br>67% 5,750<br>66% 5,500<br>65% 5,250<br>64% 5,000<br>63% 4,750<br>62% 4,500<br>Total Population Share Ages 15-64<br>2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022<br><!-- End of picture text --> Source: World Bank, WDI 2023.\\n\\nThe most striking outcome is the similarity between the observed changes in 2020, the year of fullblown crisis, and the simulation with the structure of the economy in 1912. Here, with the exceptions of the decline in growth that substantially overshoots predictions, and exports and prices of domestic goods that present reversed signs,<sup>3</sup> all other changes are within the range of the 1912 simulation. Indeed, the simulated 90 percent collapse in foreign inflows for the economy of 1912 depresses income by 39.7 percent. The actual decline in\\n\\n> 3 The simulation using the structure of 1912 predicts a decline in domestic prices and an increase in exports which attenuates GDP decline. Indeed, in 1912, exports represented over 60 percent of GDP (see Tables 2 and 3). In 2019, exports were only 21 percent of GDP, domestic prices increased because of massive increase in money supply, and, more importantly, the massive disruptions in service provisions – including energy shortages – prevented the economy from turning the decline in real exchange rate into an opportunity for a better competitive position.\\n\\n12\", \"datasets\": [{\"mention_name\": {\"text\": \"Lebanon National Accounts 2004-2020\", \"confidence\": 0.7321317791938782, \"start\": 1248, \"end\": 1283, \"clean_text\": \"Lebanon National Accounts 2004-2020\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1248, \"end\": 1283, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1248, \"end\": 1283, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1248, \"end\": 1283}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1248, \"end\": 1283, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.46272367238998413, \"start\": 123, \"end\": 158}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7321317791938782, \"start\": 123, \"end\": 158}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.46542319655418396, \"start\": 123, \"end\": 158}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.46542319655418396, \"start\": 123, \"end\": 158}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"foreign inflows in 2020 is 50.3 percent and leads to a contraction of income by 16.5 percent. Private consumption, private consumption per capita, imports, and domestic demand for domestic goods all decline within the range and direction of the simulation.\\n\\nThe RER dynamics is of particular importance. Indeed, while a decline in foreign inflows in 2017-19 should have led to a decline in RER, this did not happen, and the RER continued to increase. This is owed to the continuation of the policy of de-facto peg and is a symptom of the overvaluation of the Lebanese currency. This persistent increase in RER led to a drop in exports by 1-1.5 percent on average between 2011 and 2019. The shutdown of the main channel for foreign inflows through the banking sector in 2020 and the sharp decline in these inflows led to a decline in the RER, in parallel to a collapse of the de facto peg of the LBP. This decline in the RER happened despite the increase in domestic prices, the latter being influenced by a massive expansionary monetary policy followed by the Central Bank. Yet, the increase in the prices of imports owed to the collapse of the peg outweighed the increase in domestic prices and produced a decline in the RER.\\n\\nIn sum, it is interesting to see that the observed dynamic in 2020 is better proxied through a crisis scenario applied to the structure of the economy of 1912 rather than to the economy of 2004. This implies that the economic policy followed between 2004 and 2019 gradually distorted the economy of Lebanon towards an extraversion, dependence on foreign financial inflows, and vulnerability that prevailed a century earlier. Back then, the shock of WWI starved the population. Nowadays, the path towards the crisis and the crisis itself reduced the resident population by 900,000 individuals between 2015 and 2022, of which 300,000 between 2020 and 2022 (Figure 4), despite the COVID-19 restrictions on travel. Hence, a century later, the adjustment to the crisis in Lebanon is again happening through the society and the population, in 1916 through starvation, and in the 2020s through outmigration.\\n\\n# **D. Conclusion**\\n\\nIn the early 20th century, Mount Lebanon experienced deindustrialization due to the development of exports of raw silk. The strong development of the silk sector generated both a resource movement effect and a spending effect like those described by Corden & Neary. This deindustrialization and the monopolization of industry, agriculture, and exports by the silk sector created strong dependency and vulnerability of Mount Lebanon to shocks in the silk market. Outmigration and remittances were the spontaneous responses of the system to the fluctuation of income and consumption related to the fluctuation in exports and prices of silk. Remittances from migrants improved income and consumption and had a strong spending effect. The latter led to an increase in the demand for imported goods and in the demand for and prices of domestic goods. This in turn led to a real appreciation and to deindustrialization and eviction of exporting sectors, which further increased the reliance on remittances. The dependency on silk exports and remittances reached such levels that the interruption of international flows of goods, capital, and people drove the population into starvation in 1916.\\n\\nBetween 1972 and 2004, Lebanon moved back to characteristics close to those of Mount Lebanon of 1912 in terms of dependency on (and vulnerability to) foreign inflows. In 1972, prior to the oil boom in the Middle East, Lebanon already was receiving substantial foreign inflows, but with few effects on income, prices, and GDP. Most of these inflows were accumulated into foreign currency reserves and gold and neutralized. Between 1975 and 1990, war devastation was a major shock to the country and the spontaneous responses of the system were outmigration and high foreign inflows. These foreign 13\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"inflows sustained income and consumption during the war and allowed a rapid recovery of income, consumption, and investment after the war. In 1994, large amounts of foreign inflows had a strong spending effect on the economy with an increase in the demand for imported goods and in the demand for and prices of domestic goods. This in turn led to a real appreciation and repressed exporting sectors over the whole decade between 1994 and 2004.\\n\\nIn 2004, Lebanon became an indebted country, with large non-tradable sectors and structural dependency on foreign inflows that depressed competitiveness and promoted outmigration in patterns like those observed 92 years earlier. However, similarities with 1912 stop here. Indeed, Mount Lebanon could not make any policy choice since it was a subordinated entity within the Ottoman Empire. In the Lebanon of 2004, policy choices were clear: the country adopted a de-facto peg between the US$ and the LBP in nominal terms. Lebanon has also adopted a double currency system where US$ and LBP are equally used in all types of transactions. These two measures have reduced the role of the nominal exchange rate as a variable of adjustment that would affect RER and competitiveness. These two measures also allowed banks to attract an increasing amount of capital that has been channeled to the public sector, to importers, and to non-tradable sectors. Real appreciation and loss of competitiveness have been aggravated.\\n\\nPolitical economy weighed heavily on the direction of policy choices. Bankers used high interest rates to attract more capital, which they were sure would be highly remunerated. Politicians were comfortable with the availability of funds they could channel to their constituencies. Importers were encouraging policies that expanded markets for their products. Bankers, landlords, and real estate developers would push for the overvaluation of real estate assets and for their extensive use as collateral in lending operations. These deliberated policy choices paved the way to the collapse in 2019 when a reversal in foreign inflows over many years culminated in a major crisis with unprecedented impact on the society unseen since the famine of 1916. Instead of starvation, the Lebanese society is now bleeding its vital forces through a massive outmigration, an adjustment that weakens the society and enhances the grip of the existing power structures over its destiny.\\n\\n# **BIBLIOGRAPHY**\\n\\n- Bou Habib, Chadi Adib. 2007 - _The effects of capital inflows on a small open economy growth or Dutch disease? a study of the Lebanese case_ . American University of Beirut.\\n\\nBourbonnais, Regis. 1993 – Econométrie, cours et exercices corrigés. Published by Dunod.\\n\\n- Bussolo, Maurizio, Medvedev, Denis, 2007 _– Do remittances have a flip side? A general equilibrium analysis of remittances, labor supply responses, and policy options for Jamaica._ Policy Research Working Paper No. 4143. Published by the World Bank.\\n\\n- Chami, Ralph, Fullenkamp, Connel, & Jahjah, Samir, 2003 _– Are immigrant remittance flows a source of capital for development?_ Working Paper No. 03/189. Published by the International Monetary Fund.\\n\\n- Chami, Ralph, Fullenkamp, Connel, & Jahjah, Samir, 2005 _– Are immigrant remittance flows a source of capital for development?_ Staff Papers Vol. 52, No. 1. Published by the International Monetary Fund.\\n\\n14\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"- Corden, W. Max, & Neary, J. Peter, 1982 – _Booming sector and deindustrialization in a small open economy_ . Vol.92. No. 368. The Economic Journal.\\n\\n- Dessus, Sebastien & Astrup, Claus, 2005 – _Exporting goods or exporting labor?: long term implications for the Palestinian economy._ Published in the Review of Middle East Economics and Finance, Vol. 3, No.1, 39-61.\\n\\n- Devarajan, Shantayanan, Go, Delfin S., Lewis, Jefferey D., Robinson, Sherman, & Sinko, Peka, 1997 – _Simple general equilibrium modeling._ Chapter 6 in “Applied methods for trade policy analysis – A handbook”, Cambridge University Press.\\n\\n- Devarajan, Shantayanan & Go, Delfin S., 1998 – _The simplest general equilibrium mode of an open economy._ Journal of policy modeling, published by Elsevier Science Inc.\\n\\n- Domaç, I., & Shabsigh, G., 1999 – _Real exchange rate behavior and economic growth: evidence from Egypt, Jordan, Morocco and Tunisia_ . Working Paper No. 99/40. Published by the International Monetary Fund.\\n\\n- Firro, Kais, 1990 - _Silk and Agrarian Changes in Lebanon, 1860-1914_ . International Journal of Middle East Studies, Vol. 22, No. 2 (May, 1990), pp. 151-169\\n\\n- Harberger, Arnold C., 2004 – _The real exchange rate: issues of concept and measurement._ University of California, Los Angeles.\\n\\n- International Monetary Fund, 1999 _– Back to the future: postwar reconstruction and stabilization in Lebanon_ . Occasional Paper No. 176.\\n\\n- International Monetary Fund, 1999 _– Economic dislocation and recovery in Lebanon_ Occasional Paper No. 120.\\n\\n- Khater, Akram, 2017 - _Why did they leave? Reasons for early Lebanese migration_ . Khayrallah Center for Lebanese Diaspora Studies, North Carolina State University.\\n\\n- Khater, Akram, 2001 - _Inventing Home - Emigration, Gender, and the Middle Class in Lebanon, 18701920._ University of California Press\\n\\n- Krugman, Paul R., & Obtsfeld, Maurice, 2003 – _International Economics, theory and policy._ Sixth Edition, published by Addison Wesley.\\n\\n- Lebanese University Publications, 1970 - _Lubnan, mabaheth ilmiah wa ijtimaya (Lebanon, scientific and social studies)._ Ordered by the Governor of Mount Lebanon, Ismael Haki Bek in 1918.\\n\\n- Ministry of Planning, Central Administration of Statistics, 1974 - _Recueil de Statistiques Libanaises, Année 1973 (Lebanon Statistical Abstract, 1973)._\\n\\n- Nahas, Charbel, 2003 – _Huzuz ijtinab al azama wa churut takhatiha (Chances of avoiding the crisis and conditions of overcoming it)._ Published by Dar An-Nahar.\\n\\n- Nowak, Jean-Jacques, 1998 – _Syndrome néerlandais et théorie du commerce international (Dutch Disease and theory of international trade)_ . Published by Economica.\\n\\n15\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"- Opoku-Afari, Maxwell, Morrissey, Oliver & Lloyd, Timaballero, 2004 – _Real exchange rate response to capital inflows: a dynamic analysis for Ghana._ Research paper No. 04/12. Published by the Center for Research in Economic Development and International Trade, University of Nottingham.\\n\\n- Pindyck, Robert S., & Rubinfeld, Daniel L., 1998 – _Econometric models and economic forecasts._ Fourth edition, published by McGraw-Hill.\\n\\n- Pitts, Graham Auman, 2021 - _Was Capitalism the Crisis? Mount Lebanon’s World War I Famine_ . Environment & Society Portal, Arcadia (Spring 2021), no. 3. Rachel Carson Center for Environment and Society. doi:10.5282/rcc/8801\\n\\n- Presidency of the Council of Ministers, 2022 - Central Administration of Statistics _, Annual National Accounts 2004-2020_ .\\n\\n- Presidency of the Council of Ministers, Economic Accounts mission, 2006 - _Economic Accounts of Lebanon, 2004._\\n\\nWorld Bank, 2023 - _World Development Indicators_ .\\n\\n16\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099324301312410845/pdf/IDU16cf9281c16b28145b51b4fa17b1af694b7a9.pdf\", \"pages\": [17]}}]}"}}},{"rowIdx":93,"cells":{"doc_id":{"kind":"string","value":"34249090"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34249090\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"authors\": {\"0\": {\"author\": \"Behrer,Arnold Patrick\"}, \"1\": {\"author\": \"Heft-Neal,Sam\"}}, \"count\": \"World\", \"count_key\": \"517191\", \"docna\": {\"0\": {\"docna\": \"In Most Low- and Middle-Income Countries\\n Pollution Levels Are Higher in Wealthier Areas\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"DEC-Sustainability & Infrastruct (DECSI)\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34249090\"}, \"repnb\": \"WPS10689\", \"docdt\": \"2024-01-31T00:00:00Z\", \"datestored\": \"2024-01-31T02:49:21Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"Air pollution is a major threat to\\n health, and the dangers are particularly acute in low- and\\n middle-income countries. However, little is known about how\\n the burden of pollution is spread across the wealth\\n distribution in these countries. This paper uses new data\\n providing high-resolution wealth estimates for more than 100\\n low- and middle-income countries, combined with equally\\n high-resolution estimates of air pollution, to estimate how\\n wealth is correlated with ambient air pollution around the\\n world. The findings show that on average air pollution is\\n positively correlated with wealth, but the relationship is\\n highly heterogeneous across countries. The fact that air\\n pollution and wealth are both disproportionately high in\\n urban areas, where economic activity is largely\\n concentrated, appears to drive this relationship. When the\\n analysis is limited to anthropogenic sources of pollution,\\n the relationship becomes less heterogeneous and more\\n systematically positive. The paper also examines the\\n relationship between pollution exposure and wealth within\\n large cities around the world. Again, the findings show\\n substantial heterogeneity across cities. The paper explores\\n several hypotheses for this heterogeneity but does not find\\n a single explanation. Economic concentration within cities\\n appears to explain some of the relationship. Cities with\\n more concentrated economic opportunity tend to have more\\n positive correlations between pollution and wealth.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10689; PLANET\", \"display_title\": \"In Most Low- and Middle-Income\\n Countries Pollution Levels Are Higher in Wealthier Areas\", \"disclosure_date\": \"2024-01-31T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099449201312441737/text/IDU12354a7cd103c314993198571ea404b85aecb.txt\", \"listing_relative_url\": \"/research/2024/01/34249090/most-low--middle-income-countries-pollution-levels-higher-wealthier-areas\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099449201312441737/IDU12354a7cd103c314993198571ea404b85aecb\", \"new_url\": \"2024/01/34249090/IDU12354a7cd103c314993198571ea404b85aecb\", \"ext_pub_date\": \"2024-01-31T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Air Pollution m1327122 1791,Urban Development m1327371 699,Inequality and Shared Prosperity m1327655 700\", \"docm_id\": \"IDU-2354a7cd-03c3-4993-9857-ea404b85aecb\", \"chronical_docm_id\": \"IDU-2354a7cd-03c3-4993-9857-ea404b85aecb\", \"publishtoextweb_dt\": \"2024-01-31T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10689\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"guid\": \"099449201312441737\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"46\", \"author_keys\": {\"0\": {\"author_key\": \"1561699\"}, \"1\": {\"author_key\": \"NA\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099449201312441737/IDU12354a7cd103c314993198571ea404b85aecb\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10689 In Most Low- and Middle-Income Countries Pollution Levels Are Higher in Wealthier Areas _A. Patrick Behrer Sam Heft-Neal_ Development Economics Development Research Group January 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10689\\n\\n# **Abstract**\\n\\nAir pollution is a major threat to health, and the dangers are particularly acute in low- and middle-income countries. However, little is known about how the burden of pollution is spread across the wealth distribution in these countries. This paper uses new data providing high-resolution wealth estimates for more than 100 low- and middle-income countries, combined with equally high-resolution estimates of air pollution, to estimate how wealth is correlated with ambient air pollution around the world. The findings show that on average air pollution is positively correlated with wealth, but the relationship is highly heterogeneous across countries. The fact that air pollution and wealth are both disproportionately high in urban areas, where economic activity is largely concentrated, appears to drive this relationship. When the analysis is limited to anthropogenic sources of pollution, the relationship becomes less heterogeneous and more systematically positive. The paper also examines the relationship between pollution exposure and wealth within large cities around the world. Again, the findings show substantial heterogeneity across cities. The paper explores several hypotheses for this heterogeneity but does not find a single explanation. Economic concentration within cities appears to explain some of the relationship. Cities with more concentrated economic opportunity tend to have more positive correlations between pollution and wealth.\\n\\nThis paper is a product of the Development Research Group, Development Economics. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at abehrer@worldbank.org.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [{\"mention_name\": {\"text\": \"high-resolution wealth estimates\", \"confidence\": 0.35313716530799866, \"start\": 327, \"end\": 359, \"clean_text\": \"high-resolution wealth estimates\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 327, \"end\": 359, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 327, \"end\": 359, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 327, \"end\": 359}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 327, \"end\": 359, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9900723695755005, \"start\": 327, \"end\": 359}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.35313716530799866, \"start\": 327, \"end\": 359}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.719639003276825, \"start\": 327, \"end\": 359}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.719639003276825, \"start\": 327, \"end\": 359}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# In most low- and middle-income countries levels are pollution in wealthier areas higher\\n\\n## A. Patrick Behrer<sup>1</sup><sup>_∗_</sup> and Sam Heft-Neal<sup>2</sup>\\n\\n> 1The World Bank, Washington, DC, USA\\n\\n2 Stanford University Center on Food Security and the Environment, Stanford, CA, USA\\n\\n> _∗_ To whom correspondence should be addressed; e-mail: abehrer@worldbank.org.\\n\\nKeywords: Air pollution, inequality, wealth,LMIC countries, sustainable development\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **1 Introduction**\\n\\nAir quality poses a significant threat to health in most parts of the world. At current levels air pollution is responsible for approximately 10 million premature deaths each year.<sup>1</sup> While the number of deaths associated with indoor air pollution has declined in recent years, increases in deaths due to ambient pollution have offset this improvement and led to an overall increase in total air pollution deaths. More than half of the world’s population now lives in areas where ambient air pollution levels are increasing. The majority of those facing increasing levels of air pollution live in low- and middle-income countries (LMICs)<sup>2</sup> where more than 90% of air pollution-related deaths occur.<sup>3</sup> The health burden of air pollution is also unequally shared _within_ countries. With a few exceptions, low income populations in Europe are exposed to higher pollution levels than wealthy populations in the same country.<sup>4</sup> In the United States low income populations<sup>5–8</sup> and racial and ethnic minority populations<sup>8–13</sup> tend to be more exposed to ambient particulate pollution. These differences across groups have persisted over the past several decades despite universal reductions in pollution levels.<sup>14,15</sup> While the within-country relationship between income or wealth and air pollution has been well documented in Europe and the US, evidence from LMICs is scarce.<sup>16</sup> This scarcity of evidence in the parts of the world with the largest populations, the highest levels of ambient pollution, and where pollution levels are increasing fastest limits our understanding of the present and future patterns of vulnerability. Lack of evidence may hinder effective policy-making to reduce the harms of this exposure. Furthermore, with the exception of recent studies in Eastern Europe<sup>4</sup> and China,<sup>17</sup> examinations of income and ambient air pollution outside of high-income countries have been limited to urban areas.<sup>18–23</sup> These city-level studies have documented pollution exposures declining with income<sup>18–21,23</sup> or non-linear relationships with pollution highest at wealth extremes.<sup>22</sup> Population-scale estimates of the relationship between wealth and pollution in LMICs that include observations from both rural and urban areas are needed. Conclusions about the distribution of pollution exposures across wealth levels in Europe and the United States may not apply where settlement patterns differ.<sup>17</sup> Urban centers tend to be economic drivers that generate substantially higher incomes than are typically found in rural areas.<sup>24,25</sup> While this gradient exists in most countries it can be especially pronounced in LMICs. Higher levels of economic activity can simultaneously lead to higher levels of pollution<sup>26–28</sup> and wealth. If urban residents are on average wealthier and experience more ambient pollution than rural residents, then wealth could be positively correlated with ambient pollution exposure across the full population. This is true even in settings where within urban pollution is negatively correlated with neighborhood 2\", \"datasets\": [{\"mention_name\": {\"text\": \"LMICs\", \"confidence\": 0.6723792552947998, \"start\": 1444, \"end\": 1449, \"clean_text\": \"LMICs\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.5217396020889282, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1444, \"end\": 1449, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1444, \"end\": 1449}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1444, \"end\": 1449, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.46414241194725037, \"start\": 698, \"end\": 703}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6723792552947998, \"start\": 698, \"end\": 703}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5371822118759155, \"start\": 698, \"end\": 703}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5371822118759155, \"start\": 698, \"end\": 703}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"wealth. For these reasons the most comprehensive population-scale study<sup>17</sup> outside the U.S. and Europe, a population-scale study in China, finds a strong _positive_ relationship between pollution and income. More broadly, for countries where anthropogenic sources of pollution dominate, there is a long recognized potential trade-off between urban-based development and environmental quality.\\n\\nA primary reason that the few existing studies in LMICs have been mostly limited to urban areas is that data on air pollution and income or wealth are disproportionately available for cities. To overcome this challenge across a broad geographic scale we leverage recently released spatially resolved estimates of wealth for 103 LMICs.<sup>25</sup> These relative wealth indices (RWI) represent the most comprehensive micro-level wealth estimates to-date. They were derived from machine learning models trained on survey based measures of wealth. We combine these relative wealth estimates with global monthly spatially resolved estimates of ambient PM2 _._ 5 for 2015-2020 that are derived from a combination of satellite observations and modeled output.<sup>29</sup> Together, the data allow us to estimate the correlation between relative wealth and average ambient PM2 _._ 5 levels across grid cells in more than 100 LMICs at increasingly fine scales: across the full sample, separately within each country, and within cities.\\n\\nFor the pooled sample we regress average PM2 _._ 5 on relative wealth at the grid-cell-level and include separate intercepts for each country (Equation 1). The intercepts (fixed effects) account for baseline differences in wealth and pollution across countries and for the fact that wealth measures are normalized at the country level. The estimated coefficient on relative wealth in this specification provides a measure of average within-country correlation between wealth and pollution. Next, we estimate separate regressions for each country again regressing cell-level average PM2 _._ 5 on cell-level relative wealth (Methods). This exercise provides us with countrylevel estimates for the within-country correlation between wealth and pollution and allows us to assess which factors contribute to pollution being progressive or regressive in different settings. We next restrict our sample to large urban areas and estimate city-specific regressions in order to assess how pollution varies with wealth across neighborhoods within cities. Finally, we explore potential drivers of heterogeneity in both country- and city- level relationships.\\n\\n# **2 Results**\\n\\nWhile wealthier LMICs tend to have lower pollution on average (Section SI-1.1), we find that the average within-country correlation between wealth and ambient pollution is positive. This relationship is seen both in the raw data and in our regression results. Areas with the lowest average PM2 _._ 5 concentrations have the lowest average RWI levels while areas with the highest PM2 _._ 5 concentrations have the highest average RWI (Figure 1a). Results from our preferred 3\", \"datasets\": [{\"mention_name\": {\"text\": \"spatially resolved estimates of wealth for 103 LMICs\", \"confidence\": 0.43231430649757385, \"start\": 685, \"end\": 737, \"clean_text\": \"spatially resolved estimates of wealth for 103 LMICs\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 685, \"end\": 737, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 685, \"end\": 737, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 685, \"end\": 737}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 685, \"end\": 737, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9601503610610962, \"start\": 685, \"end\": 737}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.43231430649757385, \"start\": 685, \"end\": 737}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5827623009681702, \"start\": 685, \"end\": 737}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5827623009681702, \"start\": 685, \"end\": 737}}, {\"mention_name\": {\"text\": \"survey based measures of wealth\", \"confidence\": 0.5069124698638916, \"start\": 917, \"end\": 948, \"clean_text\": \"survey based measures of wealth\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 917, \"end\": 948, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 917, \"end\": 948, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 917, \"end\": 948}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 917, \"end\": 948, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9630182385444641, \"start\": 48, \"end\": 79}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.5069124698638916, \"start\": 48, \"end\": 79}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8701485395431519, \"start\": 48, \"end\": 79}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8701485395431519, \"start\": 48, \"end\": 79}}, {\"mention_name\": {\"text\": \"raw data\", \"confidence\": 0.4918316602706909, \"start\": 2820, \"end\": 2828, \"clean_text\": \"raw data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2820, \"end\": 2828, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2820, \"end\": 2828, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2820, \"end\": 2828}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2820, \"end\": 2828, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.8542948365211487, \"start\": 193, \"end\": 201}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4918316602706909, \"start\": 193, \"end\": 201}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7171515822410583, \"start\": 193, \"end\": 201}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7171515822410583, \"start\": 193, \"end\": 201}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"regression specification, weighting grid-cells jointly by country population and the inverse of the error on our wealth measure, indicate that a 1 standard deviation increase in the wealth index is associated with a 0.04 standard deviation increase in ambient PM2 _._ 5. This result is robust to a variety of alternative weighting schemes (Table SI-1) and the relationship is further strengthened by trimming our wealth data at the 5<sup>_th_</sup> and 95<sup>_th_</sup> percentile (Table SI-2).\\n\\nThis positive correlation appears to be driven by urban-rural differences in both income and pollution. The shading in Figure 1b-c highlights that both grid-cells with the highest pollution levels and grid cells with the highest wealth levels are the most urban in our sample. More generally, being in or near larger urban areas is associated with higher levels of wealth (Figure SI-2). Wealth increases with urbanization both within and across categories (urban, peri-urban, rural) and the differences are stark enough that on average the wealthiest rural areas are poorer than the poorest urban areas. Exposure to ambient air pollution follows similar patterns albeit less dramatically (Figure 1c). Nonetheless, average pollution in urban areas is 12% ( _t-stat:_ 47.7) higher than in peri-urban areas and 50% ( _t-stat:>_ 100) higher than in rural areas.\\n\\nOur pooled estimate hides substantial heterogeneity across countries. Country specific estimates indicate that ambient pollution increases with wealth in the majority of LMICs. However, a substantial minority of countries demonstrate the opposite relationship. While wealthier individuals live in more polluted locations in countries like Nepal, pollution decreases strongly with wealth in other countries like Nigeria (Figure 2a). When we map the country level relationship (Figure 2b-c) one immediate pattern becomes clear: countries with strong negative correlations between ambient pollution and wealth are highly concentrated in West or Central Africa. In most of the rest of the world, particularly Asia, wealthier areas have higher ambient pollution.\\n\\nOne reason the wealth-pollution relationship may differ in West and Central Africa is that a substantial portion of the ambient particulate pollution in these regions is driven by dust storms off the Sahel and Sahara.<sup>30,31</sup> Pollution from these sources does not generate the same correlation with urbanization (and hence wealth) as anthropogenic sources and is more severe in rural locations which are often in closer proximity to the dust sources (Figure 3). Across the sample dust and sea salt makeup 3.5% of PM2 _._ 5 concentrations in urban areas but more than 30% of PM2 _._ 5 in rural areas (Methods). Panels **A** and **B** of Figure 3 show the different spatial distributions of total PM2 _._ 5 and PM2 _._ 5 with dust and sea salt removed (an imperfect proxy for anthropogenic emissions) in West Africa. Total PM2 _._ 5 is concentrated inland, near the Sahel in particular, while anthropogenic PM2 _._ 5 is heavily concentrated along the coast, where urbanization and wealth are highest (Panels **C** and **D** ). When we recalculate the correlation between wealth and pollution using PM2 _._ 5 measures that remove dust and sea salt most of the correlations between wealth and total ambient pollution that were negative become positive suggesting that even in these settings anthropogenic emissions increase with wealth (Figure 2c,d 4\", \"datasets\": [{\"mention_name\": {\"text\": \"wealth data\", \"confidence\": 0.7081063389778137, \"start\": 413, \"end\": 424, \"clean_text\": \"wealth data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 413, \"end\": 424, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 413, \"end\": 424, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 413, \"end\": 424}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 413, \"end\": 424, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.2588138282299042, \"start\": 413, \"end\": 424}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7081063389778137, \"start\": 413, \"end\": 424}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.49755480885505676, \"start\": 413, \"end\": 424}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.49755480885505676, \"start\": 413, \"end\": 424}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"inset). Examining correlations based on the primary source of anthropogenic pollution does not generate similar changes in the patterns of correlation (see Section SI-1.2).\\n\\nLeveraging the fine spatial scale of the data, we estimate city-specific correlations across grid cells that lie within contiguous areas determined to be highly urban (see Methods). Estimating these correlations separately by city we find the relationship is again heterogeneous across locations (Figure 4a) and ranges from strongly negative to strongly positive. Re-estimating within-city correlations using the dust and sea salt removed pollution measures leads to more positive correlations on average across our cities but does not completely eliminate the presence of negative correlations and substantial heterogeneity remains. Figure 4b shows within-city correlations for three of the largest population centers in our sample (Rio de Janeiro, Delhi, and Lagos) and highlights the observed heterogeneity. Lagos’s location on the coast in West Africa mean that dust and sea salt contribute relatively more to total PM2 _._ 5 levels and thus the estimated correlation changes depending on whether we compare wealth with total or anthropogenic PM2 _._ 5 (Figure 4a).\\n\\nOne factor that appears to be important in many cities is how populations sort across different terrains. We find that on average in cities where the wealthy tend to live at higher elevations they also tend to live in lower pollution areas and vice versa (Figure SI-5). This relationship does not explain all of the variation in the estimated correlation between wealth and pollution across cities but we find a standard deviation increase in the correlation between wealth and elevation is associated with a 0.2 standard deviation more negative correlation between wealth and pollution [ _t_ -stat: 3.03].\\n\\nComparing the pollution-wealth and pollution-elevation gradients in Lagos, Delhi, and Rio de Janeiro emphasizes both the potential importance of elevation and its varying influence. In Rio de Janeiro the poorest households live in higher-elevation neighborhoods further removed from primary emissions sources that are located near the center of town. As a result, wealthier neighborhoods in Rio de Janeiro on average have higher pollution. The opposite relationship exists in Lagos where the wealthiest live at the highest elevations in the city further removed from emissions sources. Delhi illustrates the limits of elevation as an explanatory variable as we observe a stronger gradient between wealth and pollution than in Rio de Janeiro yet we see a relatively weak gradient between wealth and elevation due at least in part to the limited elevation changes within the city.\\n\\nWe see similar patterns in other cities including Kathmandu, Dhaka, and Islamabad (Figure 3f). In both Islamabad and Kathmandu, cities in or near the mountains, higher wealth areas are more polluted. In Dhaka, where there is little elevation change, wealthier areas have less pollution. Comparing the joint distribution of wealth and pollution across all of South Asia (Panel **E** , Figure 3) emphasizes both the importance of terrain, highlighted by the large band 5\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"of high pollution through the Indo-Gangetic plains into the Himalayan foothills, and that urban areas throughout the region are hotspots of both pollution and wealth. This is particularly notable in South India where wealth is relatively high and pollution relatively low, yet urban centers are still hotspots of high pollution. In contrast, we observe much weaker relationships between pollution and wealth when we look at within-city correlations in the US (Figure SI-10).\\n\\nWe consider several other factors that could influence neighborhood sorting patterns in Section SI-1.3. We find that economic concentration may explain some variation across cities, but find little difference in the pattern of heterogeneity based on countries’ history of European colonization (Figures SI-11 & SI-12), location in equatorial regions, whether they were part of the Soviet Union or primary source of particulate pollution within the city (Figure SI-13).\\n\\n6\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"<!-- Start of picture text --> A. RWI and binned pollution —B. PM,., concentrations and urbanicity<br>RWI value Urbanicity PM..; concentration (g/m?)<br>High- i 100°<br>90- :<br>70-<br>60-<br>50- =<br>! 40- ------—~_ :<br>30- = . = Saaeeeee<br>10-<br>i 0- 12345 6 7 8 9 10 11 12 13 14 45 16 17 18 19 20 21 22 23 24 25 26 27 26 29 30.<br>| Urban Peri-Urban Rural<br>| | C. RWI levels and urbanicity<br>RWI value<br>Low- | ao<br>' 123<br>\\\\ \\\\ \\\\ \\\\ \\\\ ; 456789 won 12 13 14 15 16 17 18 19 20 21 2 23 24 25 26 27 28 28 0<br>0-5 25-30 50-55 75-80 95+ Urban Peri-Urban Rural<br>5 wg/m? PMz2.s bins<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"<!-- Start of picture text --> A. e* B.<br>:5<br>Ho<br>i ‘ a<br>H Na<br>i<br>i'$ Pollution{Pollution measuredmeasured as totalas non-naturalPM2.s PM2.s<br>iS ~<br>Correlation between wealth and pollution<br>: 4 3 2 7+ oO tf 2 3 4<br>$} = a<br>3: Fy a<br>3 -— @<br>°°o e rN o?¢ Bg es =<br>.<br>° -15 -10 5 0 5 10<br>° comnlation betweon wealth Correlation between wealth<br>and average pollution and average pollution<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"<!-- Start of picture text --> a. Total PM, b. Anthropogenic PM,, c. Urbanicity d. Relative Wealth<br>a 4 ; Ee et tang ss cc<br>as eee See + eS lee ee ee<br>2 ae a ‘ g Wek net<br>Wealth Wealth Wealth<br>A A MM =~) _———Y<br><0 20 30 40 50 60 +70 <0 20 30 40 50 60 >70 Most Rural Most Urban Lower Higher<br>bom? gn?<br>e. Joint distribution of pollution and wealth f. City-level pollution and wealth<br>Islamabad Sedia erewii 83ea \\\"soe a. 3<7<br>be, yr 6 es<br>CCRC 5S 9, i= o 5<br>~~ fe 2 a<br>so~ ait, artnse aehlF! Sane,we Kathmandu rea’ xy _ = 7 veatn a<br>ee Sha Ga 8 20<br>wee se aise nip = edd g 3<br>‘SSoe an< gh“Raatage ee eetA T A{ a) z<br>Wear iabeetes! 59 78 92 0.5 2.0<br>Ve x 3 =e<br>I bie 3 acs<br>z Fis & = = Wealth<br>2 ei 39 65 -0.7 1.6<br>Higher Wealth — * Wealth PM, ; (g/m?) 3 Wealth<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"<!-- Start of picture text --> A.<br>10<br>@ Total PM,, .<br>@ P\\\\,, with dust and sea salt removed oid<br>Ss Rio de Janeiro<br>S65 \\\\ .<br>-3 \\\\ 6<br>oo Lagos \\\\ we<br>zo wn\\\"<br>o> Vi —— aa<br>ao y gene ee<br>#Booc© arend. aAan aaa Ye<br>EG wos<br>rs} oa Delhi<br>5:<br>B.<br>Qas<br>3<br>= = ore 8 Ke<br>SsoO a X pxa Wealtheal o>—% © Elevation,<br>wo ; . ‘ B . .<br>° .. =a 4<br>an ss)<br>a | ———— a |<br>10 20 -0.25 1.80 0 1,000<br>_ \\\\ 2 ee (| g ae<br><= bi \\\\ Wealth Elevation<br>a<br>U {<br>a<br>100a120 |0.05 165 130 ee320<br>a =z Bins =$ om 4<br>r<br>a | | eee |<br>30 40 -0.75 170 0 110<br>PM,, (ug/m?) Wealth (rwi) Elevation (m)<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"# **3 Discussion**\\n\\nUsing large-scale spatially resolved wealth and pollution data we find that ambient air pollution is, on average, positively correlated with wealth in LMICs around the world. This appears to be driven by urbanization, which is highly positively correlated with both wealth and ambient air pollution in most LMICs. However, there is substantial heterogeneity in this result which appears driven partly by the the weaker correlation between urbanization and ambient air pollution in countries where the primary source of ambient air pollution is non-anthropogenic.\\n\\nWithin cities the theoretically anticipated relationship between pollution and wealth is ambiguous. It will be driven by patterns of residential sorting that are likely influenced by cultural, geographic, and historic differences. As a result, one would expect city-level results to vary widely across different cities in different countries. We find this to be the case. Sorting across different terrains, particularly elevation, appears to be important in places where cities are built on varied geographies. Concentration of economic activity also appears to explain some of the relationship between wealth and pollution levels. Cities with more economic concentration have a stronger and more positive relationships between wealth and pollution. This may be due to the relatively wealthy wishing to live closer to areas of concentrated economic opportunity – which may also be the most polluted areas. This is analogous to the mechanism that appears to drive the correlations between pollution and wealth at a country level – living in close proximity to economic opportunity can come at the cost of living closer to emissions sources.\\n\\nThere are several limitations to our results. First, we do not account for indoor air pollution, which is negatively correlated with urbanization<sup>32</sup> because rural households are more likely to cook with pollution-generating fuels. Relatively higher concentrations of indoor air pollution in rural areas suggests that the total burden of air pollution may be less positively correlated with wealth than the ambient concentrations examined here. In the extreme case, sufficiently high levels of indoor concentrations in rural areas could completely offset the positive correlation of ambient concentrations with wealth and lead to a regressive total burden of air pollution.\\n\\nA second limitation is that we do not observe how ambient pollution levels translate to personal exposures and impacts. Personal exposures are mediated through the built environment,<sup>33</sup> patterns of commuting, and workplace conditions. Individuals living in neighborhoods with the same ambient pollution levels may not be personally exposed to the same amount of pollution because of differences in these conditions. For example, less wealthy neighborhoods may have lower quality housing and may therefore experience higher levels of infiltration than urban areas.<sup>32</sup> Greater wealth may also allow individuals to better insulate themselves from negative impacts for a given level of personal exposure<sup>34</sup> if, for example, wealth is correlated with better health care access. Therefore, even in countries where ambient pollution is higher in wealthier 11\", \"datasets\": [{\"mention_name\": {\"text\": \"large-scale spatially resolved wealth and pollution data\", \"confidence\": 0.6584222912788391, \"start\": 26, \"end\": 82, \"clean_text\": \"large-scale spatially resolved wealth and pollution data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 26, \"end\": 82, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 26, \"end\": 82, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 26, \"end\": 82}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 26, \"end\": 82, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.8767059445381165, \"start\": 26, \"end\": 82}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6584222912788391, \"start\": 26, \"end\": 82}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7733719944953918, \"start\": 26, \"end\": 82}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7733719944953918, \"start\": 26, \"end\": 82}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"areas, the least wealthy members of society may still be most vulnerable to the negative impacts of air pollution.\\n\\nWe rely on spatially resolved estimates of wealth and PM2 _._ 5 that are constructed by model predictions from observed inputs, particularly from satellite imagery. While both sets of predictions are trained using data collected on the ground for areas in which that data exists, they are useful precisely because data does not exist everywhere. In areas without ground collected data their accuracy cannot be verified. We confirm that our results are robust to using alternative measures of wealth and pollution (see Section SI-1.4) for the areas in which that data exists but these are also the areas where the algorithms would be expected to be most accurate. It is by construction uncertain how accurate the wealth and PM2 _._ 5 predictions are in areas for which we have no ground data. If, for example, wealth is systematically underestimated in less polluted areas, or pollution levels are overestimated in wealthy areas, then our approach would overestimate the degree of progressivity in the pollution-wealth relationship. Conversely, if the wealth data systematically underestimates wealth in polluted areas or pollution is underestimated in poorer areas then our approach will underestimate the progressivity.\\n\\nUnderstanding the distribution of ambient air pollution concentrations across wealth levels in LMICs is important for policymakers. The common perception that the least wealthy members of society typically live in the most polluted areas is not consistent with the empirical evidence presented here. Our results indicate the relationship around the world differs from that found in high income countries and varies across countries. The substantial heterogeneity we document highlights that the benefits of reducing ambient air pollution will accrue to different populations in different settings. Finally, our results emphasize that ambient air pollution is a problem in precisely the places that are engines of economic growth in most countries.<sup>24</sup> Failure to address the challenges of ambient air pollution in these settings is likely holding back economic growth.<sup>35</sup> This work leaves open several important questions. In particular, more research is needed to understand if there is a consistent driver of the relationship between wealth and pollution across cities or whether these relationships are explained by idiosyncratic conditions. The answers will inform our understanding of the pollution burden around the world and how cities can be designed to maximize the economic opportunities they offer while minimizing the negative consequences of the pollution they generate.\\n\\n12\", \"datasets\": [{\"mention_name\": {\"text\": \"data collected on the ground\", \"confidence\": 0.3514140546321869, \"start\": 330, \"end\": 358, \"clean_text\": \"data collected on the ground\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 330, \"end\": 358, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 330, \"end\": 358, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 330, \"end\": 358}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 330, \"end\": 358, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6215218305587769, \"start\": 330, \"end\": 358}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3514140546321869, \"start\": 330, \"end\": 358}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.592329740524292, \"start\": 330, \"end\": 358}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.592329740524292, \"start\": 330, \"end\": 358}}, {\"mention_name\": {\"text\": \"wealth data\", \"confidence\": 0.638299286365509, \"start\": 1167, \"end\": 1178, \"clean_text\": \"wealth data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1167, \"end\": 1178, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1167, \"end\": 1178, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1167, \"end\": 1178}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1167, \"end\": 1178, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.6712521314620972, \"start\": 332, \"end\": 343}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.638299286365509, \"start\": 332, \"end\": 343}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6310933232307434, \"start\": 332, \"end\": 343}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6310933232307434, \"start\": 332, \"end\": 343}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"# **4 Materials and Methods**\\n\\n## **4.1 Data**\\n\\n**Wealth data–** Our wealth data comes from recently published estimates of relative wealth on a 2.4 _×_ 2.4 km grid for populated areas in 103 countries around the world.<sup>25</sup> These data use machine learning algorithms to combine satellite data, mobile phone data, and data from Facebook users to estimate relative wealth (RWI) within countries. The algorithms are trained on Demographic and Health Survey (DHS) collected measurements of wealth. The gridded wealth data provide estimates for each grid-cell of how wealthy that grid-cell is relative to others in the same country as well as an estimated error for that estimate. The relative wealth indicator ranges from -2.5 to 2.5 over the full data in our sample, with smaller ranges within individual countries.\\n\\nAs a robustness check we also re-estimate our model with the Demographic and Health Survey (DHS) derived wealth indices that the gridded wealth estimates were trained on.<sup>36</sup> **Pollution data–** We use estimated pollution concentrations from the global dataset produced by the Van Donkelaar research group<sup>29</sup> . This provides monthly average pollution estimates on a 0.01 _×_ 0.01 decimal degree covering the entire planet. We use the average of estimates spanning 2015 to 2020 and for each wealth grid cell assign it the value of the nearest PM2 _._ 5 gridcell. To calculate the average we collapse the monthly pollution estimates in three different ways. In our primary analysis we calculate the average across all years and months from 2015 to 2020. Secondly, we calculate the median across all months and years. Finally, we use the maximum monthly average level of particulates for each grid-cell for the month-of-year that had the highest average pollution from 2015 through 2020. For example, Delhi is assigned the average December PM2 _._ 5 concentration because that is the month-of-year in which average particulate values are highest. In the analysis of the correlation between wealth and particulate pollution from non-dust and sea salt sources we use the dust and sea salt removed estimates produced by the Van Donkelaar group at the annual level and match to RWI points in the same way.\\n\\nAs a robustness check we also re-estimate our models with available ground monitor measurements of PM2 _._ 5. All ground measurements come from the OpenAQ network ( `https: //openaq.org` ).\\n\\n**Elevation data–** We use data from NASA’s ASTER project that provides a global digital elevation model (DEM) that provides elevation at 30m resolution comprehensively around the world.\\n\\n**Pollution source data–** We use data from the EU EDGAR database that provides information on contribution of different economic sectors to particulate pollution for each point on a grid covering the whole planet. We aggregate these sectoral figures into broad categories (e.g. power generation) and calculate the total contribution of each broad category within each coun-\\n\\n13\", \"datasets\": [{\"mention_name\": {\"text\": \"Wealth data\", \"confidence\": 0.4501004219055176, \"start\": 50, \"end\": 61, \"clean_text\": \"Wealth data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 50, \"end\": 61, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 50, \"end\": 61, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 50, \"end\": 61}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 50, \"end\": 61, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9956520199775696, \"start\": 50, \"end\": 61}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4501004219055176, \"start\": 50, \"end\": 61}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6046046018600464, \"start\": 50, \"end\": 61}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6046046018600464, \"start\": 50, \"end\": 61}}, {\"mention_name\": {\"text\": \"data from Facebook users\", \"confidence\": 0.3832252025604248, \"start\": 326, \"end\": 350, \"clean_text\": \"data from Facebook users\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 326, \"end\": 350, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 326, \"end\": 350, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 326, \"end\": 350}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 326, \"end\": 350, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9399489164352417, \"start\": 326, \"end\": 350}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3832252025604248, \"start\": 326, \"end\": 350}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7553380727767944, \"start\": 326, \"end\": 350}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7553380727767944, \"start\": 326, \"end\": 350}}, {\"mention_name\": {\"text\": \"Demographic and Health Survey\", \"confidence\": 0.3434469997882843, \"start\": 433, \"end\": 462, \"clean_text\": \"Demographic and Health Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 433, \"end\": 462, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 433, \"end\": 462, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 433, \"end\": 462}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 433, \"end\": 462, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8781455159187317, \"start\": 433, \"end\": 462}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3434469997882843, \"start\": 433, \"end\": 462}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8947502374649048, \"start\": 433, \"end\": 462}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8947502374649048, \"start\": 433, \"end\": 462}}, {\"mention_name\": {\"text\": \"gridded wealth data\", \"confidence\": 0.47923707962036133, \"start\": 507, \"end\": 526, \"clean_text\": \"gridded wealth data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 507, \"end\": 526, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 507, \"end\": 526, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 507, \"end\": 526}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 507, \"end\": 526, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9954387545585632, \"start\": 507, \"end\": 526}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.47923707962036133, \"start\": 507, \"end\": 526}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.44932594895362854, \"start\": 507, \"end\": 526}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.44932594895362854, \"start\": 507, \"end\": 526}}, {\"mention_name\": {\"text\": \"global dataset\", \"confidence\": 0.558080792427063, \"start\": 1078, \"end\": 1092, \"clean_text\": \"global dataset\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1078, \"end\": 1092, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1078, \"end\": 1092, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1078, \"end\": 1092}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1078, \"end\": 1092, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9938619136810303, \"start\": 373, \"end\": 387}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.558080792427063, \"start\": 373, \"end\": 387}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8767313957214355, \"start\": 373, \"end\": 387}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8767313957214355, \"start\": 373, \"end\": 387}}, {\"mention_name\": {\"text\": \"OpenAQ network\", \"confidence\": 0.36343613266944885, \"start\": 2390, \"end\": 2404, \"clean_text\": \"OpenAQ network\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2390, \"end\": 2404, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2390, \"end\": 2404, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2390, \"end\": 2404}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2390, \"end\": 2404, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.8993343114852905, \"start\": 132, \"end\": 146}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.36343613266944885, \"start\": 132, \"end\": 146}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9890304207801819, \"start\": 132, \"end\": 146}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9890304207801819, \"start\": 132, \"end\": 146}}, {\"mention_name\": {\"text\": \"ASTER project\", \"confidence\": 0.321803480386734, \"start\": 2477, \"end\": 2490, \"clean_text\": \"ASTER project\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2477, \"end\": 2490, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2477, \"end\": 2490, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2477, \"end\": 2490}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2477, \"end\": 2490, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.9886005520820618, \"start\": 219, \"end\": 232}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.321803480386734, \"start\": 219, \"end\": 232}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9108409881591797, \"start\": 219, \"end\": 232}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9108409881591797, \"start\": 219, \"end\": 232}}, {\"mention_name\": {\"text\": \"EU EDGAR database\", \"confidence\": 0.7746491432189941, \"start\": 2669, \"end\": 2686, \"clean_text\": \"EU EDGAR database\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2669, \"end\": 2686, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2669, \"end\": 2686, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2669, \"end\": 2686}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2669, \"end\": 2686, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9906944036483765, \"start\": 411, \"end\": 428}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7746491432189941, \"start\": 411, \"end\": 428}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8727380633354187, \"start\": 411, \"end\": 428}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8727380633354187, \"start\": 411, \"end\": 428}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"try based on the sum of the contributions of that broad category at each grid-cell within the country.\\n\\n**Urbanization data–** We use data measuring urban and rural catchment areas from ref.<sup>37</sup> This data provides a pixel level estimate of the travel time to the nearest urban area (URCA). Pixels within an urban areas are assigned values that indicate they are within an urban area of various sizes. These values range from 1, indicating the pixel is in a city of at least 5 million people, to 30, indicating the pixel is in a rural hinterland. Values in between 1 and 30 indicate the pixel is in an urban area of increasingly smaller sizes (e.g. 20,000-50,000 people) or is less than 1 hours travel, 1-2 hours travel, or more than 2 hours travel time from a city of various sizes (e.g. 1 hour from a city of 20,000-50,000 people). We define pixels that fall within urban areas of any size as urban, those that are less than 1 hours travel from a city of any size as peri-urban, and those that are more than an hour’s travel from a city as rural. These classifications are only for explanatory purposes, in all analysis we use the assigned score from 1-30. To classify RWI pixels we assign it the URCA value of the nearest URCA pixel. The plurality of grid-cells in our sample are in the peri-urban category (Figure SI-6).\\n\\n**Economic concentration–** We use data from the Global Human Settlement Layer (GHSL) data product BUILT-V that measures the built-up volume of a pixel as the number of cubic meters of built volume in that pixel. We use a pixel size of 100m. We collected data on the NRES version of the BUILT-V product that only include non-residential built up area in the calculation of the built volume. For each city in our data we calculate the standard deviation in the BUILT-V-NRES value across the BUILT-V-NRES cells within that city. We normalize this value by dividing by the sum of the BUILT-V-NRES values within the city.\\n\\n**Country-level income–** We use data from the World Bank World Development indicators to measure country level GDP per capita (constant 2015 US$) when we examine the correlation between pollution averaged across the entire country and income averaged across the entire country. We also use the Urban population (% of total population) at the country level in the same analysis. Data available here: `https://data.worldbank.org/` .\\n\\n**Data for additional heterogeneity analyses–** In our analysis of heterogeneity by primary source of pollution we use data from the EU EDGAR database on the amount of PM2 _._ 5 emitted from various different sources on a global grid. We assign grid points to countries (or cities) and calculate the total PM2 _._ 5 emissions from each category of source in the EDGAR data (e.g. air transportation). We then aggregate into broad categories (e.g. transportation, industry, etc.). We then determine the primary source of emissions by summing all the categories and choosing the category that accounts for the largest share of the total. Data available here: `https:// edgar.jrc.ec.europa.eu/` . For heterogeneity by history of colonization and equatorial location we assemble indicators for each country based on a review of the country’s history of colonization\\n\\n14\", \"datasets\": [{\"mention_name\": {\"text\": \"data measuring urban and rural catchment areas\", \"confidence\": 0.7195937633514404, \"start\": 134, \"end\": 180, \"clean_text\": \"data measuring urban and rural catchment areas\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 134, \"end\": 180, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 134, \"end\": 180, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 134, \"end\": 180}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 134, \"end\": 180, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.9953741431236267, \"start\": 134, \"end\": 180}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.7195937633514404, \"start\": 134, \"end\": 180}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8321170806884766, \"start\": 134, \"end\": 180}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8321170806884766, \"start\": 134, \"end\": 180}}, {\"mention_name\": {\"text\": \"Global Human Settlement Layer\", \"confidence\": 0.5568779110908508, \"start\": 1383, \"end\": 1412, \"clean_text\": \"Global Human Settlement Layer\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.34020593762397766, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1383, \"end\": 1412, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1383, \"end\": 1412}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1383, \"end\": 1412, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.6464095711708069, \"start\": 698, \"end\": 727}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5568779110908508, \"start\": 698, \"end\": 727}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8359345197677612, \"start\": 698, \"end\": 727}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8359345197677612, \"start\": 698, \"end\": 727}}, {\"mention_name\": {\"text\": \"World Bank World Development indicators\", \"confidence\": 0.39985743165016174, \"start\": 2000, \"end\": 2039, \"clean_text\": \"World Bank World Development indicators\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2000, \"end\": 2039, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2000, \"end\": 2039, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2000, \"end\": 2039}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2000, \"end\": 2039, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.985338568687439, \"start\": 635, \"end\": 674}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.39985743165016174, \"start\": 635, \"end\": 674}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8769517540931702, \"start\": 635, \"end\": 674}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8769517540931702, \"start\": 635, \"end\": 674}}, {\"mention_name\": {\"text\": \"EU EDGAR database\", \"confidence\": 0.8516243696212769, \"start\": 2519, \"end\": 2536, \"clean_text\": \"EU EDGAR database\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2519, \"end\": 2536, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2519, \"end\": 2536, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2519, \"end\": 2536}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2519, \"end\": 2536, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9854074120521545, \"start\": 342, \"end\": 359}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8516243696212769, \"start\": 342, \"end\": 359}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.590603232383728, \"start\": 342, \"end\": 359}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.590603232383728, \"start\": 342, \"end\": 359}}, {\"mention_name\": {\"text\": \"EDGAR data\", \"confidence\": 0.5278158783912659, \"start\": 2748, \"end\": 2758, \"clean_text\": \"EDGAR data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2748, \"end\": 2758, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2748, \"end\": 2758, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2748, \"end\": 2758}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2748, \"end\": 2758, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.4061202108860016, \"start\": 571, \"end\": 581}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5278158783912659, \"start\": 571, \"end\": 581}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7411676645278931, \"start\": 571, \"end\": 581}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7411676645278931, \"start\": 571, \"end\": 581}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"(or status as a post-Soviet state) from various sources.\\n\\n**Income and pollution in the United Stats–** To assess the correlation between income and ambient pollution in the United States we utilize census tract level median income from the American Community Survey 5-year estimate 2016-2020. Ambient pollution comes from the same PM2 _._ 5 data used in the global analysis but averaged to the census tract level.\\n\\n# **4.2 Methods**\\n\\nTo measure correlation between wealth and ambient concentrations of pollution we estimate fixed effects regressions of the form: where _PMg_ is the measure of pollution at RWI grid-cell _g_ aggregated over our time period in one of the three ways described in the Data section. _RWIg_ is the RWI value at that grid-cell. _ψc_ is a country fixed effect that we include in the pooled regressions (these are omitted in the country specific regressions). We employ various weighting schemes, weighting by: grid-cell population, the inverse of the reported RWI estimation error, the grid-cell population multiplied by the inverse of the error, the population of the country containing the grid-cell, and the inverse of the error times the population of the country containing the error.\\n\\nIn a separate analysis we compare the share of pollution in urban and rural areas coming from dust and sea salt to the share coming from other sources. To do this we first measure the average levels of exposure in rural and urban areas using the total PM2 _._ 5 pollution data. We then repeat this calculation using the dust and sea salt removed measures of PM2 _._ 5. We attribute the difference in these averages to dust and sea salt.\\n\\nTo assess heterogeneity we use several different approaches. To examine heterogeneity based on primary emissions sources, level of per capita income, or colonization history we simply stratify the sample by the relevant measure of heterogeneity and re-estimate equation 1 (omitting the fixed effects) within each country (or city) within each strata.\\n\\nTo assess the role that elevation plays in driving heterogeneity across cities we calculate the correlation between wealth and elevation separately within each city and the correlation between wealth and pollution within each city. We then run a simple OLS regression at the city level with the correlation between wealth and pollution as the dependent variable and the correlation between wealth and elevation as the independent variable. Our estimation, where _i_ denotes cities, takes the form: Corr(Wealth/Pollution) _i_ = _β_ Corr(Wealth/Elevation) _i_ + _ε_ (2) 15\", \"datasets\": [{\"mention_name\": {\"text\": \"American Community Survey\", \"confidence\": 0.663907527923584, \"start\": 241, \"end\": 266, \"clean_text\": \"American Community Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 241, \"end\": 266, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 241, \"end\": 266, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 241, \"end\": 266}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 241, \"end\": 266, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.9137082099914551, \"start\": 241, \"end\": 266}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.663907527923584, \"start\": 241, \"end\": 266}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8317297101020813, \"start\": 241, \"end\": 266}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8317297101020813, \"start\": 241, \"end\": 266}}, {\"mention_name\": {\"text\": \"total PM2 _._ 5 pollution data\", \"confidence\": 0.403210312128067, \"start\": 1464, \"end\": 1494, \"clean_text\": \"total PM2 _._ 5 pollution data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1464, \"end\": 1494, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1464, \"end\": 1494, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1464, \"end\": 1494}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1464, \"end\": 1494, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.2982574701309204, \"start\": 672, \"end\": 702}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.403210312128067, \"start\": 672, \"end\": 702}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9385017156600952, \"start\": 672, \"end\": 702}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9385017156600952, \"start\": 672, \"end\": 702}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"We estimate a similar equation to assess how much of the variation in the correlation between wealth and pollution can be explained by economic concentration. In that specification we replace the _Corr(Wealth/Elevation)_ term on the right hand side of equation 2 with the standard deviation of the volume of non-residential built up area in city _i_ normalized by the total volume of non-residential built up area in city _i_ from the Global Human Settlement layer.\\n\\n# **5 Acknowledgments**\\n\\nWe thank Gabriel Englander, the members of ECHOLab at Stanford, and participants in the 2022 TWEEDS workshop for useful comments. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.\\n\\n# **6 Data Availability**\\n\\nAll data used in the study is publicly available. Links to the original data sources are provided in the methods section.\\n\\n# **7 Code Availability**\\n\\nCode to replicate the results will be made available on the authors’ websites.\\n\\n16\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"# **References**\\n\\n1. Weichenthal, S. _et al._ How low can you go? Air pollution affects mortality at very low levels. _Science Advances_ **8,** eabo3381 (2022).\\n\\n2. Shaddick, G., Thomas, M., Mudu, P., Ruggeri, G. & Gumy, S. Half the world’s population are exposed to increasing air pollution. _NPJ Climate and Atmospheric Science_ **3,** 1–5 (2020).\\n\\n3. Fuller, R. _et al._ Pollution and health: a progress update. _The Lancet Planetary Health_ (2022).\\n\\n4. Richardson, E. A., Pearce, J., Tunstall, H., Mitchell, R. & Shortt, N. K. Particulate air pollution and health inequalities: a Europe-wide ecological analysis. _International journal of health geographics_ **12,** 1–10 (2013).\\n\\n5. Tessum, C. W. _et al._ PM2.5 polluters disproportionately and systemically affect people of color in the United States. _Science Advances_ **7,** eabf4491 (2021).\\n\\n6. Mohai, P., Pellow, D. & Roberts, J. T. Environmental justice. _Annual review of environment and resources_ **34,** 405–430 (2009).\\n\\n7. Clark, L. P., Millet, D. B. & Marshall, J. D. Changes in transportation-related air pollution exposures by race-ethnicity and socioeconomic status: outdoor nitrogen dioxide in the United States in 2000 and 2010. _Environmental health perspectives_ **125,** 097012 (2017).\\n\\n8. Mikati, I., Benson, A. F., Luben, T. J., Sacks, J. D. & Richmond-Bryant, J. Disparities in distribution of particulate matter emission sources by race and poverty status. _American journal of public health_ **108,** 480–485 (2018).\\n\\n9. Jbaily, A. _et al._ Air pollution exposure disparities across US population and income groups. _Nature_ **601,** 228–233 (2022).\\n\\n10. Bell, M. L. & Ebisu, K. Environmental inequality in exposures to airborne particulate matter components in the United States. _Environmental health perspectives_ **120,** 1699– 1704 (2012).\\n\\n11. Miranda, M. L., Edwards, S. E., Keating, M. H. & Paul, C. J. Making the environmental justice grade: the relative burden of air pollution exposure in the United States. _International journal of environmental research and public health_ **8,** 1755–1771 (2011).\\n\\n12. Banzhaf, S., Ma, L. & Timmins, C. Environmental justice: The economics of race, place, and pollution. _Journal of Economic Perspectives_ **33,** 185–208 (2019).\\n\\n13. Liu, J. _et al._ Disparities in Air Pollution Exposure in the United States by Race/Ethnicity and Income, 1990–2010. _Environmental Health Perspectives_ **129,** 127005 (2021).\\n\\n17\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"14. Colmer, J., Hardman, I., Shimshack, J. & Voorheis, J. Disparities in PM2. 5 air pollution in the United States. _Science_ **369,** 575–578 (2020).\\n\\n15. Rosofsky, A., Levy, J. I., Zanobetti, A., Janulewicz, P. & Fabian, M. P. Temporal trends in air pollution exposure inequality in Massachusetts. _Environmental research_ **161,** 76–86 (2018).\\n\\n16. Hajat, A., Hsia, C. & O’Neill, M. S. Socioeconomic disparities and air pollution exposure: a global review. _Current environmental health reports_ **2,** 440–450 (2015).\\n\\n17. Wang, Y., Wang, Y., Xu, H., Zhao, Y. & Marshall, J. D. Ambient air pollution and socioeconomic status in China. _Environmental Health Perspectives_ **130,** 067001 (2022).\\n\\n18. Gupta, S., Das, S. & Murty, M. Quantifying Air Pollution Vulnerability and its Distributional Consequences: Some Perspectives from Delhi. _Ecology, Economy and Society-the INSEE Journal_ **2,** 93–125 (2019).\\n\\n19. Dasgupta, S., Lall, S. V. & Wheeler, D. Traffic, air pollution, and distributional impacts in Dar es Salaam: A spatial analysis with new satellite data. _World Bank Policy Research Working Paper_ (2020).\\n\\n20. Dionisio, K. L. _et al. Air pollution in Accra neighborhoods: spatial, socioeconomic, and temporal patterns_ 2010.\\n\\n21. Rooney, M. S. _et al._ Spatial and temporal patterns of particulate matter sources and pollution in four communities in Accra, Ghana. _Science of the Total Environment_ **435,** 107–114 (2012).\\n\\n22. Liu, Q., Wang, S., Zhang, W., Li, J. & Dong, G. The effect of natural and anthropogenic factors on PM2.5: Empirical evidence from Chinese cities with different income levels. _Science of the Total Environment_ **653,** 157–167 (2019).\\n\\n23. Wu, J.-X., He, L.-Y. & Zhang, Z. On the co-evolution of PM2.5 concentrations and income in China: A joint distribution dynamics approach. _Energy Economics_ **105,** 105706 (2022).\\n\\n24. Glaeser, E. _Triumph of the city: How our greatest invention makes us richer, smarter, greener, healthier, and happier_ (Penguin, 2012).\\n\\n25. Chi, G., Fang, H., Chatterjee, S. & Blumenstock, J. E. Microestimates of wealth for all low-and middle-income countries. _Proceedings of the National Academy of Sciences_ **119** (2022).\\n\\n26. Coker, E. S. _et al._ The effects of air pollution on COVID-19 related mortality in northern Italy. _Environmental and Resource Economics_ **76,** 611–634 (2020).\\n\\n18\", \"datasets\": [{\"mention_name\": {\"text\": \"satellite data\", \"confidence\": 0.3102445900440216, \"start\": 1056, \"end\": 1070, \"clean_text\": \"satellite data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1056, \"end\": 1070, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1056, \"end\": 1070, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1056, \"end\": 1070}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1056, \"end\": 1070, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.9940779209136963, \"start\": 532, \"end\": 546}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3102445900440216, \"start\": 532, \"end\": 546}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.517465353012085, \"start\": 532, \"end\": 546}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.517465353012085, \"start\": 532, \"end\": 546}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"27. Ji, X., Yao, Y. & Long, X. What causes PM2.5 pollution? Cross-economy empirical analysis from socioeconomic perspective. _Energy Policy_ **119,** 458–472 (2018).\\n\\n28. Li, G., Fang, C., Wang, S. & Sun, S. The effect of economic growth, urbanization, and industrialization on fine particulate matter (PM2. 5) concentrations in China. _Environmental science & technology_ **50,** 11452–11459 (2016).\\n\\n29. Van Donkelaar, A., Martin, R. V., Li, C. & Burnett, R. T. Regional estimates of chemical composition of fine particulate matter using a combined geoscience-statistical method with information from satellites, models, and monitors. _Environmental science & technology_ **53,** 2595–2611 (2019).\\n\\n30. Evans, M. _et al._ Policy findings from the DACCIWA Project (2018).\\n\\n31. Toure, N. O. _et al._ Observed and modeled seasonal air quality and respiratory health in Senegal during 2015 and 2016. _GeoHealth_ **3,** 423–442 (2019).\\n\\n32. Lim, S. _et al._ Comparing human exposure to fine particulate matter in low and high-income countries: A systematic review of studies measuring personal PM2.5 exposure. _Science of The Total Environment,_ 155207 (2022).\\n\\n33. Burke, M. _et al. Exposures and behavioral responses to wildfire smoke_ tech. rep. (National Bureau of Economic Research, 2021).\\n\\n34. Behrer, A., Park, R., Wagner, G., Golja, C. & Keith, D. Heat has larger impacts on labor in poorer areas. _Environmental Research Communications_ **3,** 095001 (2021).\\n\\n35. Dechezleprˆetre, A., Rivers, N. & Stadler, B. The economic cost of air pollution: Evidence from Europe (2019).\\n\\n36. Rutstein, S. O. & Johnson, K. The DHS wealth index. DHS comparative reports no. 6. _Calverton, Md: ORC Macro_ (2004).\\n\\n37. Cattaneo, A., Nelson, A. & McMenomy, T. Global mapping of urban–rural catchment areas reveals unequal access to services. _Proceedings of the National Academy of Sciences_ **118** (2021).\\n\\n19\", \"datasets\": [{\"mention_name\": {\"text\": \"DHS wealth index\", \"confidence\": 0.6219590902328491, \"start\": 1620, \"end\": 1636, \"clean_text\": \"DHS wealth index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1620, \"end\": 1636, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1620, \"end\": 1636, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1620, \"end\": 1636}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1620, \"end\": 1636, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.8789684176445007, \"start\": 379, \"end\": 395}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6219590902328491, \"start\": 379, \"end\": 395}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7203290462493896, \"start\": 379, \"end\": 395}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7203290462493896, \"start\": 379, \"end\": 395}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"# **Supplementary Information**\\n\\n# **SI-1 Supplementary text**\\n\\n## **SI-1.1 Country level correlations**\\n\\nTo provide context for the relationship between wealth and pollution globally we draw on data from the World Bank to estimate the country-level correlation across the entire world The general intuition that pollution is higher in lower income countries is borne out by the country-level data. Two results stand out: first, as has been well documented, countrylevel average pollution is lower on average in higher income countries (Figure SI-1a). Only two countries with a per capita income above $50,000 have average pollution levels above 20 _μg/m_<sup>3</sup> (Macau and Qatar). Countries with low levels of per capita income (less than $15,000) span the full range of pollution levels but the majority have pollution concentrations above 25 _μg/m_<sup>3</sup> . Simple cross-country regression confirms this correlation, every $10,000 increase in per capita income, is associated with a decline in pollution of 1.8 _μg/m_<sup>3</sup> ( _t-stat_ : 4.3).\\n\\nSecond, consistent with the negative relationship between income and pollution the most polluted countries in the world are concentrated in Central and West Africa, the Middle East, and South Asia (Figure SI-1b-c). With the exception of the Middle East, these are among the lowest income regions in the world. Of the four countries that are notable outliers with pollution levels higher than would be expected given their income levels (labeled in Figure SI-1a), three are highly urbanized city states (The United Arab Emirates, Macau, and Monaco). The fourth, Qatar, has a significant portion of air pollution coming from natural pollution sources like dust storms.\\n\\n## **SI-1.2 Heterogeniety by per capita income and pollution source**\\n\\nThe spatial pattern of ambient particulate pollution, and hence its correlation with wealth, may also vary depending on the primary sources of particulate pollution and on country-level income. For example, in places where transportation is the primary source of emissions, high pollution areas may be less spatially concentrated than in places where pollution is primarily generated by industrial activity which is more likely to be concentrated in a handful of cities. Because the nature of economic activity changes at different levels of development<sup>1</sup> these spatial patterns may also vary by country-level per capita income.\\n\\nWe do not find evidence that suggests either of these hypotheses can explain our country level results. We observe similar patterns of heterogeneity across countries when we group by average per capita GDP or by the primary source of particulate pollution in the country.\\n\\n20\", \"datasets\": [{\"mention_name\": {\"text\": \"data from the World Bank\", \"confidence\": 0.7596397995948792, \"start\": 195, \"end\": 219, \"clean_text\": \"data from the World Bank\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 195, \"end\": 219, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 195, \"end\": 219, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 195, \"end\": 219}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 195, \"end\": 219, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9447398781776428, \"start\": 195, \"end\": 219}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.7596397995948792, \"start\": 195, \"end\": 219}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.47273632884025574, \"start\": 195, \"end\": 219}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.47273632884025574, \"start\": 195, \"end\": 219}}, {\"mention_name\": {\"text\": \"country-level data\", \"confidence\": 0.7665692567825317, \"start\": 379, \"end\": 397, \"clean_text\": \"country-level data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 379, \"end\": 397, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 379, \"end\": 397, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 379, \"end\": 397}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 379, \"end\": 397, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9447398781776428, \"start\": 379, \"end\": 397}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7665692567825317, \"start\": 379, \"end\": 397}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.47273632884025574, \"start\": 379, \"end\": 397}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.47273632884025574, \"start\": 379, \"end\": 397}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"Dividing countries into five bins based on per capita GDP (from $0-$2,000 to _>_ $8,000, all USD2015) we find that within every bin the range of correlations spans strongly negative to strongly positive (Figure SI-3a) suggesting that income alone does not explain the observed patterns. When we group countries based on their primary sources of particulate emission we observe similar patterns. Heterogeneity persists whether the primary source of pollution is the power, transport, industrial, or agricultural sectors (Figure SI-3c). While re-estimating the correlation between wealth and ambient pollution using the measure of PM2 _._ 5 that removes dust and sea salt substantially reduces the number of countries with a negative relationship across all categories, there still do not appear to be clear patterns in the relationship between wealth and anthropogenic pollution across GDP or primary pollution sources (Figure SI-3b & d).\\n\\n# **SI-1.3 City level heterogeneity**\\n\\nIn seeking to explain the observed heterogeneity across cities we first classify city heterogeneity separately by strength (proxied by statistical significance) and by sign (whether relationship is positive or negative). We do this to account for the possibility that the drivers of these two different measures of heterogeneity may not be the same. For example, cities in which the relationship is weak and positive may be more qualitatively similar to cities with a weak but negative relationship than to cities with a strong positive relationship.\\n\\nTo better understand these dimensions of heterogeneity we focus on understanding three components of cities - the degree of pollution inequality, the degree of wealth inequality, and the proximity of wealthy neighborhoods to high pollution areas. In some cities pollution and wealth vary dramatically across neighborhoods while in other cities both are more uniformly distributed. Among cities with large ranges of wealth and pollution levels, settlement patterns could have resulted in highly different proximities of wealthy neighborhoods to polluted areas. Ultimately, understanding city-level heterogeneity requires understanding why some cities have high levels of variation in pollution while others do not, why there is large variation in wealth levels in some cities and not others, and, in cities with large variation in both wealth and pollution, why settlement patterns have resulted in the wealthy living in relatively higher pollution neighborhoods in some cities and relatively lower pollution neighborhoods in other cities.\\n\\nWe explore several potential answers to these questions.\\n\\nIn the main text we highlight that elevation does appear to be an important determinant in some cities, especially those with substantial changes in elevation within the city limits. This is because pollution is consistently lower at higher elevations. How wealth sorts across the elevation gradient varies by city however. In some, like Santiago, Chile the wealthy live at higher elevations.<sup>2</sup> In others, like Tbilisi, Georgia, the opposite is true. As a result, sometimes cities with a large change in elevation exhibit a positive relationship between wealth and pollution 21\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"while in others the relationship is negative.\\n\\nConcentration of economic activity may also play an important role in determining how much variation exists in pollution levels across a city and may be related to the fact that pollution is lower at higher elevations. In our data elevation is correlated with terrain roughness. Increased terrain roughness may make it more expensive to build in high elevation areas and so these locations may have fewer local sources of pollution. We measure concentration of economic activity as the standard deviation of the volume of non-residential built up area normalized by the total volume from the Global Human Settlement layer. This provides an estimate of how much the volume of non-residential built area varies across 100m cells covering each city. We observe that within cities elevation and terrain roughness both have a significant negative correlation with the volume of non-residential built area.\\n\\nWe also find that cities that have more concentration in their economic activity have larger and more positive correlations between wealth and pollution levels. A standard deviation increase in our measure of concentration is associated with a 0.12 standard deviation change [ _t_ -stat: 1.71] in the correlation between wealth and pollution. This increases to 0.22 standard deviations [ _t_ -stat: 1.88] if we focus only on the cities with a correlation between wealth and pollution that is significant at 5%. This relationship persists when we control for correlation between elevation and wealth.\\n\\nHistorical conditions are another potential factor that could influence all three components of cities today. European colonization has had a persistent impact on settlement patterns in many LMICs around the world.<sup>3,4</sup> In particular, choices about where colonizing rulers chose to live may have led to different patterns in where the wealthy live today. These choices may have been driven in part by the presence of tropical diseases<sup>3</sup> which would imply different patterns in equatorial countries. In addition, patterns of urban settlement in the Soviet Union may have evolved differently than non-Soviet countries due to central planning of urban development<sup>5,6</sup> in part because the allocation of apartments based on employment in economically prioritized sectors may have resulted in less sorting by wealth than in non-Soviet locations that has persisted.<sup>7</sup> We examine whether any of these factors explain the patterns of heterogeneity that we measure. We find that they do not. There is little difference in the pattern of heterogeneity with respect either to magnitude, significance or sign within city groups based on their containing country’s history of European colonization (Figures SI-11 & SI-12), whether they are located in equatorial regions, or whether they were part of the Soviet Union. Like with the country level results there also do not appear to be differences based on the primary source of particulate pollution within the city (Figure SI-13).\\n\\nAs a contrast to our city level results in LMICs, in Figure SI-10 we examine these relationships in three of the largest cities in the United States (Los Angeles, Houston, and New York City).\\n\\n22\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"In all three of these cases the relationships we observe are less stark than in the LMICs in our sample. This is consistent with differential patterns of residential sorting in the United States – – in particular, greater suburbanization in the United States leading to lower pollution exposure<sup>8</sup> driving a different relationship between pollution exposure and wealth. It is also consistent with a relatively weak relationship between income and ambient pollution exposure in the United States.<sup>9</sup>\\n\\n# **SI-1.4 Robustness of Results**\\n\\nOur results are robust to a variety of alternative specification and data choices. Measuring pollution as the median across months and years, rather than the average, in our sample results in broadly similar results (Table SI-1). The choice of weighting scheme appears to matter more when pollution is measured as the median and the magnitude of the correlation is somewhat smaller. In particular, weighting by grid-cell population eliminates the correlation in the pooled sample when using the median. This suggests there are a few grid-cells with very high populations that do not exhibit any correlation between wealth and median pollution but do exhibit a correlation when pollution exposure is measured using the average level of pollution. We believe that using average pollution is a more appropriate measure than median pollution because the average is sensitive to extremely high values of pollution that may be particularly harmful (e.g. spikes in Delhi during the crop burning season).\\n\\nAlternatively, we can calculate the average level of pollution by month for each grid-cell in our sample and assign the average value of the month with the highest average level of pollution to each grid-cell. This approach focuses heavily on the spikes that might occur as a result of seasonal crop burning or forest clearing or due to seasonal increases in dust storms. Using this approach yields results that are qualitatively similar to using the average across the full sample but the magnitudes of the correlation are larger in the maximum approach (Table SI-1). Comparing the country specific estimates of the correlation between wealth and pollution exposure, Figure SI-4, indicates broad agreement in the estimated correlation when exposure is measured as median or average levels of pollution. Most countries fall on or near the 45<sup>_◦_</sup> line.\\n\\nThere is less agreement between the maximum measure and the average measure however. At the extremes the qualitative patterns seem to persist - Nigeria for example has a negative correlation with both measures and has the strongest negative correlation in the sample. The notable exception is India and Pakistan (which are treated as a single unit in the RWI data, a choice we preserve), which has a substantially higher positive correlation using the maximum measure rather than the average measure. This is likely driven by very high levels of pollution in some North Indian cities during December. One possible explanation for these somewhat divergent results using the maximum measure rather than the average measure is that the 23\", \"datasets\": [{\"mention_name\": {\"text\": \"LMICs\", \"confidence\": 0.8508294820785522, \"start\": 84, \"end\": 89, \"clean_text\": \"LMICs\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.4856620728969574, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 84, \"end\": 89, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 84, \"end\": 89}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 84, \"end\": 89, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.2979871332645416, \"start\": 84, \"end\": 89}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8508294820785522, \"start\": 84, \"end\": 89}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8016486763954163, \"start\": 84, \"end\": 89}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8016486763954163, \"start\": 84, \"end\": 89}}, {\"mention_name\": {\"text\": \"RWI data\", \"confidence\": 0.842110276222229, \"start\": 2770, \"end\": 2778, \"clean_text\": \"RWI data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2770, \"end\": 2778, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2770, \"end\": 2778, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2770, \"end\": 2778}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2770, \"end\": 2778, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.4612599313259125, \"start\": 1090, \"end\": 1098}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.842110276222229, \"start\": 1090, \"end\": 1098}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7587634325027466, \"start\": 1090, \"end\": 1098}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7587634325027466, \"start\": 1090, \"end\": 1098}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"maximum measure captures large, non- or semi-anthropogenic sources of pollution that may spike in certain months (e.g. seasonal forest burning).\\n\\nWe also examine whether our results are sensitive to alternative data on wealth, using data from DHS clusters, or pollution, using ground based monitor data. In both cases the sample we are able to analyze using the alternative datasets is smaller than our global sample because of a lack of comprehensive data covering all 103 countries in our sample.\\n\\nWhen we estimate the country-specific correlation using DHS wealth data and our global pollution data set we find the same general pattern of correlation with roughly half the countries exhibiting a positive relationship between wealth and pollution (Figure SI-7). Nigeria and Nepal remain the countries with the most negative and most positive correlations respectively. There are some countries where the sign of the correlation flips when using the DHS data relative to RWI but in general there is broad agreement in sign and magnitude across the two data sets (Figure SI-8).\\n\\nWe also examine the trend in pollution within the seven countries for which we have pollution data from ground monitors that represent at least 10 distinct RWI grid-cells. Figure SI-9 shows that in most of these countries, including all of those for which we have the largest number of ground monitors, the relationship between pollution and wealth is qualitatively similar when calculated using satellite measures of pollution or ground monitors. In the same figure we highlight which ground monitors are in urban or rural settings to underline the point that restricting analysis to those places with ground monitors will often result in the omission of most rural locations and so will fail to capture the urban-rural gradient that appears to be at the heart of our primary results. In several of the countries included in this sample there is only data from one or two rural ground monitors for the entire country. There is also often substantial temporal gaps in ground monitor data, on average ground monitors only report measurements in 45% of the months over which we have satellite measures of pollution. This can also lead to substantial bias when using ground monitor data if the ground monitor data is missing observations during particular seasons when pollution spikes (or is only available in those seasons).\\n\\nIn our city level results the choice of city boundaries is an important determinant of the sign and magnitude of the within-city correlation. The overall pattern of heterogeneity remains regardless of which precise city boundary is chosen but estimates for individual cities do change. In contrast, the threshold for the minimum level of RWI points within a city does not change our results in the range from 10 to 50.\\n\\nIn Figure SI-14 we show how the point estimates change for three cities in our data for which we have both the URCA designated boundaries as well as administrative boundaries. In some cases (e.g. Delhi) the URCA and administrative boundaries align relatively closely but in 24\", \"datasets\": [{\"mention_name\": {\"text\": \"DHS clusters\", \"confidence\": 0.5387709140777588, \"start\": 243, \"end\": 255, \"clean_text\": \"DHS clusters\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 243, \"end\": 255, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 243, \"end\": 255, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 243, \"end\": 255}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 243, \"end\": 255, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.5396518111228943, \"start\": 243, \"end\": 255}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5387709140777588, \"start\": 243, \"end\": 255}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.49980637431144714, \"start\": 243, \"end\": 255}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.49980637431144714, \"start\": 243, \"end\": 255}}, {\"mention_name\": {\"text\": \"ground based monitor data\", \"confidence\": 0.4133147895336151, \"start\": 277, \"end\": 302, \"clean_text\": \"ground based monitor data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 277, \"end\": 302, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 277, \"end\": 302, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 277, \"end\": 302}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 277, \"end\": 302, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.5396518111228943, \"start\": 277, \"end\": 302}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4133147895336151, \"start\": 277, \"end\": 302}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.49980637431144714, \"start\": 277, \"end\": 302}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.49980637431144714, \"start\": 277, \"end\": 302}}, {\"mention_name\": {\"text\": \"DHS wealth data\", \"confidence\": 0.6273350119590759, \"start\": 556, \"end\": 571, \"clean_text\": \"DHS wealth data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 556, \"end\": 571, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 556, \"end\": 571, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 556, \"end\": 571}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 556, \"end\": 571, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9510234594345093, \"start\": 556, \"end\": 571}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6273350119590759, \"start\": 556, \"end\": 571}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7457561492919922, \"start\": 556, \"end\": 571}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7457561492919922, \"start\": 556, \"end\": 571}}, {\"mention_name\": {\"text\": \"global pollution data set\", \"confidence\": 0.47508370876312256, \"start\": 580, \"end\": 605, \"clean_text\": \"global pollution data set\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 580, \"end\": 605, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 580, \"end\": 605, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 580, \"end\": 605}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 580, \"end\": 605, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9510234594345093, \"start\": 580, \"end\": 605}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.47508370876312256, \"start\": 580, \"end\": 605}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7457561492919922, \"start\": 580, \"end\": 605}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7457561492919922, \"start\": 580, \"end\": 605}}, {\"mention_name\": {\"text\": \"DHS data\", \"confidence\": 0.8162306547164917, \"start\": 952, \"end\": 960, \"clean_text\": \"DHS data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 952, \"end\": 960, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 952, \"end\": 960, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 952, \"end\": 960}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 952, \"end\": 960, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.320541650056839, \"start\": 80, \"end\": 88}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8162306547164917, \"start\": 80, \"end\": 88}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.864192008972168, \"start\": 80, \"end\": 88}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.864192008972168, \"start\": 80, \"end\": 88}}, {\"mention_name\": {\"text\": \"pollution data from ground monitors\", \"confidence\": 0.7053528428077698, \"start\": 1164, \"end\": 1199, \"clean_text\": \"pollution data from ground monitors\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1164, \"end\": 1199, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1164, \"end\": 1199, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1164, \"end\": 1199}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1164, \"end\": 1199, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.4079258441925049, \"start\": 292, \"end\": 327}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.7053528428077698, \"start\": 292, \"end\": 327}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6950193047523499, \"start\": 292, \"end\": 327}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6950193047523499, \"start\": 292, \"end\": 327}}, {\"mention_name\": {\"text\": \"ground monitor data\", \"confidence\": 0.9105110168457031, \"start\": 2244, \"end\": 2263, \"clean_text\": \"ground monitor data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2244, \"end\": 2263, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2244, \"end\": 2263, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2244, \"end\": 2263}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2244, \"end\": 2263, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.7792235016822815, \"start\": 503, \"end\": 522}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9105110168457031, \"start\": 503, \"end\": 522}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9154005646705627, \"start\": 503, \"end\": 522}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9154005646705627, \"start\": 503, \"end\": 522}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"others they differ sharply (e.g. Rio de Janeiro). In general the URCA designated boundaries we use align well with boundaries drawn using other gridded products (e.g. Global Human Settlement Layer and the World Settlement Layer). How these boundaries are defined is both somewhat arbitrary and, as the changing signs in the cases of Lagos and Rio de Janeiro show, important for the estimation of correlations. This is true, at least in part, because a more geographically expansive definition of a city’s boundaries is more likely to capture differences in ambient pollution levels driven by the gradient in pollution and wealth between urban and peri-urban/rural that we highlight in our country level results.\\n\\n## **SI-1.5 Kuznets curves discussion**\\n\\nTaking advantage of the granularity of our data we examine whether they provide evidence for the existence of Environmental Kuznets Curves (EKC) within the countries in our sample. We do not find compelling evidence that the relationship between pollution exposure and wealth fits the hypothesized inverted U pattern either across the full sample or within individual countries.\\n\\nFigure SI-15 plots the average level of pollution for all grid-cells within various RWI bands across all the countries in our sample. In Figure SI-16 we plot the same for the ten most populous countries in our sample. Lighter grey point represent the relationship at the most dis-aggregated level of RWI while green and blue dots indicate results averaging within larger bandwidths of RWI. In all cases we fail to see strong evidence of an inverted U. The relationships that we do observe vary across countries, consistent with our previous results. This variation includes countries in which pollution levels are higher at higher wealth levels to those with flat to declining or declining pollution levels moving up the wealth distribution depending on the country.\\n\\n# **SI-2 Supplementary Tables**\\n\\n25\", \"datasets\": [{\"mention_name\": {\"text\": \"Global Human Settlement Layer\", \"confidence\": 0.7182708382606506, \"start\": 167, \"end\": 196, \"clean_text\": \"Global Human Settlement Layer\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 167, \"end\": 196, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 167, \"end\": 196, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 167, \"end\": 196}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 167, \"end\": 196, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.4168665111064911, \"start\": 167, \"end\": 196}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7182708382606506, \"start\": 167, \"end\": 196}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4510380029678345, \"start\": 167, \"end\": 196}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4510380029678345, \"start\": 167, \"end\": 196}}, {\"mention_name\": {\"text\": \"World Settlement Layer\", \"confidence\": 0.6407893896102905, \"start\": 205, \"end\": 227, \"clean_text\": \"World Settlement Layer\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 205, \"end\": 227, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 205, \"end\": 227, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 205, \"end\": 227}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 205, \"end\": 227, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.4168665111064911, \"start\": 205, \"end\": 227}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6407893896102905, \"start\": 205, \"end\": 227}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4510380029678345, \"start\": 205, \"end\": 227}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4510380029678345, \"start\": 205, \"end\": 227}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"# **Table SI-1:** Correlation between RWI and PM2 _._ 5\\n\\n||Unweighted|Population<br>weighted|Country<br>population<br>weight|Error<br>weighted|Jointly<br>weighted|Country<br>jointly<br>weighted|\\n|---|---|---|---|---|---|---|\\n|(A) PM2_._5 measured as average|||||||\\n|Wealth Index|0.748<sup>_∗∗∗_</sup>|0.549<sup>_∗∗∗_</sup>|1.851<sup>_∗∗∗_</sup>|0.643<sup>_∗∗∗_</sup>|0.467<sup>_∗∗∗_</sup>|1.701<sup>_∗∗∗_</sup>|\\n||(0.018)|(0.098)|(0.060)|(0.019)|(0.094)|(0.058)|\\n||[_<_0.00009]|[_<_0.00009]|[_<_0.00009]|[_<_0.00009]|[_<_0.00009]|[_<_0.00009]|\\n|N|2,672,997|2,672,372|2,672,997|2,672,997|2,672,372|2,672,997|\\n|(B) PM2_._5 measured as median|||||||\\n|Wealth Index|0.421<sup>_∗∗∗_</sup>|0.047|0.412<sup>_∗∗∗_</sup>|0.350<sup>_∗∗∗_</sup>|-0.001|0.321<sup>_∗∗∗_</sup>|\\n||(0.016)|(0.081)|(0.050)|(0.016)|(0.077)|(0.050)|\\n||[_<_0.00009]|[0.5574]|[_<_0.00009]|[_<_0.00009]|[0.9859]|[_<_0.00009]|\\n|N<br>26|2,672,997|2,672,372|2,672,997|2,672,997|2,672,372|2,672,997|\\n|(C) PM2_._5 measured as average during maximum month|||||||\\n|Wealth Index|2.269<sup>_∗∗∗_</sup>|1.537<sup>_∗∗∗_</sup>|8.496<sup>_∗∗∗_</sup>|2.025<sup>_∗∗∗_</sup>|1.534<sup>_∗∗∗_</sup>|8.278<sup>_∗∗∗_</sup>|\\n||(0.032)|(0.208)|(0.119)|(0.031)|(0.194)|(0.113)|\\n||[_<_0.00009]|[_<_0.00009]|[_<_0.00009]|[_<_0.00009]|[_<_0.00009]|[_<_0.00009]|\\n|N|2,679,731|2,678,430|2,679,731|2,679,731|2,678,430|2,679,731|\\n|**Fixed Efects:**|||||||\\n|Country|Y|Y|Y|Y|Y|Y|\\n\\nNotes: Each column reports the results of a linear fixed effects regression of pollution against RWI at the grid-cell level across all countries in our sample. Each row is a seperate regression. RWI is a country specific wealth index that ranges across our full dataset from -2.022 to 2.456. We collapse pollution across all months and years in the sample by RWI grid-cell. In row **A** we collapse as the average across all months and years. In row **B** we collapse as the median. In row **C** we calculate the monthly average across all years and assign grid-cells the average in the month with the highest monthly average. Population weights by the population at the RWI grid-cells. Error weights by the inverse of the error reported by the creators of the RWI data for each RWI estimate. Joint weights by both population and the inverse of the error. Country population weights by the total population of the countries in which the grid-cell is located. Country jointly weights by the country population and the inverse of the error. _p_ -values reported in brackets. (* p _<_ .10 ** p _<_ .05 *** p _<_ .01).\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [27]}}, {\"page\": 28, \"chunk\": 28, \"input_text\": \"# **Table SI-2:** Correlation between RWI and PM2 _._ 5, winsored\\n\\n||||Top &|bottom 5% s|ample|\\n|---|---|---|---|---|---|\\n||Winsored across|Winsored within|Pooled|Bottom 5%|Top 5%|\\n|(A) PM2_._5 measured as average||||||\\n|Wealth Index|2.098<sup>_∗∗∗_</sup>|2.158<sup>_∗∗∗_</sup>|-0.035|-6.721<sup>_∗∗∗_</sup>|-2.383<sup>_∗∗∗_</sup>|\\n||(0.064)|(0.064)|(0.087)|(0.924)|(0.539)|\\n||[_<_0.00009]|[_<_0.00009]|[0.6920]|[_<_0.00009]|[_<_0.00009]|\\n|N|2,672,997|2,672,997|268,145|134,482|133,663|\\n|(B) PM2_._5 measured as median||||||\\n|Wealth Index|0.516<sup>_∗∗∗_</sup>|0.573<sup>_∗∗∗_</sup>|-0.955<sup>_∗∗∗_</sup>|-6.727<sup>_∗∗∗_</sup>|-2.523<sup>_∗∗∗_</sup>|\\n||(0.055)|(0.055)|(0.075)|(0.840)|(0.445)|\\n||[_<_0.00009]|[_<_0.00009]|[_<_0.00009]|[_<_0.00009]|[_<_0.00009]|\\n|N<br>2|2,672,997|2,672,997|268,145|134,482|133,663|\\n|(C) PM2_._5 measured as average during maximum month<br>7||||||\\n|Wealth Index|9.664<sup>_∗∗∗_</sup>|9.705<sup>_∗∗∗_</sup>|4.681<sup>_∗∗∗_</sup>|0.599|-2.761<sup>_∗∗∗_</sup>|\\n||(0.124)|(0.124)|(0.164)|(1.628)|(1.045)|\\n||[_<_0.00009]|[_<_0.00009]|[_<_0.00009]|[0.7129]|[0.0082]|\\n|N|2,679,731|2,679,731|268,842|134,646|134,196|\\n|**Fixed Efects:**||||||\\n|Country|Y|Y|Y|Y|Y|\\n\\nNotes: Each column reports the results of a linear fixed effects regression of pollution against RWI at the grid-cell level across all countries in our sample. Each row is a seperate regression. RWI is a country specific wealth index that ranges across our full dataset from -2.022 to 2.456. We collapse pollution across all months and years in the sample by RWI grid-cell. In row **A** we collapse as the average across all months and years. In row **B** we collapse as the median. In row **C** we calculate the monthly average across all years and assign grid-cells the average in the month with the highest monthly average. We winsorize at the fifth and ninety-fifth percentiles of the RWI distribution across the whole sample in the Winsored across column and at the same thresholds within countries in the Winsored within column. In columns 3-5 we calculate the top and bottom 5% of RWI points within each country and only include the points that fall into the categories indicated in the column headings. We weight all regressions by the inverse of the error reported in the RWI data multiplied by the population containing the grid point. _p_ -values reported in brackets. (* p _<_ .10 ** p _<_ .05 *** p _<_ .01).\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [28]}}, {\"page\": 29, \"chunk\": 29, \"input_text\": \"**Table SI-3:** Colonization history and city correlations\\n\\n||Base|Dust and<br>SS removed|Base|Dust and<br>SS removed|Base|Dust and<br>SS removed|\\n|---|---|---|---|---|---|---|\\n|British|0.174|-0.092|||||\\n||(0.269)|(0.331)|||||\\n||[]0.5179|[]0.7809|||||\\n|French|-0.039|-0.267|||||\\n||(0.418)|(0.515)|||||\\n||[]0.9253|[]0.6043|||||\\n|Spanish/Portuguese|0.098|-0.592<sup>_∗_</sup>|||||\\n||(0.287)|(0.353)|||||\\n||[]0.7333|[]0.0952|||||\\n|Other|-0.242|1.176<sup>_∗∗_</sup>|||||\\n||(0.372)|(0.458)|||||\\n||[]0.5161|[]0.0107|||||\\n|Equitorial|||0.350<sup>_∗_</sup>|0.261|||\\n||||(0.208)|(0.263)|||\\n||||[]0.0934|[]0.3225|||\\n|Post-soviet|||||0.007|-0.239|\\n||||||(0.271)|(0.343)|\\n||||||[]0.9782|[]0.4853|\\n|N|273|273|273|273|273|273|\\n\\nNotes: The unit of observation for all regressions is a city in our data. All columns regress the coefficient for the within-city correlation between RWI and average PM2 _._ 5 as calculated with or without dust and sea salt (denoted in column headings) against a series of dummies indicating the colonization history of the country containing the city. In the first and second column the omitted category are countries that were never formally colonized by a European power prior to WWI. Equitorial countries are those between the Tropic of Capricorn and the Tropic of Cancer. Post-soviet countries are those that were formally part of the Soviet Union. _p_ -values reported in brackets. (* p _<_ .10 ** p _<_ .05 *** p _<_ .01).\\n\\n28\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [29]}}, {\"page\": 30, \"chunk\": 30, \"input_text\": \"<!-- Start of picture text --> « \\\\.¢<br>QAT Pt<br>. a<br>a- Percent¥of the populationAliving in urban settingPy<br>15 | Se<br>C.<br>ie ARE %<br>S w<br>ae 50 7 “ Average annual PM2.5<br>D. 5 10 8 2 25 30 % # 4 50<br>o ___ |<br>a »<br>o) B2<br>Z t Suey Ag<br>25- q<br>MAC . ° sy<br>* Average per capita income, 2016-2020, 2015USD, 000s<br>|__|<br>MCO.<br>@) -<br>0 50000 100000 150000<br>Average per capita income,<br>2016-2020, 2015USD<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [30]}}, {\"page\": 31, \"chunk\": 31, \"input_text\": \"<!-- Start of picture text --> Full Sample Rural Peri-Urban Urban<br>100<br>80 ate<br>40<br>20<br>——_ __ _— —_ a rr a<br>-1 05 0 O05 1 15 1 -05 0 05 1 15 1-05 0 05 1 15 1-050 0 05 10 15<br>Wealth Wealth Wealth Wealth<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [31]}}, {\"page\": 32, \"chunk\": 32, \"input_text\": \"<!-- Start of picture text --> Per capita GDP groupings Primary PM, source groupings<br>PM,5 with dust and PM, with dust and<br>Total PM,5 sea salt removed Total PM... sea salt removed<br>o +, <$2000 ° % ° *, Power sector ° *,<br>o $2000-3999 ° ry @ Industry rs<br>$ $4000-5999 *% 3 ¢ Transportation $<br>% $6000-7999 3 3 Agriculture *%<br>3¢>$8000 3e3e3e<br>ee +4 eA ¢<br>° bg2 °°° oa @<br>«¢*<br>¢¢ 2 ¢ @° 3<br>*e<br>e4 * +5 4 °g4e¢2 ¢ 3 3<br>°¢ 4 59 3<br>4 64$ 2 3b4<br>¢9 ¢ B 9 D43<br>°¢4¢ ° 3<br>9? .4 2 3<br>ry9Ps 4+¢ 32>rs 3333<br>° $ Bg 3<br>° 3 e 4<br>+ a °° a4@ Pge<br>oaoa % o o ° ea3¢ 3eSes<br>° A @ °<br>°¢ $4 59¢ 23<br>+ 4 2 ey<br>¢ $ @ 9<br>¢ 3 @ °<br>¢ A e 54<br>+° AA oe 4 e2<br>+° °AS ee @e<br>* > e *<br>+ o 4 °<br>+° e> 54e e*<br>°« e ea* 7ba<br>° o « ¢<br>+ ° . ¢<br>@ * o e<br>2 + « e<br>° > « +<br>° 4 ° 4<br>o¢¥s >%3 ° ° ¢ ¢+.+ o<br>°s+ °o o ° °<br>° + + o o o> o o e e¢<br>+ o ° 3<br>¢ « 4 °<br>¢ + + °<br>2 4 ° °<br>2 > o<br>> ? 4 o<br>>Ra°° ,y}°ea +ca°° 4eo ry<br>a Py a ¢<br>a e° °+44 er¢PY<br>b4a454g® e°oee<br>¢@<br>3<br>¢¢<br>. aAoaAo . ¢+ ° +<br>6 0 5 10-5 0 5 10 5 0) 5 10-5 0 5 10<br>Correlation between wealth Correlation between wealth Correlation between wealth Correlation between wealth<br>and average pollution and average pollution and average pollution and average pollution<br>with dust and sea salt removed with dust and sea salt removed<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [32]}}, {\"page\": 33, \"chunk\": 33, \"input_text\": \"<!-- Start of picture text --> 10 .<br>s <*<br>a. —<br>2S<br>cas22= o o$ o<br>Ze + QBroe<br>OG<br>as ¢<br>co -5 ><br>3BoE yr “<br>fo} -10 Y a<br>[o) 7<br>15)<br>-15 -10 5 0) 5 10<br>Correlation between wealth<br>and average pollution<br><!-- End of picture text --> <!-- Start of picture text --> 1<br>Se / °<br>32 pets 4 et<br>=}<br>co35o> 0 e o . “3.oo 25o%.<br>2 le<br>OFS -1 f<br>ao /<br>c = oe!<br>oO 1<br>DE 2<br>oc£ cs] J<br>Oe ;<br>3 ¢<br>-15 -10 5 ) 5 10<br>Correlation between wealth<br>and average pollution<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [33]}}, {\"page\": 34, \"chunk\": 34, \"input_text\": \"<!-- Start of picture text --> c 10<br>fe)<br>3 3<br>Oo<br>(on<br>Oo @<br>2s °<br>3$<br>©© 00<br>2 @ r)<br>£© e@76 eee<br>‘<br>© O-+----------@-® 6G<br>e@ oe ~—-@<br>c oO Be —<br>® o% ar)<br>DO @ @<br>8<br>=<br>85 é<br>Cc<br>Ae)<br>® ©<br>Oo<br>=<br>Oo<br>O 410<br> aa a |<br>-.05 0 05 4<br>Correlation between wealth and elevation<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [34]}}, {\"page\": 35, \"chunk\": 35, \"input_text\": \"<!-- Start of picture text --> 100<br>ao<br>x<br>=<br>ja<br>a 60<br>=<br>ce<br>—<br>o<br>w 40<br>i]<br>=<br>a<br>20<br>0<br>o c Pa &<br>= é S<br>ae a rs<br>Share of pixels in each URCA category<br><!-- End of picture text --> <!-- Start of picture text --> 10<br>£ 9 090090900 09008,<br>° 0000<br>-5 coolgeens piiasega Gitbalabas23TMe* eel<®S°RR abbadi Plaga;g°§ Giigecibiadiidisiiis:iiebaie °pyiqeeceiseagy% ts<br>Bags ge © 8 3 ° & = £<br>-10 a FI 8<br>4 Fi<br>Ey g<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [35]}}, {\"page\": 36, \"chunk\": 36, \"input_text\": \"<!-- Start of picture text --> y 15 5<br>® R*=0.7<br>a2)<br>£<br>a<br>= O<br>g 10<br>= ©<br>ep)<br>r<br>QO<br>Ps ° °<br>SO<br>w oO<br>Si<br>~ 0 O oO<br>&O<br>©<br>Q<br>Q one) Qo<br>iS -5 e)<br>=<br>S )<br>£<br>we —10<br>a)<br>= a a i a |<br>-10 -5 ) 5 10 15<br>Wealth-pollution correlations using RWI<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [36]}}, {\"page\": 37, \"chunk\": 37, \"input_text\": \"<!-- Start of picture text --> Brazil , Colombia ao Gahana © ne<br>n=210 ° n=21 @n=14<br>. fe) 24 © gga? © 140<br>& 120 fo)<br>72 woe - eee -------@--------O o (O° fo) 2 o Oo 0 ogee , 100 fe) ° Oe<br>Fa oe %2 - O-P- z=S20 °oeo coOo© =a 80 weeroaO°<br>2040 oo 6 18 eo ° ° fe)<br>0 QoGOPSo O @OS 16 o me)|! ° ‘o°° ° °<br>15 10-6 ° ° [e)<br>14 99 JO--------<br>a a rs a | rs a | rsss@-B-@ -O-- OYas@- -O-|<br>0.6 08 1.0 1.2 1.4 1.0 15 2.0 08 1.0 1.2 1.4 1.6<br>Wealth Wealth Wealth<br>India ° n=266 ,, Mexico _ Mongolia e) n=10<br>200 fe) ° 5°?” n=32 120<br>fo} ee) 2540 Pe o 8B 8. [e)® °<br>150 ~ oes’g re) are ° eee<br>° Q § 6 20 ° ..R ---O 100 ee<br>° OF EEO 6 » 20 GE fo) © wee<br>=O 20 COgeemEES?., = 67° ° = lo<br>100 9° Oo oe She Okey o ° o %<br>o (6 & ope RKsk. Weess @ fe) .<br>oy 6 Be BROS 10 00<br>s=272>O 0: = @ ag BS BID 0% [e)<br>506 ® iS) Seam ee0.6 © 5 60 wee<br>° Be KARSO mE CPN OS ° ° weeee cee<br>o) ° 8 “086 ~~ O 0 ie) fe)<br>a a | a a rs | rrssa |<br>-0.5 0.0 0.5 1.0 1.5 0.6 0.8 1.0 1.2 1.4 1.6 18 0.4 0.6 0.8 1.0 1.2 1.4 1.6<br>Wealth Wealth Wealth<br>Peru<br>50 n=12 0<br>fe)<br>“10 fe) 9olg @ © Ground Monitor (urban site) (©) _ Satellite°, Derived. (urban site)*.,<br>. O58. 5° o<br>2° 8 os ©) — Ground Monitor (rural site) © Satellite Derived (rural site)<br>20<br>° ° _...... Estimated pollution-wealth _____.._ Estimated pollution-wealth<br>relationship using relationship using<br>- ground monitor data satellite derived data<br>(e)<br>rrrsa |<br>0.9 1.0 11 1.2 1.3<br>Wealth<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [37]}}, {\"page\": 38, \"chunk\": 38, \"input_text\": \"<!-- Start of picture text --> me=<br>n 5 \\\\<br>(o)) . ay 4 <<br>a Income iF » Income<br>he<br>a | ——L = — Sa<br>8 15 3 110 0 1,000<br>i :<br>o law yy las<br>g se i i cs ‘ite Zod<br>S . Bes : : x . ge = 320 . oe 4 ae<br>ed i a. of :<br>“<br>Income Income<br>O—F—E aaa Ee a |<br>7 10 3 120 0 85<br>x ‘ i» , Ss 8<br>> ‘eo ‘ 4 oy? ‘<br>g ( 1 a seams ( é Income<br>EE ——E SE<br>9 12 3 150 0 125<br>PM,, (ug/m’) Household Income (1,000 $) Elevation (m)<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [38]}}, {\"page\": 39, \"chunk\": 39, \"input_text\": \"<!-- Start of picture text --> c 10<br>oc ]<br>28<br>g2 5 |<br>g2os {#4 +)<br>ofi 0 j boyepar ted ect pee tite pipe Ht a | 1<br>ce I} Hy<br>osoonSe | }Md<br>ag 5<br>2S<br>[S}<br>-10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> cHo2S 10 4<br>28<br>Bo 5 ||<br>5S 1 Hil -<br>woo 0 Vath itiatarcep tp pete stene dpe 1 ytd tht ap ey!<br>EE WWF<br>Do=o & \\\\ i<br>25So 45<br>2E<br>Og<br>g -10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> c 10<br>Be<br>22<br>3S 5<br>a3<br>=<br>ESSeog Ih]<br>osSe 5<br>ag<br>2S<br>[S}<br>-10<br>Cities ordered from least to most correlated<br><!-- End of picture text --> <!-- Start of picture text --> cH32S 10<br>28<br>Be 5<br>Be<br>2 | |<br>EES56=o = |<br>oSo 45<br>a5<br>2E<br>Og<br>g -10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> i—< 10 |<br>ec<br>22<br>58<br>Seog<br>os<br>sg *<br>2S<br>[S}<br>-10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> ceS 10<br>$2<br>28<br>58 ‘|<br>5=o =<br>Os<br>Ss”<br>2E<br>Og<br>g -10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [39]}}, {\"page\": 40, \"chunk\": 40, \"input_text\": \"<!-- Start of picture text -->\\n- 1<br>ooc<br>gS<br>ESSe |<br>og<br>gsa3 ,><br>-10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text -->\\n-2 10<br>osc=fo}<br>82 \\\\h iF<br>o=EE2 I<br>oF<br>Qo85°,<br>og -10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> «<br>o25<br>aE<br>sa |<br>gS 1 |<br>ESocon I 1]<br>gsa3 ,><br>2S<br>° 10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> <£c 10<br>oS25<br>so<br>eB’ }<br>82 I iy]<br>EEo=of i<br>G535° |<br>2s<br>8 10<br>= Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text -->\\n- 1<br>ooc<br>Ss<br>ESoc i<br>og<br>osa3 ><br>-10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text -->\\n-2 10<br>osc=fo}<br>g 2 {|<br>o=EE2 I<br>oF<br>G585°<br>og -10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> « |<br>D j<br>gsgs 5 j!<br>cfs3S |}f<br>Se We<br>8S dd ap be<br>o2 0 Haga linea at aie Ae tet pre ;<br>Sseog yibil<br>osagSe 5<br>id<br>is)<br>-10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> c£c 10<br>92 H<br>Bo& |<br>coe 5 enf<br>J |e<br>3S ss dundaldhelia<br>oo 0 Abeba ae thtsyah. ih |<br>Seof I !<br>ovBeSo 5.<br>2s<br>os<br>g -10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\", \"datasets\": [{\"mention_name\": {\"text\": \"gsa3\", \"confidence\": 0.5312226414680481, \"start\": 67, \"end\": 71, \"clean_text\": \"gsa3\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 67, \"end\": 71, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 67, \"end\": 71, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 67, \"end\": 71}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 67, \"end\": 71, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.9329988956451416, \"start\": 67, \"end\": 71}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5312226414680481, \"start\": 67, \"end\": 71}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5271552205085754, \"start\": 67, \"end\": 71}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5271552205085754, \"start\": 67, \"end\": 71}}, {\"mention_name\": {\"text\": \"Cities ordered from least to most correlated\", \"confidence\": 0.5842968225479126, \"start\": 448, \"end\": 492, \"clean_text\": \"Cities ordered from least to most correlated\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 448, \"end\": 492, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 448, \"end\": 492, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 448, \"end\": 492}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 448, \"end\": 492, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.5857016444206238, \"start\": 78, \"end\": 122}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.41421887278556824, \"start\": 78, \"end\": 122}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.45635682344436646, \"start\": 78, \"end\": 122}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.45635682344436646, \"start\": 78, \"end\": 122}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [40]}}, {\"page\": 41, \"chunk\": 41, \"input_text\": \"<!-- Start of picture text --> 5oc 10<br>Zs55<br>26Ss<br>Be |<br>eS 0 level<br>82 {|<br>Bs<br>.!<br>a3 ><br>fs)2s<br>-10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> 53& 10<br>Fe$= 2<br>eg2 5 |<br>§8 |<br>Be 0 ' +}<br>EE<br>osBEBD<br>85°4<br>2sos<br>g -10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> 5oc 10<br>Zs55<br>gS<br>ES |<br>85<br>Bs !<br>3g .!5<br>fs)2s<br>-10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> 53& 10<br>Fe$= 2 5<br>§2 t<br>4 |<br>Se<br>3s<br>& 5BD 45<br>2sos<br>2 -10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> « |<br>ooc |<br>55 }<br>gS iN} |<br>Eo iar<br>Sc y<br>Bs<br>SgSS 5<br>fs)2s<br>-10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> <2c 10<br>osoF |<br>2<br>82 wt<br>4 yan Fi<br>8e |<br>3e<br>BESE 5<br>2sos<br>g -10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> 5oc 10<br>2255<br>gS nf |<br>BeEo 0 date bi eget ont oath FUT<br>al Be ie<br>BeSS | |<br>2g 5<br>fs)2s<br>-10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> 55& 10 J<br>z$= &2 5 |<br>82 i) ay<br>be 0 kdb abanadma MME FF?<br>EE a<br>osBE<br>a5Se 45<br>2sos<br>g -10<br>Cities ordered from least to most correlated<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [41]}}, {\"page\": 42, \"chunk\": 42, \"input_text\": \"<!-- Start of picture text --> =°= a4\\\\ ~ fDospee ae sg.<br>oO . P~ . Wealth Wealth<br>Ss ae \\\"<br>2 a > las<br>7 é =<br>. a)<br>—\\\"1 a21 -0.1a19 =0 390<br>7 a se<br>Wealth ilar Wealth<br>favu 4 Ne‘i. ; =<br>a a EE<br>1 133 -04 18 190 275<br>a]> Wealth ‘Wealth<br>URCA Administrative<br>= aS Oe Ee EE<br>Rural Urban 28 36 0) 18 0 97<br>Boundaries PM, (g/m?) Wealth (rwi) Elevation (m)<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [42]}}, {\"page\": 43, \"chunk\": 43, \"input_text\": \"<!-- Start of picture text --> ®@ 100<br>D<br>3S<br>oS oo oO, g 20<br>oo Frias * . o<br>03 Sp 50 “aa > An Od.<br>ss tS naga. °°<br>a. ai Be<br>“ Oo oo<br>0<br>2 1 0 1 2 3<br>RWI<br><!-- End of picture text --> RWI, 0.000 RWI,0.00 © RWI, 0.0\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [43]}}, {\"page\": 44, \"chunk\": 44, \"input_text\": \"<!-- Start of picture text --> nod23 120<br>8s 100<br><3<br>23 S5£3os 620 e e g, seaomeeeiiiemn biccadgeaotoge- Beream2B Sg ah.<br>cSs 7:<br>2A 0 1 2<br>RWI<br>RWI, 0.000 © RWI,0.00 ¢ RWI, 0.0<br><!-- End of picture text --> <!-- Start of picture text --> nod=3 40<br>8s a an<br>=5 30 oo Bg Oo<br>oO BBS5329£o 0 S29gVoaaaeaeae: F iocageeogmappire yobo © 8 3<br>aS 10 oe ote<br>2 4 0 1 2<br>RWI<br>RWI, 0.000 © RWI, 0.00 ¢ RWI, 0.0<br><!-- End of picture text --> <!-- Start of picture text --> 3B= 100<br>Beca | ae ouder) _ O .. dee<br>3858 oS geal eel cata io o<br>2s85ge=a 040 0° ae soe, ®<br>A 0 1 2<br>RWI<br>RWI, 0.000 © RWI,0.00 ° RWI, 0.0<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> B40=<br>;<br>33ca %,°o aa oof i ea iog<br>28Bage58=a 10o Sars Pres,Ai oe<br>2 4 0 1 2<br>RWI<br>RWI, 0.000 © RWI, 0.00 ¢ RWI, 0.0<br><!-- End of picture text --> <!-- Start of picture text --> B 30<br>>Begs”a) Be Pos,<br>go58 20 Paeso op Poa igaicat we”Bea if > ° ~°<br>oe > ogres ieeee MECHEee LEMS a COSe O<br>Be 3 ‘<br>Be 10 o<br>e& 5 2<br>A 0 1 2 3<br>RWI<br>RWI, 0.000 © RWI, 0.00 ¢ RWI, 0.0<br><!-- End of picture text --> <!-- Start of picture text --> B 50<br>>BeSS 40 ;<br>i)go<br>2252 30 og ape Oi OcPEP BLE, ap able Be 0 SP<br>28 20 0 - * ‘ ;<br>8a oP ptt<br>g&a 10 3<br>2 4 0 1 2<br>RWI<br>RWI, 0.000 © RWI, 0.00 ¢ RWI, 0.0<br><!-- End of picture text --> <!-- Start of picture text --> 3B 120<br>oO§ 100 3 8 S<br>25 °F oa<br>85 80 * cena oS ,<br>SDofs§© 60 a9 5 Oe a Bice—_ Pieler oe oon.oe)oF ee)Mae a o Sof a a *<br>Bea 40 % o8 8 OE GSSte Betearo Gap ® 2<br>ia x»<br>2 4 0 1 2<br>RWI<br>RWI, 0.000 © RWI,0.00 ¢ RWI, 0.0<br><!-- End of picture text --> <!-- Start of picture text --> 3B %<br>O5 30 °<br>25 op ome ee cS<br>85 235 speactgete %.% °<br>SBDofo 20 Boo yog0° a aaa maeOpe rest eatheoo”bond, Dg Bqo 5 co86 °<br>s§ aCe i .<br>BS 15 0 dig<br>ge2 10 © sue<br>2 A 0 1 2<br>RWI<br>RWI, 0.000 © RWI, 0.00 ¢ RWI, 0.0<br><!-- End of picture text --> <!-- Start of picture text --> B 50<br>52SBS5 40 p<br>52 mer,<br>d 2 39 ~ bh poppe oy Mb<br>3> ia anise aeBio ee |.<br>ss °° fai ne : got<br>28Sa= 20 %o38 > eee2 Seger So oe °<br>&10<br>2 4 0 1 2<br>RWI<br>RWI, 0.000 © RWI, 0.00 ¢ RWI, 0.0<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> B 50<br>53Ses » ototaR Re BPeg ogo<br>52S40 93.\\\"So a tel eget.BF ante?> Pw EMagescage, 20<br>oo ff Rags Brea eo! ea Sp & ® a ,<br>3 2 2 3 Tae ao ey ooo<br>? e Bes O30 8<br>2985<®= 30 5 oe BaasoP.apaniigale8o<br>ge 2<br>2 A 0 1 2 3<br>RWI<br>RWI, 0.000 © RWI, 0.00 ¢ RWI, 0.0<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [44]}}, {\"page\": 45, \"chunk\": 45, \"input_text\": \"# **SI-4 Supplementary References**\\n\\n1. Arrow, K. _et al._ Economic growth, carrying capacity, and the environment. _Ecological economics_ **15,** 91–95 (1995).\\n\\n2. Fuentes, L., Mac-Clure, O., Moya, C. & Olivos, C. Santiago Chile: city of cities? Social inequalities in local labour market zones. _CEPAL review_ (2017).\\n\\n3. Acemoglu, D., Johnson, S. & Robinson, J. A. The colonial origins of comparative development: An empirical investigation. _American economic review_ **91,** 1369–1401 (2001).\\n\\n4. Dell, M. & Olken, B. A. The development effects of the extractive colonial economy: The dutch cultivation system in java. _The Review of Economic Studies_ **87,** 164–203 (2020).\\n\\n5. Hill, F. & Gaddy, C. G. _The Siberian curse: How communist planners left Russia out in the cold_ (Brookings Institution Press, 2003).\\n\\n6. Kumo, K. & Shadrina, E. On the Evolution of Hierarchical Urban Systems in Soviet Russia, 1897–1989. _Sustainability_ **13,** 11389 (2021).\\n\\n7. Gentile, M. The rise and demise of the Soviet-made housing shortage in the baltic countries. _Housing Estates in the Baltic Countries,_ 51–70 (2019).\\n\\n8. Kahn, M. E. Economic Geography and Pollution: A Comment on Joseph S. Shapiro’s “Pollution Trends and US Environmental Policy: Lessons from the Past Half Century”. _Review of Environmental Economics and Policy_ **16,** 357–360 (2022).\\n\\n9. Jbaily, A. _et al._ Air pollution exposure disparities across US population and income groups. _Nature_ **601,** 228–233. issn: 1476-4687. `https://doi.org/10.1038/s41586-021-04190-y` (Jan. 2022).\\n\\n10. Chi, G., Fang, H., Chatterjee, S. & Blumenstock, J. E. Microestimates of wealth for all low-and middle-income countries. _Proceedings of the National Academy of Sciences_ **119** (2022).\\n\\n44\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099449201312441737/pdf/IDU12354a7cd103c314993198571ea404b85aecb.pdf\", \"pages\": [45]}}]}"}}},{"rowIdx":94,"cells":{"doc_id":{"kind":"string","value":"34249700"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34249700\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"admreg\": \"Africa,Africa\", \"admreg_key\": \"119222,119222\", \"authors\": {\"0\": {\"author\": \"Wu,Haoyu\"}, \"1\": {\"author\": \"Atamanov,Aziz\"}, \"2\": {\"author\": \"Bundervoet,Tom\"}, \"3\": {\"author\": \"Paci,Pierella\"}}, \"count\": \"Africa\", \"count_key\": \"517192\", \"docna\": {\"0\": {\"docna\": \"The Growth Elasticity of Poverty : Is Africa\\n Any Different ?\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"EFI-AFR1-POV-Poverty and Equity (EAEPV)\", \"projn\": \"3A-Africa Poverty And Inequality Flagship -- P180133\", \"subsc\": \"Social Protection\", \"prdln\": \"Advisory Services & Analytics\", \"prdln_key\": \"1416466\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34249700\"}, \"repnb\": \"WPS10690\", \"docdt\": \"2024-02-01T00:00:00Z\", \"datestored\": \"2024-02-01T01:12:51Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"On current trends, the future of global\\n poverty reduction will be determined by Sub-Saharan Africa.\\n Yet even during Sub-Saharan Africa’s period of\\n high economic growth — roughly corresponding to\\n the first decade and a half of the 2000s — the\\n extent to which this growth translated into improved living\\n standards for African households was hotly debated. This\\n paper revisits the issue of Sub-Saharan Africa’s\\n relatively low growth elasticity of poverty using a sample\\n of 575 successive and comparable growth spells between 1981\\n and 2021. The findings confirm that, even controlling for\\n initial differences in poverty, income levels, and\\n inequality, Sub-Saharan Africa consistently had a\\n significantly lower growth elasticity of poverty relative to\\n other regions over this period. The lower growth elasticity\\n of poverty, which has remained unchanged over time, is due\\n to a lower passthrough between growth in gross domestic\\n product per capita (or growth in household final consumption\\n expenditure as measured by national accounts) and growth in\\n household consumption expenditures as measured from surveys.\\n Given the low passthrough of economic growth to households,\\n Africa thus needs higher rates of economic growth than its\\n peer countries in other regions to achieve equal rates of\\n poverty reduction. Given the challenge of achieving this in\\n the current global economic environment, success in reducing\\n global poverty will require a focused effort to strengthen\\n the effect of aggregate economic growth on household welfare\\n in Sub-Saharan Africa. The results suggest that this will\\n require (i) improved provision of basic education services\\n and basic infrastructure, (ii) faster structural\\n transformation, and (iii) a decrease in the occurrence and\\n persistence of violent conflicts.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10690; PEOPLE\", \"display_title\": \"The Growth Elasticity of Poverty :\\n Is Africa Any Different ?\", \"disclosure_date\": \"2024-02-01T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099312502012418681/text/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.txt\", \"listing_relative_url\": \"/research/2024/02/34249700/growth-elasticity-poverty-africa-different\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099312502012418681/IDU1e36a6afa14bb414e6c192e61c45a54f4b453\", \"new_url\": \"2024/02/34249700/IDU1e36a6afa14bb414e6c192e61c45a54f4b453\", \"ext_pub_date\": \"2024-02-01T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Poverty m1326687 695,Inclusive Growth m1327170 477,Growth, Poverty and Inequality m1327633 1091\", \"docm_id\": \"IDU-e36a6afa-4bb4-4e6c-92e6-c45a54f4b453\", \"chronical_docm_id\": \"IDU-e36a6afa-4bb4-4e6c-92e6-c45a54f4b453\", \"publishtoextweb_dt\": \"2024-02-01T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10690\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"projectid\": \"P180133\", \"guid\": \"099312502012418681\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"34\", \"author_keys\": {\"0\": {\"author_key\": \"1717170\"}, \"1\": {\"author_key\": \"1182029\"}, \"2\": {\"author_key\": \"1208314\"}, \"3\": {\"author_key\": \"144977\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099312502012418681/IDU1e36a6afa14bb414e6c192e61c45a54f4b453\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10690 The Growth Elasticity of Poverty Is Africa Any Different?\\n\\n_Haoyu Wu Aziz Atamanov Tom Bundervoet Pierella Paci_ Poverty and Equity Global Practice February 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10690\\n\\n# **Abstract**\\n\\nOn current trends, the future of global poverty reduction will be determined by Sub-Saharan Africa. Yet even during Sub-Saharan Africa’s period of high economic growth— roughly corresponding to the first decade and a half of the 2000s—the extent to which this growth translated into improved living standards for African households was hotly debated. This paper revisits the issue of Sub-Saharan Africa’s relatively low growth elasticity of poverty using a sample of 575 successive and comparable growth spells between 1981 and 2021. The findings confirm that, even controlling for initial differences in poverty, income levels, and inequality, Sub-Saharan Africa consistently had a significantly lower growth elasticity of poverty relative to other regions over this period. The lower growth elasticity of poverty, which has remained unchanged over time, is due to a lower passthrough between growth in gross domestic product per capita (or growth in household final consumption expenditure as measured by national accounts) and growth in household consumption expenditures as measured from surveys. Given the low passthrough of economic growth to households, Africa thus needs higher rates of economic growth than its peer countries in other regions to achieve equal rates of poverty reduction. Given the challenge of achieving this in the current global economic environment, success in reducing global poverty will require a focused effort to strengthen the effect of aggregate economic growth on household welfare in Sub-Saharan Africa. The results suggest that this will require (i) improved provision of basic education services and basic infrastructure, (ii) faster structural transformation, and (iii) a decrease in the occurrence and persistence of violent conflicts.\\n\\nThis paper is a product of the Poverty and Equity Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at tbundervoet@worldbank.org.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# The Growth Elasticity of Poverty: Is Africa Any Different?\\n\\nAuthors: Wu, Haoyu r Aziz Atamanov r Tom Bundervoet<sup>1</sup> and Pierella Paci Keywords: poverty, elasticity, Sub-Saharan Africa, economic growth, inequality JEL classification: I32, I38, O1 Acknowledgments. The authors are thankful for two peer-reviewers Dean Jolliffe and Nistha Sinha for useful and constructive feedback. We would also like to thank for guidance and support Andrew Dabalen Ambar Narayan, and Johan A. Mistiaen\\n\\n> 1 The author ordering was constructed through American Economic Association’s randomization tool (confirmation code: ifVmvfTkJeyU). Contact author: Tom Bundervoet tbundervoet@worldbank.org.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# 1. Introduction\\n\\nSince the early 1990s, the world has made tremendous progress in reducing extreme poverty. The global poverty rate, defined by the World Bank’s extreme poverty line of US$2.15 a day (in 2017 Purchasing Power Parity), decreased from 38 percent in 1990 to less than 9 percent by 2019 (World Bank, 2022). This progress was mainly driven by East Asia and the Pacific (EAP), the region that accounted for the bulk of the poor in 1990, where poverty fell from 66 percent in 1990 to 1 percent by 2019. EAP’s success implies that the future of global poverty reduction now lies squarely with Sub-Saharan Africa (SSA) and, to a lesser extent, South Asia. In 2019, 60 percent of the global poor lived in SSA, and this is projected to increase to 87 percent by 2030 (World Bank, 2022a). Progress in Sub-Saharan Africa will determine the success of global poverty reduction efforts in the coming decades.\\n\\nFollowing two largely lost decades, growth in SSA accelerated to an average of 5 percent per year between 2000 and 2014, resulting in a reduction in the poverty rate from 56 percent in 2000 to 38 percent by 2014. However, growth faltered after 2014 and, given rapid population growth, per capita GDP in SSA was slightly lower in 2019 than in 2014.<sup>2</sup> Yet even during SSA’s high growth period, there were concerns around the extent to which this growth benefited the poor. In reference to the period of high growth, Bhorat and Naidoo (2017) write that “ _Africa’s socioeconomic variables have not, however, matched this impressive economic performance_ ”. Similarly, examining the growth, inequality, and poverty nexus in SSA, Thorbecke (2023, p ii81) writes that “ _It is evident that economic growth in SSA in the past two decades has not been sufficiently inclusive_ ”, while Fosu (2023, p ii167) states that “ _Nonetheless, despite the apparent decline in poverty in SSA during the latter period_ [SSA’s growth resurgence period since the late-1990s], _the pace has been slow compared to other regions_ ”. In short, there is a widespread belief that economic growth in SSA has been less poverty-reducing relative to other regions or, in other words, that the growth elasticity of poverty – the percentage change in poverty rates for a percent change in a per capita indicator of economic output – in SSA has been systematically lower.\\n\\nThis paper uses the most up-to-date database of comparable growth spells in developing countries between the early 1980s and 2021 to test this assertion and identify the main factors that make economic growth in SSA less poverty reducing. Compared to the early research on the growth elasticity of poverty (henceforth GEP),<sup>3</sup> we take advantage of important progress made in recent years in the quality and availability of household data. In particular, we benefit from larger sample size and cover a larger share of the developing world’s population thanks to the strong increase in the number of household surveys since the early 2000s, including in previously un- or under-covered countries. We also benefit from the efforts to enhance intertemporal comparability of surveys within countries and include only comparable poverty estimates. While the main focus of the paper is on assessing regional differences in the GEP, the paper seeks to address three main questions: (i) does SSA, the continent with the highest poverty rate and also the slowest progress on poverty reduction, have a GEP that is significantly lower relative to the\\n\\n> 2 In this paper, we consider the region as a whole (Sub-Saharan Africa). The reader is advised that there is large heterogeneity and several non-resource rich countries have kept on growing fast post 2014.\\n\\n> 3 See, for instance, Ravallion and Chen (1997), Ravallion (2001), Bourguignon (2003), Ravallion (2004), Adams (2004), . Kakwani (1990; 1993) introduced the concept of the growth elasticity of poverty reduction but defined this as a partial elasticity, holding the distribution constant, resulting in large elasticities in absolute value.\\n\\n2\", \"datasets\": [{\"mention_name\": {\"text\": \"database of comparable growth spells\", \"confidence\": 0.5583831667900085, \"start\": 2397, \"end\": 2433, \"clean_text\": \"database of comparable growth spells\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2397, \"end\": 2433, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2397, \"end\": 2433, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2397, \"end\": 2433}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2397, \"end\": 2433, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9902682900428772, \"start\": 213, \"end\": 249}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.5583831667900085, \"start\": 213, \"end\": 249}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9223365783691406, \"start\": 213, \"end\": 249}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9223365783691406, \"start\": 213, \"end\": 249}}, {\"mention_name\": {\"text\": \"household data\", \"confidence\": 0.6692434549331665, \"start\": 2793, \"end\": 2807, \"clean_text\": \"household data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2793, \"end\": 2807, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2793, \"end\": 2807, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2793, \"end\": 2807}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2793, \"end\": 2807, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"microdata\", \"confidence\": 0.3077930808067322, \"start\": 609, \"end\": 623}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6692434549331665, \"start\": 609, \"end\": 623}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.8129255771636963, \"start\": 609, \"end\": 623}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8129255771636963, \"start\": 609, \"end\": 623}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"rest of the world (that is, has growth systematically been less poverty-reducing in Sub-Saharan Africa relative to other regions), (ii) has the GEP changed over time (that is, has economic growth on average become more or less poverty-reducing over time), and (iii) are there variables that are amenable through policy that are robustly correlated with the extent to which economic growth translates into improved monetary living standards at the household level?\\n\\nThe analysis in this paper shows that, at the global level, the GEP is remarkably robust: Based on 575 comparable growth spells between 1981 and 2021, we find an average GEP of -2 with respect to survey means, the same as found by Ravallion (2004) in a far smaller sample. The GEP with respect to growth in GDP per capita is higher, at -2.8. In addition, the GEP has not changed through time, suggesting that the recent slowdown in global poverty reduction is mainly due to slowing growth. Our results confirm previous findings that the GEP is substantially lower in SSA, an effect that is mainly driven by a lower passthrough between growth in GDP per capita (or, alternatively, growth in household final consumption expenditure from national accounts) and growth in household consumption as measured from surveys. Finally, we find that the extent to which economic growth translates into improved monetary welfare at the household level is mediated by social stability/peace, the provision of basic public services (literacy and primary education), access to basic infrastructure (electricity, sanitation, and drinking water) and a country’s economic structure (sectoral structure of employment and value added and rents from natural resources).\\n\\nThis paper proceeds as follows: Section 2 introduces the data used in the analysis and presents descriptive statistics on the GEP, disaggregated by region, time period, and other salient dimensions. Section 3 uses a regression framework to test, conditional on differences in initial conditions, whether there are regional disparities in the extent to which economic growth translates into poverty reduction and whether elasticities have changed over time. This section also tests whether the extent to which economic growth translates into improved household welfare is mediated by certain policy variables. The final section concludes.\\n\\n# 2. Data description\\n\\nWe use data from the Poverty and Inequality Platform (World Bank, 2022b) which replaced PovcalNet, the World Bank’s database for monitoring of global poverty (see World Bank, 2022c for a description of data sources and methods used).<sup>4</sup> The database contains income or consumption distributions from nationally representative household surveys typically carried out or supervised by national statistical offices or international agencies, used for national and international poverty monitoring. Following the traditional approach of the research on the GEP, we create a database of “spells” by combining each pair of consecutive surveys within every country. By construction, only countries with at least two comparable household surveys (and hence one spell) are included in the database of spells. This results in a dataset of 1,604 spells based on 1,870 nationally representative household surveys for 139 countries between 1969 and 2021. Countries presented in the database accounted for 95.1 percent of global population in 2019 (and 95.4 percent of population in the “developing world<sup>5</sup> ”).\\n\\nThe distribution of consecutive spells by regions is shown in Table 1. The majority of spells are from two regions with frequent (mostly annual) poverty surveys: Europe and Central Asia and Latin America and\\n\\n> 4 The Poverty and Inequality Platform replaced PovcalNet in March 2022. 5 The world excluding high-income countries.\\n\\n3\", \"datasets\": [{\"mention_name\": {\"text\": \"Poverty and Inequality Platform\", \"confidence\": 0.7363517880439758, \"start\": 2397, \"end\": 2428, \"clean_text\": \"Poverty and Inequality Platform\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2397, \"end\": 2428, \"clean_text\": \"\"}, \"producer\": {\"text\": \"World Bank\", \"confidence\": 0.6671199202537537, \"start\": 2430, \"end\": 2440, \"clean_text\": \"World Bank\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2397, \"end\": 2428}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2397, \"end\": 2428, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9904869198799133, \"start\": 798, \"end\": 829}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7363517880439758, \"start\": 798, \"end\": 829}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9276273250579834, \"start\": 798, \"end\": 829}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9276273250579834, \"start\": 798, \"end\": 829}}, {\"mention_name\": {\"text\": \"PovcalNet\", \"confidence\": 0.6949872374534607, \"start\": 2464, \"end\": 2473, \"clean_text\": \"PovcalNet\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2464, \"end\": 2473, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2464, \"end\": 2473, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2464, \"end\": 2473}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2464, \"end\": 2473, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9904869198799133, \"start\": 865, \"end\": 874}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6949872374534607, \"start\": 865, \"end\": 874}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9276273250579834, \"start\": 865, \"end\": 874}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9276273250579834, \"start\": 865, \"end\": 874}}, {\"mention_name\": {\"text\": \"nationally representative household surveys\", \"confidence\": 0.8032709956169128, \"start\": 2685, \"end\": 2728, \"clean_text\": \"nationally representative household surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2685, \"end\": 2728, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2685, \"end\": 2728, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2685, \"end\": 2728}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2685, \"end\": 2728, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.4265439510345459, \"start\": 1086, \"end\": 1129}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.45770537853240967, \"start\": 1086, \"end\": 1129}, \"usage_context\": {\"text\": \"background\", \"confidence\": 0.44262179732322693, \"start\": 1086, \"end\": 1129}, \"is_used\": {\"text\": \"False\", \"confidence\": 0.44262179732322693, \"start\": 1086, \"end\": 1129}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"the Caribbean. Taken together, these regions account for 78 percent of observations and if not corrected, results will be dominated by these two regions. To correct the bias towards countries with many surveys, we construct weights equal to the inverse number of spells in each country.<sup>6</sup> By doing this, we made sure each country gets an equal weight of one in the analysis regardless of the number of spells available. This changes the regional distribution of spells by giving more weight to countries with fewer observations. The share of the Europe and Central Asia (ECA) and Latin America and the Caribbean (LAC) regions drops from 78 percent to 50 percent once weights are used, while the share of spells in Sub-Saharan Africa (SSA) increases from 5 to 24 percent. In terms of distribution through time, the lion’s share of surveys started after 2010 (almost half of all weighted observations), followed by the period from 2000 to 2010 (37 percent) and the period before 2000 (14 percent). This is in line with the substantial investments in household survey data collection in low-income countries, and SSA in particular, since 2010.\\n\\nFor each spell, we calculate mean household consumption (or income) per capita (from surveys) as well as the US$2.15 a day poverty rate at both the start and the end of the spell. Survey means are expressed in 2017 Purchasing Power Parity (PPP) terms for reasons of comparability. Each of the country-spell years is merged with national accounts data (GDP and household final consumption expenditures, expressed in constant dollars and per capita terms) from the World Development Indicators (WDI).<sup>7</sup> Next, we calculate the “descriptive elasticity”: The annualized change in poverty during a spell divided by the annualized change in survey mean consumption/income per capita during the spell (alternatively: GDP per capita). This results in an estimate of the GEP: the annualized change in the poverty rate for each percent change in per capita mean or GDP.\\n\\nFor the purposes of this paper, we are mainly interested in the elasticity of poverty to economic growth, as measured by growth in GDP per capita. It is this growth rate of GDP that is routinely reported by countries and international organizations, and which is often a target in economic development plans and policies. And it is also by this yardstick that the economic performance of countries is evaluated. We thus assess to what extent headline economic growth translates into improved household welfare, though we also report the elasticity with respect to survey means.<sup>8</sup> As one might argue that the concept of GDP includes many components that do not directly affect the lives of the poor, we also look at the elasticity between household final consumption expenditure as measured by national accounts (henceforth HFCE) and poverty.\\n\\nAnnex Table 1 presents the descriptive elasticities based on the full dataset of 1,604 spells. The summary statistics suggest that some data cleaning will be required to handle extreme values: The descriptive elasticities of poverty with respect to survey means range from -26,348 (indicating that a 1 percent increase in survey means is associated with a 26,348 percent reduction in poverty) to 2,593 (a 1 percent increase in survey means is associated with a 2,593 percent increase in poverty), while the elasticities with respect to GDP per capita range between -8,114 and 9,393. The outliers have a large influence on the average elasticity in our dataset of consecutive spells: The average GEP across all spells amounts to an implausibly large (in absolute value) -17.6 when growth is measured by survey means and -3.8 when\\n\\n> 6 If country X has five “spells” each of them will get weight of 1/5, while country Y with one spell will get the weight of one.\\n\\n> 7 Whenever a survey spans two years, we use weighted GDP across both years as our measure of GDP per capita.\\n\\n> 8 Given that poverty rates and household consumption are both measured from the same surveys, there is a clear and approximate identity-like relationship between both variables (see for instance Bergstrom, 2022). This is not the case for GDP, which is measured through national accounts.\\n\\n4\", \"datasets\": [{\"mention_name\": {\"text\": \"household survey data\", \"confidence\": 0.6649639010429382, \"start\": 1058, \"end\": 1079, \"clean_text\": \"household survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1058, \"end\": 1079, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1058, \"end\": 1079, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1058, \"end\": 1079}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1058, \"end\": 1079, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9875693917274475, \"start\": 292, \"end\": 313}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6649639010429382, \"start\": 292, \"end\": 313}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.43945279717445374, \"start\": 292, \"end\": 313}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.43945279717445374, \"start\": 292, \"end\": 313}}, {\"mention_name\": {\"text\": \"national accounts data\", \"confidence\": 0.32957983016967773, \"start\": 1480, \"end\": 1502, \"clean_text\": \"national accounts data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1480, \"end\": 1502, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1480, \"end\": 1502, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1480, \"end\": 1502}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1480, \"end\": 1502, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9692617058753967, \"start\": 714, \"end\": 736}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.32957983016967773, \"start\": 714, \"end\": 736}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5905897617340088, \"start\": 714, \"end\": 736}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5905897617340088, \"start\": 714, \"end\": 736}}, {\"mention_name\": {\"text\": \"World Development Indicators\", \"confidence\": 0.4670250713825226, \"start\": 1615, \"end\": 1643, \"clean_text\": \"World Development Indicators\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1615, \"end\": 1643, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1615, \"end\": 1643, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1615, \"end\": 1643}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1615, \"end\": 1643, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9692617058753967, \"start\": 849, \"end\": 877}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4670250713825226, \"start\": 849, \"end\": 877}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5905897617340088, \"start\": 849, \"end\": 877}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5905897617340088, \"start\": 849, \"end\": 877}}, {\"mention_name\": {\"text\": \"full dataset of 1,604 spells\", \"confidence\": 0.6567427515983582, \"start\": 2940, \"end\": 2968, \"clean_text\": \"full dataset of 1,604 spells\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2940, \"end\": 2968, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2940, \"end\": 2968, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2940, \"end\": 2968}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2940, \"end\": 2968, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9066469669342041, \"start\": 658, \"end\": 686}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6567427515983582, \"start\": 658, \"end\": 686}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5816032290458679, \"start\": 658, \"end\": 686}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5816032290458679, \"start\": 658, \"end\": 686}}, {\"mention_name\": {\"text\": \"dataset of consecutive spells\", \"confidence\": 0.38935622572898865, \"start\": 3527, \"end\": 3556, \"clean_text\": \"dataset of consecutive spells\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3527, \"end\": 3556, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3527, \"end\": 3556, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3527, \"end\": 3556}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3527, \"end\": 3556, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.38666099309921265, \"start\": 403, \"end\": 432}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.38935622572898865, \"start\": 403, \"end\": 432}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6614301800727844, \"start\": 403, \"end\": 432}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6614301800727844, \"start\": 403, \"end\": 432}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"growth is based on GDP per capita. Using the median elasticity across spells results in more reasonable values that are in line with previous estimates in the literature (a median GEP of -1.8 with respect to the survey mean and -2.1 with respect to per capita GDP). The outliers will however need to be dealt with in light of the forthcoming regression analysis.\\n\\nData intricacies and data comparability issues that give rise to implausible elasticity estimates have in previous research been dealt with by making defensible though largely arbitrary decisions. One approach has been to simply drop spells that are deemed incomparable. This approach has typically been adopted for the transition countries in Europe and Central Asia (ECA), where the exceptional increases in poverty during the post-Soviet Union economic meltdown result in implausibly large elasticities in absolute terms (see Ravallion, 2001). This is also borne out by our dataset, in which the average GEP across spells in ECA amounts to an implausibly large -42 (with respect to survey means) and a positive elasticity with respect to GDP growth (indicating that growth in GDP per capita has been associated with increases in poverty). Another approach has been to drop spells with positive elasticities from the analysis to avoid artificially reducing the average elasticity (Christiaensen and Hill, 2019). More recent research has also attempted to weaken the influence of outliers by trimming the top and bottom 5 percent of elasticity estimates. Having considered different options, we have decided to drop spells for which both the initial-year and end-year poverty rates (as measured by the US$2.15 2017 PPP a day line) are below 2 percent. We believe this is justified by the fact that (i) we are mostly interested in the relationship between growth and poverty reduction in countries that still have some extreme poverty to begin with and (ii) small absolute changes in poverty where baseline poverty rates are low tend to be large in relative terms, mechanically pushing up the absolute values of the estimated elasticities.<sup>9</sup> The curated sample consists of 575 comparable spells between 1981 and 2021, based on 715 nationally representative household surveys for 89 countries representing over 92 percent of the population in the “developing world”. The regional distribution of spells in the curated sample is shown in Panel B of Table 1. Compared to the full sample of 1,604 spells, the largest changes are observed in the ECA and SSA regions. As a result of their low poverty rates, the weighted share of ECA in the sample drops from 35 to 20 percent, while the share of SSA increases from 24 to 36 percent. There are no substantial changes in the weighted distribution of spells by time period (Table 2). About 20 percent of spells started before 2000, 40 percent started in the first decade of the 2000s and another 40 percent started after 2010. The distribution of spells by type of welfare aggregate is more skewed towards consumption in the curated sample, with 70 percent of spells using consumption as a measure of welfare and 30 percent using income. In this sample, the average descriptive GEP with respect to survey means and GDP per capita amount to - 2.1 and -0.9, respectively, while the passthrough between growth in GDP per capita and growth in survey means is approximately unity (indicating that a 1 percent change in per capita GDP has on average been associated with a 1 percent change in household consumption as measured from surveys).\\n\\n> 9 For instance, a reduction in poverty from 1.7 percent to 1.1 percent, though small in absolute terms, comes down to a relative decrease of over 35 percent, which will result in a large elasticity (absolute value). Richer countries therefore tend to have larger elasticity estimates than poor countries. Klasen and Misselhorn (2007) propose the growth semi-elasticity of poverty as a preferred measured. The semi-elasticity measures the _percentage point_ change in poverty for a percent change in mean consumption/income or GDP per capita. While the semi-elasticity is less sensitive to base effects, it tends to be lower for richer countries, as a, say, 5 percentage point reduction in poverty tends to be easier for a country with a baseline poverty rate of 70 percent than for a country with a baseline poverty rate of 8 percent.\\n\\n5\", \"datasets\": [{\"mention_name\": {\"text\": \"715 nationally representative household surveys\", \"confidence\": 0.3989417254924774, \"start\": 2201, \"end\": 2248, \"clean_text\": \"715 nationally representative household surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2201, \"end\": 2248, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2201, \"end\": 2248, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2201, \"end\": 2248}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2201, \"end\": 2248, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8946256041526794, \"start\": 529, \"end\": 576}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3989417254924774, \"start\": 529, \"end\": 576}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6525908708572388, \"start\": 529, \"end\": 576}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6525908708572388, \"start\": 529, \"end\": 576}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"# 3. Empirical analysis\\n\\n## 3.1 Specification and main findings\\n\\nTo obtain a more robust estimate of the GEP and explore differences across regions and time periods, we estimate the following basic specification (Ravallion and Chen, 1997): ∆L i i + ∆∈i Where the rate of poverty reduction is regressed on the rate of growth (either in household consumption LLPP = αα+ ββ∆L LLμμ per capita, GDP per capita, or HFCE per capita). ∆L i denotes the change in log poverty rates between time _t_ and _t-1_ in country _i_ (expressed in annual terms) while ∆L i is the change in real GDP per capita LLPP (or real per capita survey means) in the same country during the same period. We estimate this basic LLμμ specification with and without country fixed effects. While this basic specification forms the core of the analysis, we will sequentially add several other variables to control for initial differences (in poverty rates, in inequality, etc.) and assess whether elasticities differ significantly across regions or time periods.\\n\\nResults of the basic specification are reported in Columns (1) and (2) of Table 3 for elasticity with respect to survey means, Columns (3) and (4) for elasticity with respect to GDP per capita, and Columns (5) and (6) when growth is measured by HFCE per capita. Overall, using household consumption as the measure of growth, the GEP amounts to -1.8 and a slightly higher -2.0 when country effects are controlled for. In other words, on average within countries, a 1 percent increase in household per capita consumption (or income) as measured from surveys has been associated with a 2 percent decrease in the poverty rate. The GEP with respect to GDP per capita is higher, with a 1 percent increase in GDP per capita being associated with a 2.8 percent decrease in the poverty rate, on average within countries (Column (4) of Table 3). If economic growth is measured by HFCE per capita, the GEP is only slightly lower compared to using GDP per capita and ranges from -2.2 to -2.5 depending on whether or not country effects are controlled for (Columns 5 and 6 of Table 3).\\n\\nThe first salient observation from these results is just how robust the estimated elasticities are to different samples. Based on a sample of 120 spells, Ravallion (2004) writes “ _The elasticity of the “$/day” poverty rate to growth in the survey mean is around –2,_ [...]”, which is remarkably close to the GEP with respect to the survey mean we find in our sample of 575 spells.<sup>10</sup> The second salient observation is that in our sample, the GEP with respect to per capita GDP is larger than the GEP with respect to survey means, which is contrary to earlier research on smaller samples. Based on 126 spells, Adams (2004) finds a GEP of -2.3 with respect to GDP per capita, which is statistically indistinguishable from zero (compared to a significant -2.8 with respect to survey means). Ravallion (2001) also finds a lower elasticity when growth is measured by GDP per capita. We however find elasticities based on GDP per capita to be consistently larger than those based on survey means.<sup>11</sup> Next, we add a number of variables that influence the relative pace of poverty reduction (our left-hand side variable). As expected, a higher poverty rate at the beginning of a spell is associated with a significantly slower relative reduction in poverty, as large relative reductions in poverty are more difficult to achieve when the base poverty rate is high (this relates to the earlier point that poor countries usually\\n\\n> 10 Ram (2011) finds a lower GEP but this is based on a higher poverty line. The GEP decreases when higher poverty lines are used (Ravallion and Chen, 1997).\\n\\n> 11 This result is independent of the sample used. Replicating the analysis on the full sample of 1,604 spells also results in a higher GEP with respect to GEP per capita than with respect to survey means.\\n\\n6\", \"datasets\": [{\"mention_name\": {\"text\": \"survey means\", \"confidence\": 0.4036024808883667, \"start\": 2886, \"end\": 2898, \"clean_text\": \"survey means\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2886, \"end\": 2898, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2886, \"end\": 2898, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2886, \"end\": 2898}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2886, \"end\": 2898, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8476978540420532, \"start\": 93, \"end\": 105}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4036024808883667, \"start\": 93, \"end\": 105}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.45862165093421936, \"start\": 93, \"end\": 105}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.45862165093421936, \"start\": 93, \"end\": 105}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"have a lower GEP – see Column (1) of Table 4). However, initial poverty does not matter when growth is measured by GDP or HFCE per capita (see Column (1) of Table 5 and Column (1) of Table 6). Income or consumption inequality at the beginning of a spell is not significantly related to the relative pace of poverty reduction during the spell (see Column (1) of Table 4, Column (1) of Table 5, and Column (1) of Table 6). This echoes Ravallion (2012), who finds that controlling for the initial poverty rate, there is no evidence that higher inequality inhibits the pace of poverty reduction.\\n\\nAmong the key objectives of this paper is to test whether SSA has lower elasticities than other regions, whether elasticities have changed over time, and whether the effect of growth on poverty reduction is mediated by the level of inequality. To do so, we augment the earlier specification as ∆L i i i ) +∈i Where is the relevant characteristic whose mediating influence we wish to test (alternatively: region, ii0 ii0 LLPP = αα+ ββ∆L LLμμ + γγXX + δδ(∆L LLμμ ∗XX time period, and initial inequality) and is the coefficient of the interaction between household ii0 consumption growth and the relevant characteristic. If XX is statistically significant, characteristic has δδ an influence on the extent to which growth translates into poverty reduction. ii0 δδ XX Table 4 summarizes the results for survey means, while Table 5 and Table 6 summarize the results when growth is measured by GDP per capita or HFCE per capita, respectively. Column (2) in Table 4, Table 5 and Table 6 show the results of interacting growth ( ∆L i ) during a spell with a dummy variable for spells in SSA, with growth measured by survey means per capita (Table 4), GDP per capita (Table 5), and HFCE LLμμ per capita (Table 6). Overall, controlling for initial poverty rates, initial welfare means and initial inequality, growth in SSA is associated with a significantly slower relative pace of poverty reduction compared to other regions, especially when growth is measured by GDP per capita (Table 5) and HFCE per capita (Table 6). The coefficient of the interaction term between GDP growth and the SSA dummy is positive and statistically significant at the five percent level (Column (2) in Table 5), while the interaction term between growth in HFCE per capita and SSA is significant at the one percent level (Column (2) in Table 6). The interaction between growth in survey means and the SSA dummy is also positive, but only marginally significant at the 10% level (Column (2) in Table 4). Overall, the estimated GEP for SSA amounts to -0.9 based on survey means, -1.0 based on GDP per capita, and -0.5 based on HFCE per capita, while the corresponding figures for other regions (non SSA) are -2.0, -2.5, and -2.4, respectively (these are marginal effects that control for differences in initial poverty and inequality). Regardless of the indicator of growth, growth between 1981 and 2021 was less poverty reducing in SSA relative to other developing countries. Column (3) in Tables 4-6 examine the effect of inequality in mediating the effect of growth on poverty reduction. Consistent with earlier studies, we find that higher levels of inequality greatly weaken the poverty-reducing effect of growth in household consumption. Column (3) of Table 4 shows that the interaction term between initial inequality and growth in survey means during a spell is large, positive, and highly significant, confirming the well-established finding that high levels of inequality reduce the transmission effect between growth in survey means and poverty reduction (Bourguignon, 2003; Bergstrom, 2022). To illustrate, the average GEP amounts to -2.3 for spells with an initial Gini coefficient lower than 40 but weakens to -1.4 when the initial Gini is 40 or higher. In other words, countries with a high level of inequality need to achieve faster growth to reduce poverty by the same factor, which may 7\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"be particularly challenging in light of inequality’s impact on growth in the first place.<sup>12</sup> The same effect is however not found when growth is measured by GDP per capita (the interaction term between initial inequality and GDP growth is statistically insignificant in Column 3 of Table 5) or HFCE per capita (Column 3 of Table 6). In our sample, initial inequality does not affect the extent to which growth in GDP per capita or HFCE per capita translates into poverty reduction.<sup>13</sup> Column (4) in Tables 4-6 tests whether the GEP has changed through time. For this, we group the growth spells in three categories, depending on the timing of the end year of the spell. The three categories we consider are spells that ended before 2000, spells that ended between 2000 and 2010, and spells that ended in 2011 or later. Results show that the extent to which growth translates into poverty reduction has not changed through time, with none of the interaction effects between time period and growth (in either survey means, GDP per capita, or HFCE per capita) being statistically discernible from zero (spells that ended before 2000 are the reference category). Estimated coefficients are negative, indicating that, if any, elasticities have actually strengthened over time (though not significantly so). Lack of significant changes in GEP over time is robust to different categorization of spells.<sup>14</sup> Therefore, the recent slowdown in poverty reduction is not due to a weakening transmission between growth and poverty reduction, but rather due to the slowdown in growth.\\n\\n3.2 The passthrough between economic growth and household welfare The finding from the previous section that the GEP with respect to GDP per capita is significantly lower in SSA relative to other regions while the GEP with respect to survey means is only marginally lower suggests that the passthrough between aggregate economic growth growth and household consumption growth is lower in Africa (that a given growth in per capita GDP results in less growth in survey mean consumption in SSA than elsewhere). This is confirmed by the results in Table 7, which show the results of regressing changes in survey mean consumption/income per capita during a spell on growth in GDP per capita, controlling for initial differences. On average, the passthrough between growth in GDP per capita and growth in survey means is statistically equivalent to unity, indicating that on average a 1 percent growth in GDP per capita is associated with a 1 percent growth in household welfare as measured from surveys. This result is robust to including country dummies (Column (2) of Table 7). The passthrough between economic growth and household consumption growth is however significantly lower in SSA (Column 4 of Table 7), suggesting that, relative to other regions, countries in SSA need higher economic growth rates to achieve a similar growth in household consumption expenditures as measured by surveys. This result holds when regressing household consumption expenditures from surveys on household final consumption expenditure as measured by national accounts (Table 8).<sup>15</sup> Strikingly, the elasticity between\\n\\n> 12 For literature on the impact of inequality on economic growth, see, for instance, Galor and Zeira, 1993, Bénabou, 1996; Aghion and Bolton, 1997. Dabla-Norris et al. (2015) for instance, find that rising inequality is associated with a medium-term decline in GDP growth.\\n\\n> 13 Inequality as measured from surveys tends however to be substantially underestimated (see, for instance, Chancel et al. (2023) for Africa).\\n\\n> 14 As a robustness check, we assigned spells to time periods depending on which time-period “dominates” the spell. For instance, a spell that started in 2006 and ended in 2011 would be assigned to the 2011-2021 time-period according to our first definition, but would be assigned to the 2000-2010 period according to the second definition. The definition of the spell does not alter any of the results: Estimated coefficients of the interaction terms between growth and time-period are always negative but never statistically significant.\\n\\n> 15 Household Final Consumption Expenditure measures the expenditure incurred by resident households on goods and services and is conceptually closest to household consumption expenditures as measured from surveys, though important differences remain – see Prydz et al., 2022.\\n\\n8\", \"datasets\": [{\"mention_name\": {\"text\": \"surveys on household final consumption expenditure\", \"confidence\": 0.430202841758728, \"start\": 3069, \"end\": 3119, \"clean_text\": \"surveys on household final consumption expenditure\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3069, \"end\": 3119, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3069, \"end\": 3119, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3069, \"end\": 3119}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3069, \"end\": 3119, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9324167966842651, \"start\": 698, \"end\": 748}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.430202841758728, \"start\": 698, \"end\": 748}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6493916511535645, \"start\": 698, \"end\": 748}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6493916511535645, \"start\": 698, \"end\": 748}}, {\"mention_name\": {\"text\": \"surveys\", \"confidence\": 0.49525755643844604, \"start\": 3523, \"end\": 3530, \"clean_text\": \"surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3523, \"end\": 3530, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3523, \"end\": 3530, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3523, \"end\": 3530}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3523, \"end\": 3530, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8364514708518982, \"start\": 310, \"end\": 317}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.49525755643844604, \"start\": 310, \"end\": 317}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5706111788749695, \"start\": 310, \"end\": 317}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5706111788749695, \"start\": 310, \"end\": 317}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"household consumption from surveys and HFCE in SSA is equal to 0.18 and is not significantly different from zero, compared to 0.87 for the rest of the regions. In other words, in our sample of consecutive spells, private consumption as measured by national accounts has no predictive value for household consumption as measured from surveys in SSA.<sup>16</sup> As mentioned earlier in this paper, the lower GEP in SSA should not be a surprise, given that elasticities tend to be lower in countries with high baseline poverty rates (and hence a given percentage point reduction in poverty will be smaller in relative terms). Although the regressions in Tables 4-6 control for both baseline poverty rates and baseline mean welfare, one could still argue that these controls are insufficient and that the result of lower elasticities in SSA is an artifact of the way the GEP is calculated.<sup>17</sup> While this argument can be valid, it cannot be extended to the regressions of survey means on GDP per capita or HFCE (presented in Table 7 and Table 8), as both the dependent (survey means) and key independent variable (GDP per capita or private consumption expenditure per capita) have a similar low base in SSA. As such, we consider the results in Table 7 and Table 8 as our strongest results, suggesting that economic growth (measured by either GDP per capita or HFCE per capita) has a lower passthrough to household consumption as measured from surveys in SSA relative to other regions. And given that poverty in most countries in SSA is measured based on household consumption expenditures from surveys, it suggests that GDP per capita growth must have a weaker effect on poverty reduction in SSA relative to other regions, independent of SSA’s higher baseline poverty rates.\\n\\nTwo other issues need to be addressed before we can have reasonable confidence in the results presented so far. The first relates to surveys not adequately capturing rich households. One might argue that if richer households in Africa have higher consumption growth but are inadequately captured by household surveys, the lower passthrough between GDP growth and survey mean consumption growth in Africa may be due to sampling biases rather than a genuine lower effect of economic growth on household welfare.<sup>18</sup> Replicating the analysis by focusing on the passthrough between GDP growth and survey median consumption (which should be less affected by inadequate coverage of the rich) shows a similar pattern, with the interaction term between growth in GDP per capita and the SSA dummy being negative ( _-0.56_ ) and statistically significant with a p. value of _0.011_ . It is thus unlikely that under coverage of rich households can explain the lower passthrough between economic growth and household welfare in SSA. The second issue relates to the quality of national accounts. If lower income countries, many of which are in Africa, measure growth less well due to a variety of constraints, the associations between GDP growth and poverty reduction may be driven by quality issues. While it is not possible to completely dispel this concern, Angrist et al. (2021) do not find evidence that growth is on average measured less well in developing countries. We also tried to check the role of data quality by including the score on the World\\n\\n> 16 Replicating this analysis on all 1,604 consecutive spells gives a similar result, with the passthrough between household consumption from national accounts and household consumption from surveys being much lower in SSA.\\n\\n> 17 The concept of “semi-elasticities” has been proposed to address this, but arguably suffers from the reverse issue (a country with a baseline poverty rate of 9 percent cannot achieve a 10 percentage-point poverty reduction, while a poorer country with baseline poverty of 60 percent can).\\n\\n> 18 This would also mean that inequality in SSA is higher than what is measured by household surveys, which is what is argued by Chancel et al. (2023).\\n\\n9\", \"datasets\": [{\"mention_name\": {\"text\": \"household consumption expenditures from surveys\", \"confidence\": 0.6923010945320129, \"start\": 1561, \"end\": 1608, \"clean_text\": \"household consumption expenditures from surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1561, \"end\": 1608, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1561, \"end\": 1608, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1561, \"end\": 1608}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1561, \"end\": 1608, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8164702653884888, \"start\": 38, \"end\": 85}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6923010945320129, \"start\": 38, \"end\": 85}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.838588535785675, \"start\": 38, \"end\": 85}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.838588535785675, \"start\": 38, \"end\": 85}}, {\"mention_name\": {\"text\": \"score on the World\", \"confidence\": 0.34044602513313293, \"start\": 3318, \"end\": 3336, \"clean_text\": \"score on the World\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3318, \"end\": 3336, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3318, \"end\": 3336, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3318, \"end\": 3336}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3318, \"end\": 3336, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.4292827844619751, \"start\": 926, \"end\": 944}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.34044602513313293, \"start\": 926, \"end\": 944}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4909150004386902, \"start\": 926, \"end\": 944}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4909150004386902, \"start\": 926, \"end\": 944}}, {\"mention_name\": {\"text\": \"household surveys\", \"confidence\": 0.8431032299995422, \"start\": 3942, \"end\": 3959, \"clean_text\": \"household surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3942, \"end\": 3959, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3942, \"end\": 3959, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3942, \"end\": 3959}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3942, \"end\": 3959, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9735283851623535, \"start\": 754, \"end\": 771}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8431032299995422, \"start\": 754, \"end\": 771}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7017532587051392, \"start\": 754, \"end\": 771}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7017532587051392, \"start\": 754, \"end\": 771}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"Banks Statistical Performance Indicators, a measure of statistical data quality, in all GEP regressions.<sup>19</sup> Adding this variable did not change any previous results in a significant way.\\n\\n3.3 What mediates the effect of economic growth on household welfare? As a final step in the analysis, we assess which factors mediate the effect of economic growth (measured by GDP per capita or HFCE per capita) on household welfare, as measured by household consumption or income per capita from surveys. In the absence of a strong theoretical framework on how the GEP is determined, we draw on the vast (and now largely defunct) cross-country growth literature and the more recent and country-specific microeconomic analysis on the determinants of household welfare to identify a list of variables that potentially mediate the relationship between economic growth and household welfare. The cross-country growth literature, particularly popular in the end of the 1980s and in the 1990s, used cross-country growth regressions to search for empirical linkages between long-run growth rates and a variety economic policy, institutional and political indicators.<sup>20</sup> While this literature has identified over 50 variables that are significantly correlated with growth in at least one specification, ranging from human capital and investment to levels of democracy and ethno-linguistic fractionalization, sensitivity analyses have shown that only a limited number of variables are robustly and systematically correlated with economic growth (Levine and Renelt, 1992): Human and physical capital (measured as a share of investment in GDP) and initial income levels. While the literature on the determinants of household welfare is less straightforward to summarize given substantial country heterogeneity and different data sources, robust correlates of household welfare include the level of education of household members, urban vs rural location (the location effect usually persists even after controlling for other socioeconomic factors), health status of household members and access to basic health services, presence of basic infrastructure (roads, electricity, schools, health centers, etc.) and exposure to conflicts and violence.<sup>21</sup> Indicators proxying good governance have also been shown to support income growth and poverty reduction (Doumbia, 2018). Based on the macro- and micro growth literature, we group the potential mediators of the passthrough between GDP and household consumption growth into four categories<sup>22</sup> :\\n\\n- i. **Human capital:** This category includes indicators of school enrollment and completion at different levels, literacy rates, child and maternal mortality rates, fertility, and life expectancy. We also include an indicator of intergenerational mobility (the probability that a child exceeds the education level of her/his parents for the cohort born in the 1980s). This last variable only varies by country (time series for this variable is available only for a limited set of countries<sup>23</sup> ).\\n\\n- ii. **Economic and demographic structure** : Share of different sectors in GDP, employment share of different sectors, rents from natural resources as a share of GDP, inflation (average in the five years preceding the spell), gross fixed capital formation (with different lags), share of\\n\\n> 19 The Statistical Performance Indicators measure the capacity and maturity of national statistical systems by assessing the use of data, the quality of services, the coverage of topics, the sources of information, and the infrastructure and availability of resources (https://www.worldbank.org/en/programs/statistical-performance-indicators). We used a simple average for each country for the period 2014-2019.\\n\\n> 20 See, for instance, Barro (1991), Levine and Renelt (1992), and Durlauf (2009) for an overview.\\n\\n> 21 See, for instance, Glewwe (1991), Mensah et al. (2014), Christiaensen and Hill (2019), Corral et al. (2020) and the numerous World Bank country-specific poverty assessments.\\n\\n> 22 The list of potential mediators is also influenced by the pragmatic reason of data availability. To be included as a potential mediating variable, we require data for all countries and time periods included in our sample of spells.\\n\\n> 23 Data on intergenerational educational mobility are taken from Narayan et al. (2018).\\n\\n10\", \"datasets\": [{\"mention_name\": {\"text\": \"Banks Statistical Performance Indicators\", \"confidence\": 0.328485906124115, \"start\": 0, \"end\": 40, \"clean_text\": \"Banks Statistical Performance Indicators\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 0, \"end\": 40, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 0, \"end\": 40, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 0, \"end\": 40}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 0, \"end\": 40, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9985702037811279, \"start\": 0, \"end\": 40}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.328485906124115, \"start\": 0, \"end\": 40}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9258618950843811, \"start\": 0, \"end\": 40}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9258618950843811, \"start\": 0, \"end\": 40}}, {\"mention_name\": {\"text\": \"indicator of intergenerational mobility\", \"confidence\": 0.4516088366508484, \"start\": 2780, \"end\": 2819, \"clean_text\": \"indicator of intergenerational mobility\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2780, \"end\": 2819, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2780, \"end\": 2819, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2780, \"end\": 2819}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2780, \"end\": 2819, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9989222288131714, \"start\": 1040, \"end\": 1079}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.4516088366508484, \"start\": 1040, \"end\": 1079}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.649286687374115, \"start\": 1040, \"end\": 1079}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.649286687374115, \"start\": 1040, \"end\": 1079}}, {\"mention_name\": {\"text\": \"Data on intergenerational educational mobility\", \"confidence\": 0.6684600114822388, \"start\": 4301, \"end\": 4347, \"clean_text\": \"Data on intergenerational educational mobility\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4301, \"end\": 4347, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4301, \"end\": 4347, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4301, \"end\": 4347}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 4301, \"end\": 4347, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.4734880030155182, \"start\": 971, \"end\": 1017}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6684600114822388, \"start\": 971, \"end\": 1017}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7631797790527344, \"start\": 971, \"end\": 1017}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7631797790527344, \"start\": 971, \"end\": 1017}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"domestic credit to the private sector, urban population share, population growth, and the share of working age in total population.\\n\\n- iii. **Access to basic infrastructure:** This includes the share of the population with access to electricity, the share of people using at least basic drinking water and sanitation services, and the share of people using the internet. We also include an indicator for rural connectivity (the Rural Access Index<sup>24</sup> ).\\n\\n- iv. **Governance:** This includes type of regime (authoritarian, democratic or hybrid) and a set of World Bank Governance indicators such as voice and accountability, political stability and absence of violence, government effectiveness, regulatory quality, rule of law and control of corruption. This category also includes indicators of violent conflict from the Uppsala Conflict Data Program (UCDP)<sup>25</sup> ( average number of total fatalities per 1,000 people, dummy for years when total fatalities exceeded 0.08 per 1,000 people and dummy for years when total fatalities exceeded 1,000 people<sup>26</sup> ).\\n\\nTo assess whether the effect of economic growth on household welfare as measured by surveys is mediated by any of the specific variables mentioned above, we estimate the following specifications: ∆L i i i ) +∈i Where L i is the change in household consumption or income measured from surveys during time ii0 ii0 LLYY = αα+ ββ∆L LLG PP_C CC + γγXX + δδ(∆L LLG PP_C CC ∗XX period t (a spell between two successive surveys), ∆L i is the change in GDP per capita (from LLYY national accounts) during the time period, and is the variable that potentially mediates the effect of LLG PP_C CC GDP growth on household welfare growth. All regressions also include regional dummies (to control for ii0 XX time-invariant omitted variables at the regional level), the baseline survey mean (survey mean consumption or income at the start of the spell), and a dummy indicating the type of welfare aggregate variable used in the survey (consumption or income). This specification is estimated on two samples: (i) the core analysis sample of 575 successive spells, excluding spells where start- and end-year poverty is below 2 percent (as we are mainly interested in the passthrough between growth and household welfare in poorer countries) and (ii) the full sample of 1,604 successive spells, which is however dominated by upper middle income and high-income countries (these were excluded from the earlier analysis of the GEP due to their near-zero poverty rates, which biases the estimated elasticities). In addition, we also test whether these factors mediate the effect of economic growth on poverty reduction (where the left-hand side variable in equation is replaced by ∆L i ). This analysis is based on the core sample of 575 spells. We also repeated this analysis using HFCE per capita instead of GDP per capita as a measure of economic LLPP growth.\\n\\nGiven that each potential mediating factor is introduced in isolation from others, we run separate regressions for each mediating factor. The specification is always the same, except for the mediating variable. Given that we consider 68 variables as potential mediators between economic growth and household welfare (see Annex Table 2 for the full list of variables we consider), and hence are likely to\\n\\n> 24 The Rural Access Index (RAI) measures the share of the rural population who live within two kilometers (typically equivalent to a walk of 20-25 minutes) of an all-season road as a proportion of the total rural population.\\n\\n> 25 Sundberg and Melander (2013).\\n\\n> 26 See Corral et al. (2020) on measuring conflicts and their impact on socio-economic wellbeing.\\n\\n11\", \"datasets\": [{\"mention_name\": {\"text\": \"Rural Access Index\", \"confidence\": 0.7977691292762756, \"start\": 428, \"end\": 446, \"clean_text\": \"Rural Access Index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.5149404406547546, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"Uppsala Conflict Data Program\", \"confidence\": 0.31043291091918945, \"start\": 831, \"end\": 860, \"clean_text\": \"Uppsala Conflict Data Program\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 428, \"end\": 446}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 428, \"end\": 446, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9999749660491943, \"start\": 428, \"end\": 446}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7977691292762756, \"start\": 428, \"end\": 446}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.47351306676864624, \"start\": 428, \"end\": 446}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.47351306676864624, \"start\": 428, \"end\": 446}}, {\"mention_name\": {\"text\": \"Uppsala Conflict Data Program\", \"confidence\": 0.7332714200019836, \"start\": 831, \"end\": 860, \"clean_text\": \"Uppsala Conflict Data Program\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 831, \"end\": 860, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 831, \"end\": 860, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 831, \"end\": 860}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 831, \"end\": 860, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.980074405670166, \"start\": 831, \"end\": 860}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7332714200019836, \"start\": 831, \"end\": 860}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5921198725700378, \"start\": 831, \"end\": 860}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5921198725700378, \"start\": 831, \"end\": 860}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"have several Type I errors, we control for the expected false discovery rate by calculating sharpened q- values (rather than ordinary p-values) proposed by Benjamini et al. (2006) and Anderson (2008).\\n\\nAt the outset, it is important to keep in mind that none of the associations, however economically and statistically significant they may be, have a causal interpretation. Rather, the aim is to identify policy variables that have been correlated with a higher or lower passthrough between aggregate economic growth on the one side and household welfare and poverty on the other. In addition, the reader should keep in mind that household welfare growth and poverty reduction typically require a combination of policies and investments, and that acting on any one of the mediating variables may have only limited effects if progress is not made on others.\\n\\nTable 9 summarizes the results from the analysis. This table only includes variables that were statistically significant in at least one of the three regressions. Looking at the “current values”, which indicate the interaction of the mediating variable measured at the first year of the spell with growth in GDP per capita during the spell, we find that basic education, access to basic infrastructure, and the share of natural resources in GDP are significantly associated with the extent to which growth translates into household welfare. All associations have the expected sign: A higher net primary school enrollment rate and a higher youth literacy rate are associated with a higher passthrough between economic growth and household welfare, at least in some of the regressions. The finding that net primary school enrollment amplifies the effect of growth on poverty reduction but is not significantly related to the elasticity between economic growth and mean consumption suggests that primary education is especially important for the extent to which the poor benefit from aggregate growth. Among the infrastructure variables, more widespread access to electricity is associated with a stronger elasticity between economic growth and household consumption and poverty, as are the share of the population with access to at least basic drinking water and basic sanitation, which can perhaps be interpreted as indicating a lower burden of disease among a country’s population. The variables on economic structure find, as expected, that being commoditydependent, measured by a higher share of natural resources in GDP, is associated with a lower passthrough between economic growth and household welfare levels.<sup>27</sup> Somewhat surprisingly, oil, gas, or coal rents are not robustly associated with the elasticity between economic growth and household welfare, while forest rents are. While this result is somewhat puzzling, one hypothesis relates to increased household vulnerability to climate shocks in areas with high deforestation. High forest rents are indicative of overharvesting of forests, which amplifies the effects of increasingly frequent climate shocks and hence hamper poverty reduction. Violent conflict is associated with a lower passthrough between GDP growth and household welfare, but only in one of the specifications.\\n\\nFor certain mediating variables, their effect can only be expected to materialize in the longer-term (for instance, one would not expect the level of gross domestic capital formation to influence the passthrough between growth and poverty reduction in the short term). To allow for this, we also interacted growth during a spell with the level of the mediating variable 5, 10, 15 or even 20 years prior to the start of the spell. Using lagged levels of mediating variables points towards the role of economic and employment structure in influencing the elasticity between economic growth and household-level welfare outcomes. A higher share of value added in agriculture five years prior to the start of a spell is associated with a lower passthrough between growth and household welfare, while a higher employment share in services has\\n\\n> 27 Focusing on African countries, Berardi and Marzo (2015) also find that resource- or commodity-driven growth only has a weak effect on poverty.\\n\\n12\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"the opposite effect (Table 9). Using 10-year lags gives qualitatively similar results, with higher shares of economic activity and employment in services and industry strengthening the elasticity while a higher weight of agriculture lowers it. Next to boosting economic growth in itself, the process of structural transformation also seems to boost the extent to which growth benefits households. None of the other variables listed in Annex Table 2 were statistically significant in any of the regressions after adjusting for multiple hypotheses testing.<sup>28</sup> Repeating the analysis using HFCE instead of GDP confirms significance of the mediating factors found earlier (for GDP), and identifies several others (Table 10). Most notably, higher fertility rates, population growth and higher age dependency ratio are associated with a lower passthrough between economic growth and household welfare levels. This is a relevant finding since population growth and fertility rates remain the highest in SSA. Second, more education variables turned out to be significant in HFCE regressions than in regressions using GDP as a measure of economic growth, highlighting the importance of human capital in spreading the benefits of growth. Secondary education in particular amplifies the extent to which growth in HFCE translated into poverty reduction. Higher public spending on secondary education is also associated with a larger passthrough between HFCE and survey mean consumption. Lower rural accessibility and higher incidence of violent conflicts are associated with a significantly lower passthrough between economic growth and wellbeing. While the analysis in this section is correlational and exploratory, it suggests that the effective provision of basic public services, in particular basic education, and infrastructure and the process of structural transformation strengthen the effect of economic growth on household welfare and poverty reduction. The provision of basic education services increases the passthrough between growth and household monetary welfare, as do basic infrastructure assets such as access to electricity and safe drinking water and basic sanitation. High fertility and rapid population growth constrain the growth-poverty relationship through increasing the age dependency ratio. The process of structural transformation whereby labor shifts from agriculture to services and industry, and from rural to urban areas, is associated with a greater welfare-enhancing effect of aggregate economic growth. Finally, the extent to which economic growth benefits households is weakened by exposure to violence and conflict. The findings of our analysis complement those of Thorbecke and Ouyang (2022), who argued for “pro-growth poverty reduction strategies”, as direct interventions in poverty reduction were found to result in faster growth, which in its turn results in faster poverty reduction. Our results suggest another channel, that direct interventions that are good for the poor (basic education and literacy, access to electricity, access to drinking water and sanitation, etc.) also amplify the passthrough between aggregate economic growth and household welfare, sharing the benefits of economic growth more widely.\\n\\n# 4. Conclusions\\n\\nIn this paper we revisit the growth elasticity of poverty based on an updated database of successive and comparable growth spells covering the period 1981-2021. We focus on three main questions: (i) does SSA, the continent with the highest poverty rate and also the slowest progress on poverty reduction, have a GEP that is significantly lower relative to the rest of the world, (ii) has the GEP changed over time (that is, has economic growth on average become more or less poverty-reducing over time), and (iii) are there\\n\\n> 28 Certain human capital variables like mortality rates and intergenerational mobility had significant p-values, but these disappeared once using Anderson’s q values.\\n\\n13\", \"datasets\": [{\"mention_name\": {\"text\": \"updated database of successive and comparable growth spells\", \"confidence\": 0.657606840133667, \"start\": 3346, \"end\": 3405, \"clean_text\": \"updated database of successive and comparable growth spells\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3346, \"end\": 3405, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3346, \"end\": 3405, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3346, \"end\": 3405}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3346, \"end\": 3405, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.9653733372688293, \"start\": 534, \"end\": 593}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.657606840133667, \"start\": 534, \"end\": 593}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6648695468902588, \"start\": 534, \"end\": 593}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6648695468902588, \"start\": 534, \"end\": 593}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"variables that are amenable through policy that are robustly correlated with the extent to which economic growth translates into improved monetary living standards at the household level?\\n\\nWe find that the extent to which GDP per capita growth translates into poverty reduction is significantly lower in SSA than in the rest of the world. While this is expected given the higher baseline poverty rates in SSA (and hence mechanically lower elasticities given that large relative reductions in poverty are more difficult to achieve if the base poverty rate is high), the effect persists when controlling for initial levels of poverty, income, and inequality. The lower growth elasticity of poverty in SSA is mainly driven by a low passthrough between GDP per capita growth (or growth in HFCE) and household consumption growth measured by surveys. The analysis of the passthrough between GDP per capita (and HFCE) and household consumption does not suffer from the same base effects as the poverty analysis, which is why we consider those as our strongest results. Splitting our sample into three time periods, we find that elasticities have not significantly changed over time. The temporal variation in poverty reduction is thus mainly due to variations in growth rather than differences in the extent to which growth translates into poverty reduction.\\n\\nIn trying to shed light on the lower elasticity in SSA, we examine which factors are robustly associated with the extent to which aggregate economic growth translates into household welfare. With little theory to guide us, we drew on the cross-country growth literature and the research on the determinants of household welfare to identify three groups of variables that potentially influence the passthrough between economic growth and household welfare: Variables proxying investments in human capital, variables proxying investments in basic infrastructure, and variables describing a country’s economic and demographic structure. In the human capital category, better basic education (primary education and literacy) is significantly associated with a higher passthrough between growth and household welfare. In the basic infrastructure category, higher levels of electrification and access to clean drinking water and basic sanitation amplify the effect of economic growth on household consumption. Finally, a higher share of employment and value added in agriculture and a higher share of natural resources (mineral rents and forest rents) in GDP consistently inhibit the poverty-reducing effect of GDP growth, while more employment in services and industry increases the elasticity, on average. Exposure to violent conflicts was found to be a separate significant factor limiting passthrough between economic growth and wellbeing.\\n\\nThough only associations, the findings point towards the importance of basic service delivery and the provision of basic infrastructure in spreading more widely the benefits of economic growth. Getting the basics in place appears crucial. In human capital, this includes primary education and, as a result, literacy. In basic infrastructure, electricity, basic sanitation, and drinking water consistently amplify the effect of economic growth on poverty and household monetary living standards. In addition, the process of structural transformation during which labor shifts from agriculture to services and industry and from rural to urban also appears to boost the poverty-reducing effect of growth (next to driving growth itself). In general, variables that amplify the effect of economic growth (such as primary school enrollment, literacy rates, access to electricity and drinking water, etc.) are scarcer in African countries, while variables that inhibit the poverty-reducing effect of growth (employment and value-added in agriculture, natural resource dependency etc.) are more abundant (Annex Table 3). Deeper analysis would however be necessary to more comprehensively assess the reasons behind Africa’s lower passthrough between economic growth and household welfare.\\n\\n14\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"# References\\n\\nAdams Jr., R. (2004). Economic Growth, Inequality and Poverty: Estimating the Growth Elasticity of Poverty. _World Development_ 32(12): 1989-2014.\\n\\nAghion, P. and Bolton, P. (1997). “A Theory of Trickle-Down Growth and Development”. _Review of Economic Studies_ 64: 151-172.\\n\\nAnderson, M. (2008). “Multiple inference and gender differences in the effects of early intervention: A reevaluation of the Abecedarian, Perry Preschool, and Early Training Projects”. _Journal of the American statistical Association_ 103(484): 1481–1495.\\n\\nAnnual Development Effectiveness Review 2013. Towards sustainable growth for Africa Barro, R. (1991). “Economic Growth in a Cross Section of Countries”. _The Quarterly Journal of Economics_ 106(2): 407-443.\\n\\nBatuo, M., Kararach, G., and Malki, I. (2022). The dynamics of income inequality in Africa: An empirical investigation on the role of macroeconomic and institutional forces. World Development 157: Bénabou, R. (1996). “Inequality and Growth”. In Bernanke, B. and Rotemberg, J. (eds). National Bureau of Economic Research Macroeconomics Annual, Cambridge: MIT Press.\\n\\nBenjamini, Y., Krieger, A., and Yekutieli, D. (2006), “Adaptive Linear Step-up Procedures that Control the False Discovery Rate,” _Biometrika_ , 93, 491–507.\\n\\nBerardi, N. and Marzo, F. (2017). “The Elasticity of Poverty with Respect to Sectoral Growth in Africa”. _Review of Income and Wealth_ 63(1): 147-168.\\n\\nBergstrom, K. (2022). “The Role of Income Inequality for Poverty Reduction”. _The World Bank Economic Review_ 36(3): 583-604.\\n\\nBourguignon, F. (2003). “The Growth Elasticity of Poverty Reduction: Explaining Heterogeneity across Countries and Time Period”. In: Eicher, T. and Turnovsky, S., Eds., _Inequality and Growth, Theory and Policy Implications_ , The MIT Press, Cambridge.\\n\\nChambers, D. and Dhongde, S. (2011). “A Non-Parametric Measure of Poverty Elasticity.” _Review of Income and Wealth_ 57(4) Chancel, L., Cogneau, D., Gethin, A., Myczkowski, A., and Robilliard, A.S. (2023). “Income Inequality in Africa: 1990-2019: Measurement, patterns, determinants. _World Development_ 163(3): Christiaensen, L. and Hill, R. (2019). “Poverty in Africa”. In: Beegle, K. and Christiaensen, L. (2019). “Accelerating Poverty Reduction in Africa”. Washington DC: The World Bank.\\n\\nCorral, P., Irwin, A., Krishnan, N., Gerszon Mahler, D. and Vishwanath. T. (2020). Fragility and Conflict: On the Front Lines of the Fight against Poverty. Washington, DC: World Bank.\\n\\nDabla-Norris, E., Kochhar, K., Suphaphiphat, N., Ricka, F. and Tsounta, E. (2015). “Causes and Consequences of Income Inequality: A Global Perspective”. _IMF Staff Discussion Note_ SDN/15/13. Washington DC: The International Monetary Fund.\\n\\n15\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"Devarajan, S. and Fengler, W. (2013). “Africa’s Economic Boom. Why the Pessimists and the Optimists Are Both Right”. _Foreign Affairs_ 92(3): 68-81.\\n\\nDoumbia. D. (2019) The quest for pro-poor and inclusive growth: the role of governance, _Applied Economics_ , 51:16, 1762-1783, DOI: 10.1080/00036846.2018.1529392 Durlauf, S. (2009). “The Rise and Fall of Cross-Country Growth Regressions”. _History of Political Economy_ 41.\\n\\nFigueiredo, E. and Laurini, M. (2016). “Poverty elasticity: A note on a new empirical approach.” _Review of Income and Wealth_ 62(2) Fosu, A, (2015). “Growth, inequality and poverty in Sub-Saharan Africa: Recent progress in a global context”. _Oxford Development Studies_ , 43(1), 44–59.\\n\\nFosu, A. (2023). “Progress on Poverty in Africa: The Importance of Growth and Inequality”. _Journal of African Economies_ 32, ii164–ii182.\\n\\nGalor, O. and Zeira, J. (1993). “Income Distribution and Macroeconomics”. _Review of Economic Studies_ 60(1): 35–52.\\n\\nGlewwe, P. (1991). “Investigating the determinants of household welfare in Cote d’Ivoire”. _Journal of Development Economics_ 35(2): 307-337.\\n\\nKlasen, S. and Misselhorn, M. (2008), “Determinants of the Growth Semi-elasticity of Poverty Reduction,” Ibero America Institute for Economic Research Discussion Paper 176. Göttingen: GeorgAugust-Universität Göttingen.\\n\\nLevin, R. and Renelt, D. (1992). “A Sensitivity Analysis of Cross-Country Growth Regressions”. The American Economic Review 82(4): 942-963.\\n\\nMensah, A., Hutchet-Bourdon, M, and Latruffe, L. (2014). « Infrastructure Access and Household Welfare in rural Ghana”. _African Development Review_ 26(3): 508-519.\\n\\nNarayan, A., Vander Weide, R., Cojocaru, A., Lakner, C., Redaelli, S., Mahler, D., Ramasubbaiah, R., and Thewissen, S. ( 2018). _Fair Progress? Economic Mobility Across Generations Around the World_ , Equity and Development. Washington: The World Bank.\\n\\nPrydz, E., Jolliffe, D. and Serajuddin, U. (2022). “Disparities in Assessments of Living Standards Using National Accounts and Household Surveys”. _The Review of Income and Wealth_ 68(S2): S385S418.\\n\\nRam, R. (2011). “Growth elasticity of poverty: Direct estimates from recent data”. _Applied Economics_ 43(19).\\n\\nRavallion, M. (2001). “Growth, Inequality and Poverty: Looking Beyond Averages. _World Development_ 29(11): 1803-1815.\\n\\nRavallion, M. (2003). “Measuring Aggregate Welfare in Developing Countries: How Well Do National Accounts and Surveys Agree?”. _The Review of Economics and Statistics_ 85(3): 645-652.\\n\\n16\", \"datasets\": [{\"mention_name\": {\"text\": \"National Accounts and Household Surveys\", \"confidence\": 0.4073275327682495, \"start\": 2002, \"end\": 2041, \"clean_text\": \"National Accounts and Household Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2002, \"end\": 2041, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2002, \"end\": 2041, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2002, \"end\": 2041}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2002, \"end\": 2041, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8106454610824585, \"start\": 134, \"end\": 173}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.4073275327682495, \"start\": 134, \"end\": 173}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.889619767665863, \"start\": 134, \"end\": 173}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.889619767665863, \"start\": 134, \"end\": 173}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"Ravallion, M. (2004). “Pro-Poor Growth: A Primer”. _World Bank Policy Research Working Paper_ 3242. Washington DC: The World Bank.\\n\\nRavallion, M. (2012). “Why Don’t we See Poverty Convergence”. _American Economic Review_ 102(1): 504-523.\\n\\nRavallion, M. and S. Chen. (1997). “What Can New Survey Data Tell Us About Recent Changes In Distribution And Poverty?” The World Bank Economic Review, 11, 357–82.\\n\\nRavallion, M. Datt, G. and Van De Walle, D. (1991). “Quantifying absolute poverty in the developing world.” _Review of Income and Wealth_ 37(4): 345-361.\\n\\nSundberg, R., and Melander, E. (2013), “Introducing the UCDP Georeferenced Event Dataset”, _Journal of Peace Research_ , vol.50, no.4, 523-532 Shimeles, A., and Nabassaga, T. (2018). Why is inequality high in Africa? Journal of African Economies, 27 (1), pp. 108-126 Thorbecke, E. (2023). The Interrelationships Among Growth, Inequality and Poverty in Africa _. Journal of African Economies_ 32, ii81–ii86.\\n\\nThorbecke, E. and Ouyang, Y. (2022). “Towards A Virtuous Spiral Between Poverty Reduction And Growth:Comparing Sub Saharan Africa With The Developing World”. World Development 152: 1- 18.\\n\\nWorld Bank. (2018). “ _Poverty and Shared Prosperity Report: Piecing together the poverty puzzle_ ”. Washington DC: The World Bank.\\n\\nWorld Bank. (2022a). “ _Poverty and Shared Prosperity 2022: Correcting Course_ ”. Washington DC: The World Bank.\\n\\nWorld Bank. (2022b). “ _Poverty and Inequality Platform (version 20220909_2017_01_02_PROD)._ World Bank Group. www.pip.worldbank.org. Accessed Septermber 2022.\\n\\nWorld Bank. (2022c). “ _Poverty and Inequality Platform Methodology Handbook_ ”. Washington DC: The World Bank.\\n\\n17\", \"datasets\": [{\"mention_name\": {\"text\": \"Survey Data\", \"confidence\": 0.33437401056289673, \"start\": 288, \"end\": 299, \"clean_text\": \"Survey Data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 288, \"end\": 299, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 288, \"end\": 299, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 288, \"end\": 299}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 288, \"end\": 299, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9479219913482666, \"start\": 288, \"end\": 299}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.33437401056289673, \"start\": 288, \"end\": 299}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.824397087097168, \"start\": 288, \"end\": 299}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.824397087097168, \"start\": 288, \"end\": 299}}, {\"mention_name\": {\"text\": \"UCDP Georeferenced Event Dataset\", \"confidence\": 0.9472869038581848, \"start\": 615, \"end\": 647, \"clean_text\": \"UCDP Georeferenced Event Dataset\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 615, \"end\": 647, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 615, \"end\": 647, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 615, \"end\": 647}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 615, \"end\": 647, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.7471842169761658, \"start\": 615, \"end\": 647}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.9472869038581848, \"start\": 615, \"end\": 647}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5919215679168701, \"start\": 615, \"end\": 647}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5919215679168701, \"start\": 615, \"end\": 647}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"# Tables\\n\\n## <u>Table 1. Distribution of spells by regions</u>\\n\\n||**Pa**|**nel A: All spel**<br>|**ls**<br>|**Panel B:**<br>**hig**|**Spells with po**<br>**her than 2 per**<br>|**verty rates**<br>**cent**<br>|\\n|---|---|---|---|---|---|---|\\n||Number of<br>spells|Share of<br>spells,<br>unweighted|Share of<br>spells,<br>weighted|Number<br>of spells|Share of<br>spells,<br>unweighted|Share of<br>spells,<br>weighted|\\n|East Asia and Pacific|126|8|12|69|12|10|\\n|Europe and Central Asia|884|55|35|109|19|20|\\n|Latin America and Caribbean|362|23|15|278|48|22|\\n|Middle East and North Africa|72|4|9|11|2|6|\\n|North America|46|3|1|0|0|0|\\n|South Asia|31|2|4|28|5|6|\\n|Sub-Saharan Africa|83|5|24|80|14|36|\\n|Total|1,604|100|100|575|100|100|\\n\\n<u>Notes: Panel A shows the regional distribution of all 1,604 consecutive spells, while Panel B focuses on the 575 spells where</u> both start and end-year poverty rates are above two percent.\\n\\n18\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"<u>Table 2. Distribution of spells withpoverty rates higher than 2 percent across region and time period</u>\\n\\n||Num<br>before<br>2000|ber of sp<br>2000-<br>2010|ells<br>2011<br>and<br>above|Sha<br>u<br>before<br>2000|re of spe<br>nweighte<br>2000-<br>2010|lls,<br>d<br>2011<br>and<br>above|Sha<br> <br>before<br>2000|re of spe<br>weighted<br>2000-<br>2010|lls,<br> <br>2011<br>and<br>above|\\n|---|---|---|---|---|---|---|---|---|---|\\n|East Asia and Pacific|26|23|20|38|33|29|48|31|21|\\n|Europe and Central Asia|7|59|43|6|54|39|8|57|35|\\n|Latin America and Caribbean|82|117|79|29|42|28|38|40|22|\\n|Middle East and North Africa|7|3|1|64|27|9|53|27|20|\\n|South Asia|6|11|11|21|39|39|15|53|32|\\n|Sub-Saharan Africa|12|28|40|15|35|50|7|31|62|\\n|Total||575||24|42|34|21|39|39|\\n\\n<u>Notes: Table shows the number of spells by region and time period. Only 575 spells with both start- and end-year poverty rates of</u> at least 2 percent are included.\\n\\n19\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"Table 3: The growth elasticity of poverty\\n\\n||(1)|(2)|(3)|(4)|(5)|(6)|\\n|---|---|---|---|---|---|---|\\n|VARIABLES|Model 1|Model 2|Model 3|Model 4|Model 5|Model 6|\\n|gr_change|-1.826***|-1.974***|-2.307***|-2.762***|-2.204***|-2.491***|\\n||(0.164)|(0.199)|(0.326)|(0.501)|(0.345)|(0.516)|\\n|Constant|-0.0152**<br>(0.00731)|-0.109<br>(0.0841)|0.0132<br>(0.00976)|-0.0883<br>(0.0825)|-0.00144<br>(0.0122)|-0.125*<br>(0.0759)|\\n|country fixed effect|No|Yes|No|Yes|No|Yes|\\n|Observations|575|575|575|575|505|505|\\n|R-squared|0.432|0.590|0.174|0.390|0.208|0.386|\\n|Notes: Table shows results o<br>survey means (Columns (1) a<br>consists of all consecutive sp<br>percent. Robust standard erro<br>significant at 10%.|f a regression of th<br>nd (2)) or GDP pe<br>ells between 1981<br>rs in brackets. ***|e annualized lo<br>r capita (Colum<br>and 2021, drop<br>: Statistically s|g change in po<br>ns (3) and (4)<br>ping spells wit<br>ignificant at 1%|verty headcoun<br>or HFCE per c<br>h baseline and<br>; **: Statistica|t on the annual<br>apita (Columns<br>endline poverty<br>lly significant|ized log change in<br>(5) and (6). Sample<br>rates lower than 2<br>at 5%; *: Statistically|\\n\\n20\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"Table 4: The growth elasticity of poverty based on household consumption as measured from surveys\\n\\n||(1)|(2)|(3)|(4)|\\n|---|---|---|---|---|\\n|VARIABLES|||||\\n|gr_change|-1.847***|-1.966***|-4.030***|-1.664***|\\n||(0.160)|(0.159)|(0.607)|(0.289)|\\n|initial poverty rate|0.171***|0.1000**|0.181***|0.183***|\\n||(0.0373)|(0.0418)|(0.0381)|(0.0374)|\\n|initial inequality|-0.148|-0.181*|-0.285***|-0.170*|\\n||(0.0909)|(0.101)|(0.0937)|(0.0936)|\\n|intial survey mean|0.00791**|0.00634*|0.009***|0.00858***|\\n||(0.00327)|(0.00331)|(0.003)|(0.00324)|\\n|interval|0.00902***|0.00612**|0.009***|0.00842***|\\n||(0.00325)|(0.00305)|(0.003)|(0.00310)|\\n|Baseline: other<br>regions|||||\\n|SSA||0.0182|||\\n|||(0.0234)|||\\n|gr_change x|||||\\n|SSA||1.072*|||\\n|||(0.571)|||\\n|initial headcount gini|||5.111***||\\n||||(1.239)||\\n|Baseline: <2000|||||\\n|2000-2010||||-0.0402**<br>(0.0186)|\\n|2011 and above||||-0.0270*<br>(0.0159)|\\n|gr_change x|||||\\n|2000-2010||||-0.176|\\n|||||(0.369)|\\n|2011 and above||||-0.298|\\n|||||(0.386)|\\n|Baseline:|||||\\n|Consumption|||||\\n|Income|0.0170|0.0265|0.0163|0.00977|\\n||(0.0292)|(0.0301)|(0.0279)|(0.0287)|\\n|Constant|-0.0914***|-0.0515|-0.0383|-0.0585*|\\n||(0.0350)|(0.0387)|(0.0345)|(0.0349)|\\n|country fixed effect|No|No|No|No|\\n|Observations|574|574|574|574|\\n|R-squared|0.466|0.481|0.489|0.473|\\n\\nNotes: Table shows results of a regression of the annualized log change in poverty headcount on the annualized log change in survey means. Sample consists of all consecutive spells between 1981 and 2021, dropping spells with baseline and endline poverty rates lower than 2 percent. Robust standard errors in brackets. ***: Statistically significant at 1%; **: Statistically significant at 5%; *: Statistically significant at 10%.\\n\\n21\", \"datasets\": [{\"mention_name\": {\"text\": \"annualized log change in poverty headcount\", \"confidence\": 0.34376218914985657, \"start\": 1335, \"end\": 1377, \"clean_text\": \"annualized log change in poverty headcount\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1335, \"end\": 1377, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1335, \"end\": 1377, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1335, \"end\": 1377}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1335, \"end\": 1377, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.28157106041908264, \"start\": 164, \"end\": 206}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.34376218914985657, \"start\": 164, \"end\": 206}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.46643543243408203, \"start\": 164, \"end\": 206}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.46643543243408203, \"start\": 164, \"end\": 206}}, {\"mention_name\": {\"text\": \"annualized log change in survey means\", \"confidence\": 0.3036058247089386, \"start\": 1385, \"end\": 1422, \"clean_text\": \"annualized log change in survey means\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1385, \"end\": 1422, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1385, \"end\": 1422, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1385, \"end\": 1422}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1385, \"end\": 1422, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.28157106041908264, \"start\": 214, \"end\": 251}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3036058247089386, \"start\": 214, \"end\": 251}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.46643543243408203, \"start\": 214, \"end\": 251}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.46643543243408203, \"start\": 214, \"end\": 251}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"Table 5: The growth elasticity of poverty based on GDP per capita\\n\\n||(1)|(2)|(3)|(4)|\\n|---|---|---|---|---|\\n|VARIABLES|||||\\n|gr_change|-2.263***|-2.464***|-3.148**|-1.264|\\n||(0.307)|(0.352)|(1.437)|(0.784)|\\n|initial poverty rate|0.0313|-0.00292|0.0313|0.0426|\\n||(0.0365)|(0.0405)|(0.0366)|(0.0345)|\\n|initial inequality|-0.0736|-0.123|-0.130|-0.0988|\\n||(0.0968)|(0.105)|(0.120)|(0.101)|\\n|intial GDP per capita|4.04e-06|3.62e-06|4.63e-06|5.31e-06|\\n||(4.64e-06)|(4.67e-06)|(4.68e-06)|(4.77e-06)|\\n|Interval|0.00523|0.00302|0.00496|0.00412|\\n||(0.00366)|(0.00355)|(0.0037)|(0.00379)|\\n|Baseline: other regions|||||\\n|SSA||-0.00856|||\\n|||(0.0297)|||\\n|gr_change x|||||\\n|SSA||1.487**|||\\n|||(0.605)|||\\n|initial headcount gini|||2.133||\\n||||(3.107)||\\n|Baseline: <2000|||||\\n|2001-2009||||-0.0335|\\n|||||(0.0286)|\\n|2010 and above||||-0.0359|\\n|||||(0.0271)|\\n|gr_change x|||||\\n|2001-2009||||-1.273|\\n|2010 and above||||(0.949)<br>-0.880|\\n|||||(0.882)|\\n|Baseline: Consumption|||||\\n|Income|0.00572|0.00762|0.0048|-0.00448|\\n||(0.0345)|(0.0374)|(0.0346)|(0.0354)|\\n|Constant|0.00317|0.0395|0.0281|0.0405|\\n||(0.0461)|(0.0480)|(0.0591)|(0.0546)|\\n|country fixed effect|No|No|No|No|\\n|Observations|574|574|574|574|\\n|R-squared|0.178|0.186|0.178|0.203|\\n\\nNotes: Table shows results of a regression of the annualized log change in poverty headcount on the annualized log change in GDP per capita. Sample consists of all consecutive spells between 1981 and 2021, dropping spells with baseline and endline poverty rates lower than 2 percent. Robust standard errors in brackets. ***: Statistically significant at 1%; **: Statistically significant at 5%; *: Statistically significant at 10%.\\n\\n22\", \"datasets\": [{\"mention_name\": {\"text\": \"annualized log change in poverty headcount\", \"confidence\": 0.33239269256591797, \"start\": 1273, \"end\": 1315, \"clean_text\": \"annualized log change in poverty headcount\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1273, \"end\": 1315, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1273, \"end\": 1315, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1273, \"end\": 1315}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1273, \"end\": 1315, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.4481217563152313, \"start\": 164, \"end\": 206}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.33239269256591797, \"start\": 164, \"end\": 206}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5552003979682922, \"start\": 164, \"end\": 206}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5552003979682922, \"start\": 164, \"end\": 206}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"Table 6: The growth elasticity of poverty based on HFCE per capita\\n\\n|VARIABLES|(1)|(2)|(3)|(4)|\\n|---|---|---|---|---|\\n|gr_change|-2.139***|-2.379***|-3.724**|-1.771**|\\n||(0.364)|(0.455)|(1.482)|(0.833)|\\n|initial poverty rate|0.0165|-0.0242|0.0155|0.0439|\\n||(0.0482)|(0.0446)|(0.0483)|(0.0408)|\\n|initial inequality|0.0779|-0.0480|-0.0436|0.00392|\\n||(0.109)|(0.125)|(0.134)|(0.110)|\\n|intial HFCE|1.49e-05|1.20e-05|1.53e-05|1.88e-05|\\n||(1.12e-05)|(1.14e-05)|(1.12e-05)|(1.15e-05)|\\n|interval|0.00544|0.00221|0.005|0.00511|\\n||(0.005)|(0.005)|(0.005)|(0.0045)|\\n|Baseline: other regions|||||\\n|SSA||0.00223|||\\n|||(0.0270)|||\\n|gr_change x|||||\\n|SSA||1.846***|||\\n|||(0.690)|||\\n|initial headcount gini|||3.878||\\n||||(3.061)||\\n|Baseline: <2000|||||\\n|2000-2010||||-0.0427|\\n|||||(0.0434)|\\n|2011 and above||||-0.0604|\\n|||||(0.0410)|\\n|gr_change x|||||\\n|2000-2010||||-0.587|\\n|||||(0.918)|\\n|2011 and above||||0.0531|\\n|||||(0.891)|\\n|Baseline: Consumption|||||\\n|Income|-0.0169|-0.00369|-0.0171|-0.0257|\\n||(0.0411)|(0.0467)|(0.0410)|(0.0395)|\\n|Constant|-0.0827*|-0.0105|-0.0294|-0.0194|\\n||(0.0474)|(0.0616)|(0.0683)|(0.0692)|\\n|country fixed effect|No|No|No|No|\\n|Observations|505|505|505|505|\\n|R-squared|0.219|0.233|0.223|0.237|\\n\\nNotes: Table shows results of a regression of the annualized log change in poverty headcount on the annualized log change in HFCE per capita. Sample consists of all consecutive spells between 1981 and 2021, dropping spells with baseline and endline poverty rates lower than 2 percent. Robust standard errors in brackets. ***: Statistically significant at 1%; **: Statistically significant at 5%; *: Statistically significant at 10%.\\n\\n23\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"Table 7: Passthrough between GDP growth and household welfare growth from surveys\\n\\n|VARIABLES|(1)|(2)|(3)|(4)|(5)|(6)|\\n|---|---|---|---|---|---|---|\\n|gdp_change|0.897***|1.103***|0.911***|0.998***|0.459|0.550**|\\n||(0.0999)|(0.149)|(0.107)|(0.119)|(0.549)|(0.258)|\\n|initial inequality|||0.0131|0.0127|-0.0156|0.0247|\\n||||(0.0454)|(0.0487)|(0.0594)|(0.0453)|\\n|intial survey mean|||-0.003**|-0.003**|-0.003**|-0.003**|\\n|interval|||(0.00125)<br>0.001|(0.00139)<br>0.002|(0.00126)<br>0.001|(0.00123)<br>0.002|\\n||||(0.002)|(0.001)|(0.002)|(0.002)|\\n|Baseline: other<br>regions|||||||\\n|SSA||||0.0155|||\\n|||||(0.0113)|||\\n|gdp_change x|||||||\\n|SSA||||-0.499**|||\\n|||||(0.217)|||\\n|Initial inequality|||||1.091||\\n||||||(1.293)||\\n|Baseline: <2000|||||||\\n|2000-2010||||||0.00769|\\n|||||||(0.0125)|\\n|2011 and above||||||0.0183|\\n|||||||(0.0119)|\\n|gdp_change x|||||||\\n|2000-2010||||||0.543*|\\n|||||||(0.303)|\\n|2011 and above||||||0.167|\\n|||||||(0.293)|\\n|Baseline:<br>Consumption|||||||\\n|Income|||0.0319**|0.0351**|0.0318**|0.0355***|\\n||||(0.0132)|(0.0139)|(0.0134)|(0.0130)|\\n|Constant|-0.00495|0.00117|-0.00374|-0.0112|0.00911|-0.0203|\\n||(0.00436)|(0.00614)|(0.0194)|(0.0189)|(0.0265)|(0.0236)|\\n|country fixed effect|No|Yes|No|No|No|No|\\n|Observations|575|575|574|574|574|574|\\n|R-squared|0.203|0.532|0.227|0.234|0.228|0.257|\\n\\nNotes: Table shows results of a regression of the annualized log change in survey mean consumption or income per capita on the annualized log change in GDP per capita. Sample consists of all consecutive spells between 1981 and 2021, dropping spells with baseline and endline poverty rates lower than 2 percent. Robust standard errors in brackets. ***: Statistically significant at 1%; **: Statistically significant at 5%; *: Statistically significant at 10%.\\n\\n24\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"<u>Table 8: Passthrough between Household Final Consumption Expenditure (HFCE) growth from national</u> accounts and household welfare growth from surveys\\n\\n|VARIABLES|(1)|(2)|(3)|(4)|(5)|(6)|\\n|---|---|---|---|---|---|---|\\n|hfce_mean|0.711***|0.847***|0.747***|0.867***|1.031*|0.607**|\\n||(0.117)|(0.176)|(0.123)|(0.147)|(0.549)|(0.269)|\\n|initial headcount gini|||-0.00794|0.0192|0.0136|0.0133|\\n||||(0.0407)|(0.0493)|(0.0581)|(0.0405)|\\n|intial survey mean|||-0.005***|-0.004***|-0.00453***|-0.004***|\\n||||(0.00135)|(0.00133)|(0.0013)|(0.00132)|\\n|interval|||-0.000424|0.000170|-0.0002|-0.000270|\\n||||(0.00175)|(0.00168)|(0.002)|(0.00155)|\\n|Baseline: other regions|||||||\\n|SSA||||0.0162*|||\\n|||||(0.00864)|||\\n|hfce_mean x|||||||\\n|SSA||||-0.685***|||\\n|||||(0.255)|||\\n|initial headcount gini|||||-0.695||\\n||||||(1.314)||\\n|Baseline: <2000|||||||\\n|2001-2009||||||0.00777<br>(0.0148)|\\n|2010 and above||||||0.0189<br>(0.0134)|\\n|hfce_change x|||||||\\n|2001-2009||||||0.281<br>(0.309)|\\n|2010 and above||||||-0.0769|\\n|||||||(0.284)|\\n|Baseline:|||||||\\n|Consumption|||||||\\n|Income|||0.0354**|0.0356**|0.0351**|0.0371***|\\n||||(0.0151)|(0.0150)|(0.0149)|(0.0141)|\\n|Constant|0.00466|0.0203***|0.0312**|0.00924|0.0214|0.00862|\\n||(0.00490)|(0.00594)|(0.0142)|(0.0181)|(0.0232)|(0.0204)|\\n|country fixed effect|No|Yes|No|No|No|No|\\n|Observations|505|505|505|505|505|505|\\n|R-squared|0.198|0.444|0.249|0.267|0.250|0.268|\\n\\nNotes: Table shows results of a regression of the annualized log change in survey mean consumption or income per capita on the annualized log change in private household expenditure per capita as measured from national accounts. Sample consists of all consecutive spells between 1981 and 2021, dropping spells with baseline and endline poverty rates lower than 2 percent. Sample drops from 575 to 505 because for some countries data on private household expenditures from national accounts are missing. Robust standard errors in brackets. ***: Statistically significant at 1%; **: Statistically significant at 5%; *: Statistically significant at 10%.\\n\\n25\", \"datasets\": [{\"mention_name\": {\"text\": \"national accounts\", \"confidence\": 0.8697432279586792, \"start\": 1868, \"end\": 1885, \"clean_text\": \"national accounts\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1868, \"end\": 1885, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1868, \"end\": 1885, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1868, \"end\": 1885}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1868, \"end\": 1885, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.24948999285697937, \"start\": 573, \"end\": 590}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8697432279586792, \"start\": 573, \"end\": 590}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4620213806629181, \"start\": 573, \"end\": 590}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4620213806629181, \"start\": 573, \"end\": 590}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"Table 9: Variables affecting the passthrough between growth in GDP per capita and household welfare\\n\\n||Survey me<br>(N=5|an to GDP<br>75)|Survey me<br>(N=1|an to GDP<br>,604)|Poverty<br>(N=|to GDP<br>575)|\\n|---|---|---|---|---|---|---|\\n||(1|)<br>|(2|)<br>|(|3)<br>|\\n||Coefficient|Sharpened<br>qvalue|Coefficient|Sharpened<br>qvalue|Coefficient|Sharpened<br>qvalue|\\n|**current values**<br>|||||||\\n|Net primary school enrolment|0.014|0.14|0.008|0.28|-0.04**|0.044|\\n|Literacy rate, youth total (% of people<br>ages 15-24)|0.018|0.208|0.013*|0.076|-0.068***|0.001|\\n|Forest rents (% of GDP)|-0.039***|0.003|-0.023|0.229|0.103***|0.003|\\n|Mineral rents (% of GDP)|-0.103**|0.037|-0.011|0.945|0.069|0.81|\\n|Ores and metals exports (% of<br>merchandise exports)|-0.028**|0.033|-0.003|0.945|0.011|0.933|\\n|Access to electricity (% of population)|0.008**|0.037|0.004|0.237|-0.026***|0.003|\\n|People using at least basic drinking water<br>services (% of population)|0.013*|0.085|0.007|0.334|-0.03|0.215|\\n|People using at least basic sanitation<br>services (% of population)|0.007|0.208|0.004|0.441|-0.031**|0.047|\\n|Number of deaths from conflicts per<br>1000 people|-7.475*|0.066|-4.527|0.222|12.486|0.563|\\n|**lagged 5 years**|||||||\\n|Agriculture, value added (% of GDP)|-0.016**|0.041|-0.01|0.173|0.014|0.491|\\n|Employment in services (% of total<br>employment) (modeled ILO estimate)|0.014*|0.099|0.006|0.099|-0.026|0.099|\\n|Ores and metals exports (% of<br>merchandise exports)|-0.028***|0.008|-0.007|0.008|0.008|0.008|\\n|**lagged 10 years**|||||||\\n|Agriculture, value added (% of GDP)|-0.019***|0.004|-0.011*|0.055|0.037|0.151|\\n|Industry, value added (% of GDP)|0.019**|0.031|0.012|0.12|-0.037|0.176|\\n|Employment in services (% of total<br>employment) (modeled ILO estimate)|0.018*|0.061|0.005|0.507|-0.03|0.176|\\n|Employment in industry, (% of total<br>employment) (modeled ILO estimate)|0.05***|0.001|0.016*|0.096|-0.037|0.262|\\n|Ores and metals exports (% of<br>merchandise exports)|-0.022*|0.063|-0.004|0.79|0.005|0.654|\\n|People using at least basic sanitation<br>services (% ofpopulation)<br>Notes:Numbers in the table show the estimated|0.009*<br>coefficients of|0.095<br>the interaction|0.007*<br>term between|0.082<br>rowth in GDP|-0.034<br>per capita and|0.151<br>the|\\n\\n<u>Notes: Numbers in the table show the estimated coefficients of the interaction term between growth in GDP per capita and the</u> corresponding variable listed in the first column, as well as the Anderson’s sharpened q values to control for multiple hypotheses testing. “Current values” indicates that the mediating variable is measured at the beginning year of the spell. “Lagged 5 years” indicates that the mediating variable is measured five years prior to the beginning year of the spell. “Lagged 10 years” indicates that the mediating variable is measured ten years prior to the beginning year of the spell. All regressions include regional dummies to control for time-invariant omitted variables at the regional level. In the regressions of survey means on GDP per capita ((1) and (2)), a positive coefficient indicates that higher levels of the mediating variable amplify the effect of GDP growth on household consumption (and vice versa). In the regression of poverty on GEP per capita (3), a positive coefficient indicates that higher levels of the mediating variable weaken the effect of GDP growth on poverty. ***: Statistically significant at 1%; **: Statistically significant at 5%; *: Statistically significant at 10%.\\n\\n26\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [27]}}, {\"page\": 28, \"chunk\": 28, \"input_text\": \"Table 10: Variables affecting the passthrough between growth in HCFE per capita and household welfare\\n\\n||Survey mean<br>(N=5|to HFCE<br>05)|Survey mea<br>(N=1,|n to HFCE<br>482)|Poverty t<br>(N=5|o HFCE<br>05)|\\n|---|---|---|---|---|---|---|\\n||-1||-2||-3||\\n||Coefficient|<sup>Sharpene</sup><br>dqvalue|<sup>Coefficient</sup>|<sup>Sharpened</sup><br>qvalue|<br>Coefficient|<sup>Sharpene</sup><br>dqvalue|\\n|**current values**|||||||\\n|Literacy rate, adult total (% of people ages 15<br>|0.016*|0.058|0.009*|0.064|-0.034**|0.025|\\n|and above)|||||||\\n|<br>Literacy rate, youth total (% of people ages|0.025**|0.011|0.016**|0.017|-0.051***|0.004|\\n|15-24)|||||||\\n|School enrollment, secondary (% gross)|0.016***|0.001|0.009***|0.002|-0.033*|0.081|\\n|Primary completion rate, total (% of relevant|0.018**|0.021|0.013**|0.029|-0.037|0.126|\\n|age group)|||||||\\n|School enrollment, primary (% gross)|0.019*|0.084|0.011*|0.087|-0.037|0.202|\\n|<br>School enrollment, tertiary (% gross)|0.022***|0.001|0.009***|0.002|-0.036|0.16|\\n|<br>Expenditure on secondary education (% of<br>government expenditure on education)|0.022***|0.005|0.014*|0.054|0.024|0.268|\\n|Urban population (% of total)|0.02***|0.001|0.01***|0.002|-0.044**|0.036|\\n|Rural population (% of total population)|-0.02***|0.001|-0.01***|0.002|0.044**|0.036|\\n|Urban population growth (annual %)|-0.157**|0.029|-0.094***|0.007|0.432*|0.087|\\n|Age dependency ratio (% of working-age<br>population)|-0.022***|0.004|-0.011***|0.006|0.059**|0.031|\\n|<br>Fertility rate, total (births per woman)|-0.274***|0.004|-0.164***|0.002|0.716**|0.031|\\n|Mortality rate, under-5 (per 1,000 live births)|-0.011***|0.001|-0.007***|0.002|0.023**|0.026|\\n|Mortality rate, under-5, female (per 1,000 live<br>births)|<br>-0.012***|0.001|-0.007***|0.002|0.025**|0.025|\\n|Mortality rate, under-5, male (per 1,000 live<br>births)|-0.011***|0.001|-0.006***|0.002|0.022**|0.026|\\n|<br>Maternal mortality ratio (modeled estimate,<br>per 100000 live births)|-0.001***|0.001|-0.001***|0.01|0.003**|0.026|\\n|<br>Population growth (annual %)|-0.261**|0.012|-0.15***|0.002|0.585|0.148|\\n|Rural population growth (annual %)|-0.255***|0.001|-0.161***|0.002|0.351|0.173|\\n|<br>Adolescent fertility rate (births per 1,000|-0.009***|0.007|-0.006***|0.002|0.018|0.111|\\n|women ages 15-19)|||||||\\n|<br>Life expectancy at birth, total (years)|0.051***|0.001|0.028***|0.002|-0.049|0.189|\\n|Wage and salaried workers, total (% of total<br>employment) (modeled ILO estimate)|0.02***|0.001|0.007**|0.015|-0.043*|0.072|\\n|<br>Employment in agriculture (% of total|-0.023***|0.001|-0.012***|0.001|0.038*|0.081|\\n|employment)|||||||\\n|Employment in services (% of total<br>employment) (modeled ILO estimate)|0.026***|0.001|0.015***|0.002|-0.045*|0.081|\\n|Agriculture, value added (% of GDP)|-0.028**|0.014|-0.022***|0.002|0.049|0.148|\\n|<br>Employment in industry, (% of total|0.053***|0.001|0.02**|0.015|-0.09|0.111|\\n|<br>employment) (modeled ILO estimate)|||||||\\n|<br>Rural accessibility index|1.242**|0.029|1.075***|0.002|-1.109|0.385|\\n|<br>Access to electricity (% of population)|0.012***|0.005|0.005**|0.038|-0.034**|0.025|\\n|<br>People using at least basic drinking water|0.028***|0.001|0.018***|0.003|-0.045*|0.078|\\n|services (% of population)|||||||\\n|<br>People using at least basic sanitation services<br>(% of population)|0.013***|0.004|0.007**|0.015|-0.035**|0.025|\\n\\n27\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [28]}}, {\"page\": 29, \"chunk\": 29, \"input_text\": \"||Survey mean<br>(N=5|to HFCE<br>05)|Survey mea<br>(N=1,|n to HFCE<br>482)|Poverty t<br>(N=5|o HFCE<br>05)|\\n|---|---|---|---|---|---|---|\\n||-1||-2||-3||\\n||Coefficient|<sup>Sharpene</sup><br>dqvalue|<sup>Coefficient</sup>|<sup>Sharpened</sup><br>qvalue|Coefficient|<sup>Sharpene</sup><br>dqvalue|\\n|Access to electricity, rural (% of rural<br>population)|0.01***|0.007|0.004*|0.054|-0.03**|0.026|\\n|Conflict in any year of spell based on 0.08|-0.429|0.18|-0.275*|0.054|1.036|0.173|\\n|death rate per 1000|||||||\\n|Average conflict deaths per 1000|-1.599|0.522|-6.051**|0.015|-14.478|0.359|\\n\\n<u>Notes: Numbers in the table show the estimated coefficients of the interaction term between growth in HFCE per capita and the</u> corresponding variable listed in the first column, as well as the Anderson’s sharpened q values to control for multiple hypotheses testing. “Current values” indicates that the mediating variable is measured at the beginning year of the spell. All regressions include regional dummies to control for time-invariant omitted variables at the regional level. In the regressions of survey means on GDP per capita ((1) and (2)), a positive coefficient indicates that higher levels of the mediating variable amplify the effect of GDP growth on household consumption (and vice versa). In the regression of poverty on GEP per capita (3), a positive coefficient indicates that higher levels of the mediating variable weaken the effect of GDP growth on poverty. ***: Statistically significant at 1%; **: Statistically significant at 5%; *: Statistically significant at 10%.\\n\\n28\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099312502012418681/pdf/IDU1e36a6afa14bb414e6c192e61c45a54f4b453.pdf\", \"pages\": [29]}}]}"}}},{"rowIdx":95,"cells":{"doc_id":{"kind":"string","value":"34249732"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34249732\", \"last_modified_date\": \"2024-07-16T00:00:00Z\", \"authors\": {\"0\": {\"author\": \"Xuan Hoang,Trung\"}}, \"count\": \"Iran, Islamic Republic of,Iraq\", \"count_key\": \"82604,82603\", \"docna\": {\"0\": {\"docna\": \"The Long-Term and Gender-Equalizing Impacts\\n of the Islamic Republic of Iran-Iraq War in 1980–88 on\\n School Access and Labor Market Performance\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"Office of the Chief Economist (MNACE)\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34249732\"}, \"repnb\": \"WPS10691\", \"docdt\": \"2024-02-02T00:00:00Z\", \"datestored\": \"2024-02-01T01:49:14Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"This paper uses the context of the\\n Islamic Republic of Iran-Iraq war in 1980–88 to\\n study the long-term impacts of exposure to the war during\\n school years on educational attainment and labor market\\n outcomes in Iraq. The analysis uses an event study and the\\n Iraq Household Socio-Economic Surveys 2006–2007.\\n The findings show that the Islamic Republic of Iran-Iraq\\n conflict had a negative impact on the social welfare of men\\n who were exposed to the war, including on social security,\\n pension plan, health care, paid leave, and job permanence,\\n while little impact on women is found. Additionally, the\\n conflict reduced wages for both men and women. Furthermore,\\n men who were exposed to the conflict were more likely to\\n work in dangerous jobs or without air conditioning, while no\\n evidence on this is found for women. The paper also shows\\n the impact of the intensity of the Islamic Republic of\\n Iran-Iraq war on educational attainment and labor market\\n outcomes. It documents the education channel through which\\n the war affects labor market outcomes, showing that the war\\n decreased the educational levels of men and women born\\n between 1971 and 1981. The findings are robust to a variety\\n of robustness checks and falsification tests.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10691; MNACE labor and gender research programs; PEOPLE\", \"display_title\": \"The Long-Term and Gender-Equalizing\\n Impacts of the Islamic Republic of Iran-Iraq War in 1980–88\\n on School Access and Labor Market Performance\", \"disclosure_date\": \"2024-02-01T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099349102012434554/text/IDU1f87e964d10cb614d3219bdd1759b52b0527b.txt\", \"listing_relative_url\": \"/research/2024/02/34249732/long-term-gender-equalizing-impacts-islamic-republic-iran-iraq-war-1980–88-school-access-labor-market-performance\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099349102012434554/IDU1f87e964d10cb614d3219bdd1759b52b0527b\", \"new_url\": \"2024/02/34249732/IDU1f87e964d10cb614d3219bdd1759b52b0527b\", \"ext_pub_date\": \"2024-02-01T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Education m1327048 470,Conflict and Violence m1327549 2433,Gender m1327621 615\", \"docm_id\": \"IDU-f87e964d-0cb6-4d32-9bdd-759b52b0527b\", \"chronical_docm_id\": \"IDU-f87e964d-0cb6-4d32-9bdd-759b52b0527b\", \"publishtoextweb_dt\": \"2024-02-01T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10691\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"guid\": \"099349102012434554\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"27\", \"author_keys\": {\"0\": {\"author_key\": \"NA\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099349102012434554/IDU1f87e964d10cb614d3219bdd1759b52b0527b\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10691 The Long-Term and Gender-Equalizing Impacts of the Islamic Republic of Iran-Iraq War in 1980–88 on School Access and Labor Market Performance _Trung Xuan Hoang_ Middle East and North Africa Region Office of the Chief Economist February 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10691\\n\\n# **Abstract**\\n\\nThis paper uses the context of the Islamic Republic of IranIraq war in 1980–88 to study the long-term impacts of exposure to the war during school years on educational attainment and labor market outcomes in Iraq. The analysis uses an event study and the Iraq Household Socio-Economic Surveys 2006–2007. The findings show that the Islamic Republic of Iran-Iraq conflict had a negative impact on the social welfare of men who were exposed to the war, including on social security, pension plan, health care, paid leave, and job permanence, while little impact on women is found. Additionally, the conflict reduced wages for both men and women. Furthermore, men who were exposed to the conflict were more likely to work in dangerous jobs or without air conditioning, while no evidence on this is found for women. The paper also shows the impact of the intensity of the Islamic Republic of Iran-Iraq war on educational attainment and labor market outcomes. It documents the education channel through which the war affects labor market outcomes, showing that the war decreased the educational levels of men and women born between 1971 and 1981. The findings are robust to a variety of robustness checks and falsification tests.\\n\\nThis paper is a product of the Office of the Chief Economist, Middle East and North Africa Region. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The author may be contacted at hoangxuantrung3012@gmail.com.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [{\"mention_name\": {\"text\": \"Iraq Household Socio-Economic Surveys\", \"confidence\": 0.8728059530258179, \"start\": 308, \"end\": 345, \"clean_text\": \"Iraq Household Socio-Economic Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 308, \"end\": 345, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 308, \"end\": 345, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 308, \"end\": 345}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 308, \"end\": 345, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9562353491783142, \"start\": 308, \"end\": 345}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8728059530258179, \"start\": 308, \"end\": 345}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8361061811447144, \"start\": 308, \"end\": 345}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8361061811447144, \"start\": 308, \"end\": 345}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# **The Long-Term and Gender-Equalizing Impacts of the Islamic Republic of Iran–Iraq War in 1980–88 on School Access and Labor Market Performance**<sup>**1**</sup>\\n\\nTrung Xuan Hoang Thuongmai University JEL classification: J1; J3; O1; Keywords: Islamic Republic of Iran-Iraq war 1980-1988, education, event study, and Iraq.\\n\\n> 1 This work was supported by the MENA Chief Economist Office under the labor and gender research programs. <mark>This paper is a product of the Office of the Chief Economist, Middle East and North Africa region. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The author(s) may be contacted at hoangxuantrung3012@gmail.com.</mark>\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **1. Introduction**\\n\\nWars have serious consequences for a country, including the destruction of physical infrastructure, loss of lives, displacement of people, and reduced economic growth. Several papers have investigated the long-run effects of war from a macro perspective, particularly bombing. Studies examining the persistent impacts of U.S. bombing in Japan (Davis and Weinstein, 2002) and in Germany (Brakman et al., 2004) during World War II find no evidence of long-lasting effects on local population or economic performance. Similarly, Miguel and Roland (2011) show that the massive U.S. bombings in Vietnam during the Vietnam War from 1967 to 1975, which is cited as one of the most severe bombing incidents in human history, did not have long-term impacts on physical infrastructure, population density, literacy, and poverty level, following 27 years after the war ended. However, a growing body of research on the micro-level effects of armed conflicts reveals that armed wars inflict direct and external costs on survivors that continue longer and can be as harmful as the physical consequences.\\n\\nMost of the existing literature focuses on the long-term impacts of individuals who were exposed to war during early childhood on later life. Particularly, children that have been exposed to the conflict have significantly lower height-for-age in Ethiopia and Eritrea (Akresh et al. 2012) and in Burundi (Bundervoet et al, 2009). Gianmarco (2012) shows that the short-term effects of political violence on educational attainment are larger than those in the long-run in Peru. In addition to the effects of exposure to war on education, the existing literature also shows the negative effects of war on health. For instance, exposure to the war related to the German invasion of France during the first 5 years of life has adverse effects on health in adulthood in France (Allais et al. 2021). Individuals exposed to WWII destruction during pregnancy and the early postnatal period have higher BMIs and are more likely to be obese as adults (Yuksel, 2017). Also, Akresh et al. (2023) show that women who were exposed to the 1967-1970 Nigerian Civil War throughout their childhood had shorter adult stature, are more likely to be overweight or obese, and have their first child at a younger age, and men who have been exposed to war marry later and have fewer children. <mark>Singhal (2018) finds that early-life exposure to the war in Vietnam has adverse and long-term consequences on mental health in adulthood.</mark> The main purpose of this study is to investigate the long-term impact of the Islamic Republic of Iran–Iraq war in 1980-1988 on school-age children between the ages of 6 and 12 in Iraq, during which primary school is of importance. The Islamic Republic of Iran–Iraq war was a military conflict between the Islamic Republic of Iran and Iraq that lasted from September 1980 to August 2\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"1988. It was one of the bloodiest and longest battles that occurred in the Middle East in the 20<sup>th</sup> century. On September 22, 1980, Iraqi armed forces launched a full-scale invasion of neighboring Islamic Republic of Iran. The war was ended by a 1988 cease-fire. The conflict caused devastating effects on these two nations and their people, with total casualties of both sides estimated to range from 1 million to 2 million.\\n\\nTo quantitatively evaluate the long-lasting effects of the war on children aged 6 to 12 during the war during the 1980-1988 period, this study uses the Iraq Household Socio-Economic Surveys (IHSES) 2006-2007 and the event study method. The treatment event cohort is 1968 because the children would reach the age of 12 in 1980 when the Islamic Republic of Iran–Iraq war started. The treatment cohorts include individuals born between 1969 and 1981.\\n\\nWe find that the Islamic Republic of Iran–Iraq war has adverse impacts on labor market performance. The war decreases the social welfare of individuals who were exposed to the war during the school-age years. Specifically, the war decreases the probability of men accessing social security, pension plan, health care, paid leave and job permanence while we find little impact on women. However, we find significant and negative impacts of the war on wages for both men and women. Furthermore men, who were exposed to the war during their school-age years are more likely to work in a hazardous job and without air conditioning; we find no evidence for the women. We document the education channel through which the war affects labor market outcomes. We expect that the war would decrease educational attainment, leading to negative impacts on labor market outcomes. Indeed, the war decreases the educational level of the cohorts of 1971-1981 for both men and women. The Islamic Republic of Iran–Iraq conflict had a greater impact on female education than on male education. The illiterate rate of women increases due to the conflict while we find no evidence of the impact of the war on the illiteracy of men. Our findings are robust to a variety of robustness checks and falsification tests.\\n\\nOur study makes several contributions to the nascent literature on the long-term impacts of war on school-age children. First, our study focuses on Iraq. Existing studies on the effects of war on labor market outcomes and educational attainment in the Middle East and North Africa (MENA) region are rare. This study will fill in this gap by considering the impact of the Islamic Republic of Iran– Iraq war. Iraq presents a remarkable study case. Poverty remains one of Iraq's most pressing social issues, the poverty rate was 22.9 percent in 2007, this rate is still high, and exceeds 40 percent in some governorates, meanwhile 11.3 percent of girls and 5.4 percent of boys aged between 6 and 11 years old did not participate in primary school in 2013 <mark>(UNICEF, 2014).</mark> Second, most of the existing studies focus on the early exposure to the war. In other words, these studies examine the long-term impacts of war on children who were in utero and in their early childhood years during the war. Little is known about the long-term effects of war on children, who were exposed to the war during school-age years of 6 to 12. Therefore, our study sharpens understanding of the longterm impacts of war on school-age children in their later adulthood. Third, we provide heterogeneous analysis. Specifically, we examine the long-term impacts of the war on male and female labor market outcomes, and male and female educational attainment. Iraqi women now face insufficient educational and health care possibility, limited labor market access, as well as high levels of abuse and discrimination. These situations are frequently worsened by cultural and 3\", \"datasets\": [{\"mention_name\": {\"text\": \"Iraq Household Socio-Economic Surveys\", \"confidence\": 0.7481120824813843, \"start\": 589, \"end\": 626, \"clean_text\": \"Iraq Household Socio-Economic Surveys\"}, \"acronym\": {\"text\": \"IHSES\", \"confidence\": 0.5087947249412537, \"start\": 628, \"end\": 633, \"clean_text\": \"IHSES\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 589, \"end\": 626, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 589, \"end\": 626}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 589, \"end\": 626, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.856753408908844, \"start\": 589, \"end\": 626}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7481120824813843, \"start\": 589, \"end\": 626}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9652557373046875, \"start\": 589, \"end\": 626}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9652557373046875, \"start\": 589, \"end\": 626}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"societal norms, incorrect beliefs and a lack of awareness of women’s rights, and institutional and legal restrictions (Vilardo and Bittar, 2018).\\n\\nOnly a handful of studies currently examine the long-term impacts of war on school-age children. We review the most pertinent studies. Olga (2010) examines the impacts of war on school-age children and shows that girls at school age who were exposed to the 1992-1998 armed conflict had lower probability of completing their mandatory schooling than those who were not exposed to the conflict in Tajikistan, but there is no effect of the conflict on education of boys. Similarly, Ichino and Webmer (2004) show that children who were ten years old during World War II received less education and experienced a sizable earnings loss some 40 years later. Blattman and Annan (2010) find adverse impacts on children-abductees who were between 11 and 24 years old during the years they spent with the rebel forces on schooling, skilled employment and earnings in Uganda.\\n\\n# **2.** **<mark>Data</mark>**\\n\\nThis study uses the Iraq Household Socio-Economic Surveys (IHSES) 2006-2007. IHSES 20062007 surveyed 18,144 households, including all three layers of the governorates: one layer for the urban area of the center of the governorate, one layer for the rest of the urban areas, and one layer for the rural areas. This survey served as the foundation for revising the Consumer Price Index (CPI) from an out-of-date 1990 one to a revised 2007 index. IHSES 2006-2007 is nationally representative and covers socio-economic data on migration, labor, education, health, anthropometrics, agriculture, nonfarm, expenditure and income. This survey was carried out by the Central Bureau of Statistics and the Kurdistan Regional Statistics Office, in the field for a year, between October 2006 and November 2007. It covered all 18 governorates in Iraq.\\n\\n# **3. Methodology**\\n\\nThe empirical analysis relies on the variation in the timing of when children finished primary school and the war started. Children who finished primary school right after the onset of the Islamic Republic of Iran–Iraq war have little or no impact of the war compared with those children who had not completed primary school at the timing of the war. Children are more likely to drop out of school if they skip the critical years of primary school education. According to the standard human capital theory (Becker, 1993), enrolling a child in school as early as possible is optimal, as this is when the opportunity cost of the child's time is lowest. Psychological research also indicates that enrolling children in school at an earlier age can be beneficial, as brain development related to cognitive skills becomes increasingly less sensitive with age after the age of 5-6 (Shonkoff and Phillips, 2000). Indeed, Chen (2015) shows that delaying enrollment in school for one year results in a 10 percentage point increase in first-grade retention among boys and a 6 percentage point reduction in the likelihood of middle school enrollment for both boys and girls in rural China. Also, Wils (2004) finds that late school enrollment is associated with higher school dropout rates in Mozambique. Similarly, No et al. (2016) indicate that late school entry of students in grades 1– 4 is one of the causes of school dropouts in Cambodia. According to IHSES 2006-2007, the average percentage of completing primary education or above for cohorts of 1963-1969, who finished primary school during the Islamic Republic of Iran–Iraq war, was 88%. In contrast, 4\", \"datasets\": [{\"mention_name\": {\"text\": \"Iraq Household Socio-Economic Surveys\", \"confidence\": 0.754174530506134, \"start\": 1064, \"end\": 1101, \"clean_text\": \"Iraq Household Socio-Economic Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1064, \"end\": 1101, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1064, \"end\": 1101, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1064, \"end\": 1101}, \"geography\": {\"text\": \"Iraq\", \"confidence\": 0.3764883279800415, \"start\": 1064, \"end\": 1068, \"clean_text\": \"Iraq\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9986953139305115, \"start\": 246, \"end\": 283}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.754174530506134, \"start\": 246, \"end\": 283}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9901911020278931, \"start\": 246, \"end\": 283}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9901911020278931, \"start\": 246, \"end\": 283}}, {\"mention_name\": {\"text\": \"IHSES\", \"confidence\": 0.6546493172645569, \"start\": 1121, \"end\": 1126, \"clean_text\": \"IHSES\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1121, \"end\": 1126, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1121, \"end\": 1126, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1121, \"end\": 1126}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1121, \"end\": 1126, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9980906844139099, \"start\": 303, \"end\": 308}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5189636945724487, \"start\": 303, \"end\": 308}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.809698760509491, \"start\": 303, \"end\": 308}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.809698760509491, \"start\": 303, \"end\": 308}}, {\"mention_name\": {\"text\": \"IHSES 2006-2007\", \"confidence\": 0.6668671369552612, \"start\": 1488, \"end\": 1503, \"clean_text\": \"IHSES 2006-2007\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1488, \"end\": 1503, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1488, \"end\": 1503, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1488, \"end\": 1503}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1488, \"end\": 1503, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9562544226646423, \"start\": 670, \"end\": 685}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5486858487129211, \"start\": 670, \"end\": 685}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6589397192001343, \"start\": 670, \"end\": 685}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6589397192001343, \"start\": 670, \"end\": 685}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"cohorts of 1974-1979 who did not finish primary school at the time of the war had a completion rate of 81%. The survey also provides statistical information about the completion rates of secondary education or higher for two different cohorts, one that finished primary school during the war (1963-1969) and the other that did not complete primary school during the same time (1974-1979). The first cohort had a higher completion rate of 61%, while the second cohort had a lower completion rate of 49%.\\n\\nWe will exploit these significant variations in our identification strategy. Specifically, we conduct an event study along the lines of Jacobson et al. (1993) for the interest outcomes. i 1 + ∑13 i i 2 i kk where i are the outcome variables including educational attainment, social welfare, and working YY ii = αα + γγii kk≥−4,kk≠0 DD ii δδkk + XX iiαα + εε ii conditions of an individual i in governorate p at time t. The dummy variables, i , jointly ii represent the treatment event. In particular, we select the year 1968 as the treatment event because YY kk ii the children would be 12 years old in 1980. Those individuals born before 1968 were not affected DD by Islamic Republic of Iran–Iraq war, and the extent of the impact of the war would increase for those born after 1968 until 1980. We define i =1 if t - 1968 = k, and 0 otherwise, k=-4, -3, -2, - 1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13. And, kk is a set of dummy variables for each governorate. ii Xipt is a dummy variable for gender. All standard errors are clustered at the governorate level. DD ii γγ\\n\\n# **4. Long-run impact of the Islamic Republic of Iran–Iraq war**\\n\\n# **4.1 Long-run impact of the Islamic Republic of Iran–Iraq war on labor market outcomes**\\n\\nTable 1 provides the impact of the Islamic Republic of Iran–Iraq war 1980-1988 on the social welfare of workers. This time, we estimate the long-term effect of the war on workers. We use different measures to reflect the social welfare of workers, they are a dummy variable of having social security, a dummy variable for having pension plan, a dummy variable for health care, a dummy variable for paid leave, a dummy variable for job permanence, and log of wage. The findings show that cohorts of 1970-1980, who were exposed to the war, have lower probability of having social security, the results are not statistically significant for cohorts of 1974 and 1975 (Column 1). Also, most of cohorts of 1970-1980 are less likely to have pension plan, health care, paid leave, and job permanence (Columns 2-5). Importantly, workers of cohorts of 1969-1981 who exposed to the war have lower wage. Those findings suggest that the Islamic Republic of Iran– Iraq war had long-term and negative effects on social welfare of workers who were exposed at ages 6-11 years.\\n\\n**Table 1. Impact of Islamic Republic of Iran–Iraq war on Social welfare of workers**\\n\\n||Dummy for<br>social security|Dummy for<br>pension plan|Dummy for<br>health care|Dummy for<br>paid leave|Dummy for job<br>permanence|Log of<br>wage|\\n|---|---|---|---|---|---|---|\\n||(1)|(2)|(3)|(4)|(5)|(6)|\\n|Year 1964|-0.014|-0.000|-0.030|0.017|0.020|0.117|\\n||(0.041)|(0.039)|(0.030)|(0.042)|(0.022)|(0.133)|\\n|Year 1965|<br>-0.008|<br>-0.002|<br>-0.012|<br>0.011|<br>-0.030|<br>-0.072|\\n\\n5\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"|Year 1966|(0.034)<br>-0.018|(0.034)<br>-0.007|(0.030)<br>-0.008|(0.032)<br>0.019|(0.023)<br>-0.002|(0.122)<br>-0.052|\\n|---|---|---|---|---|---|---|\\n||(0.025)|(0.026)|(0.029)|(0.021)|(0.019)|(0.115)|\\n|Year 1967|-0.044|-0.038|-0.054<sup>*</sup>|-0.004|-0.004|-0.098|\\n||(0.035)|(0.033)|(0.029)|(0.043)|(0.020)|(0.111)|\\n|Year 1969 (Year of<br>change)|-0.057|-0.061|-0.074<sup>***</sup>|-0.061|-0.019|-0.199<sup>*</sup>|\\n||(0.036)|(0.039)|(0.025)|(0.038)|(0.022)|(0.112)|\\n|Year 1970|-0.060<sup>*</sup>|-0.062<sup>*</sup>|-0.053<sup>*</sup>|-0.039|-0.021|-0.221<sup>**</sup>|\\n||(0.029)|(0.032)|(0.026)|(0.029)|(0.016)|(0.083)|\\n|Year 1971|-0.043|-0.043|-0.065<sup>***</sup>|-0.032|0.002|-0.157|\\n||(0.028)|(0.029)|(0.022)|(0.027)|(0.017)|(0.110)|\\n|Year 1972|-0.065<sup>**</sup>|-0.065<sup>*</sup>|-0.042<sup>**</sup>|-0.048|-0.017|-0.259<sup>***</sup>|\\n||(0.028)|(0.031)|(0.019)|(0.033)|(0.014)|(0.079)|\\n|Year 1973|-0.092<sup>***</sup>|-0.095<sup>***</sup>|-0.052<sup>*</sup>|-0.072<sup>**</sup>|-0.028<sup>*</sup>|-0.254<sup>**</sup>|\\n||(0.029)|(0.030)|(0.028)|(0.033)|(0.015)|(0.099)|\\n|Year 1974|-0.032|-0.037|-0.042|-0.016|0.005|-0.100|\\n||(0.036)|(0.039)|(0.025)|(0.038)|(0.019)|(0.119)|\\n|Year 1975|-0.039|-0.056<sup>*</sup>|-0.039|-0.023|-0.011|-0.160|\\n||(0.031)|(0.030)|(0.033)|(0.029)|(0.015)|(0.107)|\\n|Year 1976|-0.063<sup>*</sup>|-0.053|-0.043|-0.027|-0.006|-0.200<sup>**</sup>|\\n||(0.036)|(0.037)|(0.028)|(0.032)|(0.017)|(0.093)|\\n|Year 1977|-0.066<sup>*</sup>|-0.071<sup>*</sup>|-0.049<sup>**</sup>|-0.039|-0.021|-0.355<sup>***</sup>|\\n||(0.032)|(0.039)|(0.019)|(0.027)|(0.019)|(0.086)|\\n|Year 1978|-0.069<sup>*</sup>|-0.062|-0.048<sup>*</sup>|-0.044|0.003|-0.339<sup>***</sup>|\\n||(0.036)|(0.038)|(0.025)|(0.035)|(0.019)|(0.104)|\\n|Year 1979|<br>-0.093<sup>***</sup>|<br>-0.092<sup>***</sup>|<br>-0.051<sup>**</sup>|<br>-0.067<sup>**</sup>|<br>-0.048<sup>**</sup>|<br>-0.439<sup>***</sup>|\\n||(0.028)|(0.030)|(0.024)|(0.026)|(0.019)|(0.093)|\\n|Year 1980|<br>-0.092<sup>***</sup>|<br>-0.097<sup>***</sup>|<br>-0.074<sup>***</sup>|<br>-0.063<sup>*</sup>|<br>-0.030|<br>-0.389<sup>***</sup>|\\n||(0.029)|(0.031)|(0.024)|(0.031)|(0.023)|(0.085)|\\n|Year 1981|<br>-0.128<sup>***</sup>|<br>-0.130<sup>***</sup>|<br>-0.059<sup>**</sup>|<br>-0.081<sup>**</sup>|<br>-0.041<sup>**</sup>|<br>-0.510<sup>***</sup>|\\n||(0.029)|(0.032)|(0.022)|(0.032)|(0.018)|(0.102)|\\n|_N_|10538|10538|10538|10538|10538|10238|\\n|adj._R_<sup>2</sup>|0.139|0.151|0.124|0.139|0.037|0.074|\\n\\nNotes: All regressions add birth governorate fixed-effects, household size and gender. Standard errors are clustered at the birth governorate level. *** p<0.01, ** p<0.05, * p<0.1.\\n\\nTo measure the working conditions of workers, we use a dummy variable for a worker without a hazardous job,<sup>2</sup> a dummy variable for a worker with air conditioning at the workplace and a dummy variable for health impact. A dummy variable for health impact is defined when the health of workers was adversely affected by one of the following factors: soil and gases, chemicals and radiation, heat, humidity, noise, lighting, insects, and fatigue. The results of Table 2 indicate that most of the cohorts of 1971-1980 work in hazardous conditions. The cohorts of 1969-1981 work without air conditioning. And working conditions had adverse impacts on the health of most of the cohorts of 1970-1981.\\n\\n> 2 The question is asked of interviewees in the questionnaires as follows: What is your evaluation of the degree of danger at work? And the responses include: very dangerous, dangerous, relatively, not dangerous. If an interviewee answers “not dangerous”, the dummy variable for a worker without hazardous jobs would be equal to 1, 0 otherwise.\\n\\n6\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"**Table 2. Working conditions of workers**\\n\\n||Hazardous job|Air conditioning|Health impact|\\n|---|---|---|---|\\n||(1)|(2)|(3)|\\n|Year 1964|0.014|0.000|0.129|\\n||(0.043)|(0.035)|(0.187)|\\n|Year 1965|-0.020|-0.028|-0.090|\\n||(0.026)|(0.027)|(0.131)|\\n|Year 1966|0.002|-0.000|0.157|\\n||(0.026)|(0.033)|(0.171)|\\n|Year 1967|0.011|-0.031|0.069|\\n||(0.032)|(0.034)|(0.121)|\\n|Year 1969 (Year of change)|-0.019|-0.066<sup>*</sup>|0.086|\\n||(0.030)|(0.032)|(0.119)|\\n|Year 1970|-0.038|-0.083<sup>**</sup>|0.200<sup>*</sup>|\\n||(0.036)|(0.038)|(0.097)|\\n|Year 1971|-0.058<sup>*</sup>|-0.037<sup>*</sup>|0.069|\\n||(0.030)|(0.019)|(0.115)|\\n|Year 1972|-0.056<sup>*</sup>|-0.084<sup>**</sup>|0.275<sup>**</sup>|\\n||(0.028)|(0.030)|(0.122)|\\n|Year 1973|-0.103<sup>***</sup>|-0.079<sup>***</sup>|0.085|\\n||(0.032)|(0.024)|(0.118)|\\n|Year 1974|-0.098<sup>***</sup>|-0.054<sup>*</sup>|0.324<sup>**</sup>|\\n||(0.029)|(0.030)|(0.127)|\\n|Year 1975|-0.053|-0.077<sup>***</sup>|0.081|\\n||(0.032)|(0.024)|(0.109)|\\n|Year 1976|-0.050|-0.032|0.168|\\n||(0.037)|(0.026)|(0.120)|\\n|Year 1977|-0.055<sup>*</sup>|-0.057<sup>*</sup>|0.255<sup>*</sup>|\\n||(0.031)|(0.031)|(0.130)|\\n|Year 1978|-0.034|-0.077<sup>**</sup>|0.265<sup>*</sup>|\\n||(0.022)|(0.033)|(0.135)|\\n|Year 1979|-0.036|-0.085<sup>***</sup>|0.382<sup>***</sup>|\\n||(0.031)|(0.026)|(0.112)|\\n|Year 1980|<br>-0.066<sup>**</sup>|<br>-0.074<sup>**</sup>|<br>0.188|\\n||(0.024)|(0.026)|(0.120)|\\n|Year 1981|<br>-0.070<sup>**</sup>|<br>-0.119<sup>***</sup>|<br>0.364<sup>***</sup>|\\n||(0.027)|(0.030)|(0.102)|\\n|_N_|10538|10538|10538|\\n|adj._R_<sup>2</sup>|0.125|0.087|0.172|\\n\\nNotes: All regressions add birth governorate fixed-effects, household size and gender. Standard errors are clustered at the birth governorate level. *** p<0.01, ** p<0.05, * p<0.1.\\n\\n# **4.2 Channels**\\n\\nEducation is critical for preparing people to join the labor market and providing them with the skills they need to engage in lifelong learning experiences. Vast research literature provides evidence of the value of investing in education to develop human capital and its contribution to increasing individual earnings (Ashenfelter and Krueger, 1994; Card, 2001; Duflo, 2001; Harmon, 2003; Oreopoulos, 2006; Fiszbein et al. 2007; Clark and Martorell, 2014; Peet et al. 2015; <mark>Silliman and Virtanen, 2022</mark> ). Besides a robust literature examining the impact of schooling on earnings, a large body of literature documents the relationship between educational attainment and labor 7\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"market outcomes. Oreopoulos and Salvanes (2011) show that individuals with better levels of education have higher work satisfaction, autonomy, occupational prestige, and feelings of achievement, they also have lower unemployment rates . Crystal at al. (1992) find that educational attainment has positive impacts on private pensions and other retirement income sources. Oreopoulos (2007) shows negative effects of schooling on the odds of unemployment. <mark>Those findings suggest that education may be a channel through which</mark> the Islamic Republic of Iran–Iraq war affects the labor market outcomes of workers in Iraq.\\n\\nTable 3 presents the estimation results of equation (1) for educational attainment. We use a variety of measures of educational attainment, they are a dummy variable for literacy, a dummy variable for primary education completion or above, a dummy variable for secondary education completion or above and years of education. Without controlling for birth governorate fixed-effects, household size and gender, Columns 1- 4 of Table 1 show that the effect on literacy is negative and statistically significant for cohorts of 1971-1981, except for cohort 1974. The results are also negative for cohorts of 1969 and 1970 for literacy, however the results are not statistically significant. The results on primary education completion or above are highly statistically significant at the 1 percent level for cohorts of 1971-1981. For example, children of cohorts of 1971-1981 were between 3.9 and 7.8 percent lower in the probability of completing primary education or above than those of cohort of 1968. We also find that the children of cohorts of 19711981 have lower probability of completing secondary education or above compared with child of cohort of 1968. Similarly, years of education decreased for the cohorts of 1972-1981. The results are quantitatively similar when we control for birth governorate, household size and gender in Columns 5-8. Educational attainment does not show any clear trend in the cohorts before the cohort of 1968. The educational attainment of cohorts of 1964-1967 is statistically significant but positive, or statistically insignificant.\\n\\n**Table 3. Impact of Islamic Republic of Iran–Iraq war on educational attainment**\\n\\n||Literacy|Primary<br>education<br>completion<br>or above|Secondary<br>education<br>completion<br>or above|Years of<br>education|Literacy|Primary<br>education<br>completion<br>or above|Secondary<br>education<br>completion<br>or above|Years of<br>education|\\n|---|---|---|---|---|---|---|---|---|\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|\\n|Year 1964|0.037<sup>***</sup>|0.024|0.100<sup>***</sup>|0.225|0.037<sup>***</sup>|0.023|0.100<sup>***</sup>|0.107|\\n||(0.010)|(0.017)|(0.019)|(0.155)|(0.010)|(0.017)|(0.020)|(0.152)|\\n|Year 1965|<br>0.015<sup>**</sup>|<br>0.017|<br>0.088<sup>***</sup>|<br>0.195|<br>0.015<sup>**</sup>|<br>0.015|<br>0.086<sup>***</sup>|<br>0.117|\\n||(0.007)|(0.016)|(0.022)|(0.186)|(0.007)|(0.015)|(0.022)|(0.172)|\\n|Year 1966|<br>0.032<sup>***</sup>|<br>0.012|<br>0.072<sup>***</sup>|<br>0.128|<br>0.032<sup>***</sup>|<br>0.015|<br>0.074<sup>***</sup>|<br>0.087|\\n||(0.009)|(0.014)|(0.022)|(0.121)|(0.009)|(0.013)|(0.021)|(0.130)|\\n|Year 1967|0.007|-0.002|0.052<sup>**</sup>|0.193|0.007|-0.001|0.054<sup>**</sup>|0.187|\\n||(0.005)|(0.013)|(0.021)|(0.178)|(0.005)|(0.014)|(0.021)|(0.174)|\\n|Year 1969|-0.004|-0.002|0.013|-0.035|-0.004|0.000|0.013|0.005|\\n|(Year of|||||||||\\n|change)|||||||||\\n||(0.006)|(0.010)|(0.017)|(0.127)|(0.006)|(0.010)|(0.017)|(0.139)|\\n|Year 1970|-0.004|-0.027|-0.023|-0.223|-0.004|-0.026|-0.029|-0.181|\\n||(0.005)|(0.017)|(0.021)|(0.134)|(0.004)|(0.015)|(0.021)|(0.127)|\\n|Year 1971|<br>-0.013<sup>**</sup>|<br>-0.048<sup>***</sup>|<br>-0.044<sup>*</sup>|<br>-0.247|<br>-0.013<sup>**</sup>|<br>-0.041<sup>***</sup>|<br>-0.040<sup>*</sup>|<br>-0.121|\\n||(0.005)|(0.012)|(0.023)|(0.145)|(0.005)|(0.011)|(0.023)|(0.136)|\\n\\n8\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"|Year 1972|-0.011<sup>*</sup>|-0.057<sup>***</sup>|-0.058<sup>***</sup>|-0.477<sup>***</sup>|-0.011<sup>*</sup>|-0.051<sup>***</sup>|-0.057<sup>***</sup>|-0.354<sup>**</sup>|\\n|---|---|---|---|---|---|---|---|---|\\n||(0.006)|(0.018)|(0.018)|(0.154)|(0.006)|(0.017)|(0.017)|(0.131)|\\n|Year 1973|-0.010<sup>*</sup>|-0.064<sup>***</sup>|-0.059<sup>***</sup>|-0.570<sup>***</sup>|-0.011<sup>**</sup>|-0.059<sup>***</sup>|-0.057<sup>***</sup>|-0.439<sup>***</sup>|\\n||(0.005)|(0.017)|(0.014)|(0.105)|(0.005)|(0.017)|(0.015)|(0.108)|\\n|Year 1974|-0.007|-0.059<sup>***</sup>|-0.054<sup>***</sup>|-0.483<sup>***</sup>|-0.007|-0.057<sup>***</sup>|-0.053<sup>***</sup>|-0.398<sup>***</sup>|\\n||(0.006)|(0.011)|(0.015)|(0.122)|(0.006)|(0.011)|(0.015)|(0.117)|\\n|Year 1975|-0.013<sup>**</sup>|-0.039<sup>***</sup>|-0.015|-0.184|-0.013<sup>**</sup>|-0.037<sup>**</sup>|-0.017|-0.134|\\n||(0.005)|(0.013)|(0.021)|(0.179)|(0.005)|(0.013)|(0.021)|(0.179)|\\n|Year 1976|-0.011<sup>**</sup>|-0.053<sup>***</sup>|-0.056<sup>***</sup>|-0.322<sup>**</sup>|-0.011<sup>**</sup>|-0.051<sup>***</sup>|-0.055<sup>**</sup>|-0.247<sup>*</sup>|\\n||(0.005)|(0.014)|(0.019)|(0.141)|(0.005)|(0.014)|(0.020)|(0.139)|\\n|Year 1977|-0.011<sup>**</sup>|-0.050<sup>***</sup>|-0.035<sup>**</sup>|-0.228<sup>*</sup>|-0.011<sup>**</sup>|-0.048<sup>***</sup>|-0.031<sup>**</sup>|-0.135|\\n||(0.005)|(0.012)|(0.015)|(0.117)|(0.005)|(0.012)|(0.014)|(0.117)|\\n|Year 1978|-0.012<sup>**</sup>|-0.070<sup>***</sup>|-0.081<sup>***</sup>|-0.548<sup>***</sup>|-0.012<sup>**</sup>|-0.067<sup>***</sup>|-0.078<sup>***</sup>|-0.434<sup>**</sup>|\\n||(0.006)|(0.017)|(0.016)|(0.177)|(0.006)|(0.016)|(0.017)|(0.174)|\\n|Year 1979|-0.013<sup>**</sup>|-0.059<sup>***</sup>|-0.058<sup>***</sup>|-0.274<sup>*</sup>|-0.014<sup>***</sup>|-0.057<sup>***</sup>|-0.056<sup>***</sup>|-0.140|\\n||(0.005)|(0.012)|(0.016)|(0.131)|(0.005)|(0.011)|(0.017)|(0.123)|\\n|Year 1980|-0.014<sup>***</sup>|-0.074<sup>***</sup>|-0.050<sup>**</sup>|-0.344<sup>*</sup>|-0.015<sup>***</sup>|-0.070<sup>***</sup>|-0.046<sup>**</sup>|-0.189|\\n||(0.004)|(0.011)|(0.018)|(0.195)|(0.004)|(0.010)|(0.019)|(0.197)|\\n|Year 1981|-0.018<sup>***</sup>|-0.078<sup>***</sup>|-0.088<sup>***</sup>|-0.564<sup>***</sup>|-0.019<sup>***</sup>|-0.068<sup>***</sup>|-0.080<sup>***</sup>|-0.341<sup>**</sup>|\\n||(0.004)|(0.012)|(0.020)|(0.124)|(0.004)|(0.011)|(0.021)|(0.144)|\\n|_N_|29050|29050|29050|25124|29050|29050|29050|25124|\\n|adj._R_<sup>2</sup>|0.010|0.006|0.010|0.003|0.020|0.038|0.023|0.078|\\n\\nNotes: Columns 5-8 add birth governorate fixed-effects, household size and gender. Standard errors are clustered at the birth governorate level. *** p<0.01, ** p<0.05, * p<0.1.\\n\\n# **4.3. Robustness checks**\\n\\nOur estimation results may be affected by potential channels; therefore, we offer a battery of robustness checks. We expect that individuals who were born in other countries would not be affected by the Islamic Republic of Iran–Iraq war. Those individuals offer us an important falsification test for the impact of the Islamic Republic of Iran–Iraq war, where our hypothesis is that there is no effect on individuals born abroad. We do not have information when the individuals returned to Iraq and those born abroad could return to Iraq to study at the primary school during the Islamic Republic of Iran–Iraq war. However, children born abroad had higher opportunity to send back to foreign countries if they returned to Iraq during the Islamic Republic of Iran–Iraq war. This suggests that those children would not be affected by the Islamic Republic of Iran–Iraq war.\\n\\nIndeed, Columns 1-4 of Table 4 show that cohorts of 1969-1981 are not statistically significant for different measures of educational attainment. The results are similar when we control for birth governorate fixed-effects, household size and gender. The findings generally support our hypothesis that it is the Islamic Republic of Iran–Iraq war 1980-1988 – rather than other warrelated factors (e.g., specific political regime) or unobserved region trends – that led to worse longterm educational outcomes for those who were exposed to the war during school age.\\n\\n**Table 4. Robustness checks**\\n\\n|Literacy|Primary|Secondary|Years of|Literacy|Primary|Secondary|Years of|\\n|---|---|---|---|---|---|---|---|\\n||education|education|education||education|education|education|\\n||completion|completion|||completion|completion||\\n||or above|or above|||or above|or above||\\n|(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|\\n\\n9\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"|Year 1964|0.333<sup>**</sup><br>(0.160)|0.250<br>(0.332)|0.250<br>(0.386)|8.500<sup>*</sup><br>(4.513)|0.331<sup>**</sup><br>(0.163)|0.324<br>(0.325)|0.358<br>(0.366)|9.465<sup>**</sup><br>(4.303)|\\n|---|---|---|---|---|---|---|---|---|\\n|Year 1965|<br>0.000|<br>0.250|<br>0.250|<br>0.000|<br>0.001|<br>0.275|<br>0.285|<br>0.000|\\n||(0.234)|(0.486)|(0.564)|(.)|(0.237)|(0.472)|(0.532)|(.)|\\n|Year 1966|<br>0.000|<br>-0.083|<br>-0.083|<br>1.000|<br>-0.010|<br>0.024|<br>0.075|<br>0.454|\\n||(0.160)|(0.332)|(0.386)|(3.685)|(0.163)|(0.325)|(0.367)|(3.517)|\\n|Year 1967|<br>0.000|<br>0.000|<br>-0.000|<br>-5.500|<br>0.012|<br>-0.156|<br>-0.230|<br>-5.164|\\n||(0.148)|(0.307)|(0.357)|(4.513)|(0.152)|(0.303)|(0.342)|(4.295)|\\n|Year 1969<br>(Year of|<br>0.000|<br>0.250|<br>-0.083|<br>1.000|<br>-0.010|<br>0.363|<br>0.084|<br>0.475|\\n|change)|||||||||\\n||(0.160)|(0.332)|(0.386)|(3.685)|(0.163)|(0.325)|(0.367)|(3.495)|\\n|Year 1970|<br>0.000|<br>0.250|<br>0.250|<br>3.000|<br>-0.011|<br>0.392|<br>0.459|<br>3.420|\\n||(0.181)|(0.377)|(0.437)|(3.685)|(0.185)|(0.368)|(0.416)|(3.493)|\\n|Year 1971|<br>0.000|<br>-0.750|<br>-0.750|<br>-2.500|<br>-0.007|<br>-0.596|<br>-0.524|<br>-1.325|\\n||(0.234)|(0.486)|(0.564)|(4.513)|(0.239)|(0.476)|(0.537)|(4.308)|\\n|Year 1972|<br>0.250<sup>*</sup>|<br>0.250|<br>-0.000|<br>3.833|<br>0.244|<br>0.313|<br>0.094|<br>3.386|\\n||(0.148)|(0.307)|(0.357)|(3.364)|(0.150)|(0.299)|(0.337)|(3.193)|\\n|Year 1973|<br>0.000|<br>-0.250|<br>-0.375|<br>-0.500|<br>-0.006|<br>-0.199|<br>-0.299|<br>-1.074|\\n||(0.128)|(0.266)|(0.309)|(3.009)|(0.130)|(0.260)|(0.294)|(2.867)|\\n|Year 1974|<br>0.000|<br>0.250|<br>-0.250|<br>2.250|<br>-0.005|<br>0.261|<br>-0.232|<br>0.876|\\n||(0.148)|(0.307)|(0.357)|(3.191)|(0.151)|(0.301)|(0.340)|(3.059)|\\n|Year 1975|<br>0.000|<br>-0.062|<br>-0.313|<br>1.100|<br>-0.006|<br>0.006|<br>-0.212|<br>0.644|\\n||(0.117)|(0.243)|(0.282)|(2.774)|(0.119)|(0.237)|(0.268)|(2.632)|\\n|Year 1976|<br>0.250<sup>*</sup>|<br>-0.250|<br>-0.375|<br>-2.000|<br>0.245<sup>*</sup>|<br>-0.184|<br>-0.277|<br>-2.343|\\n||(0.128)|(0.266)|(0.309)|(3.009)|(0.130)|(0.259)|(0.292)|(2.856)|\\n|Year 1977|<br>0.000|<br>0.250|<br>-0.000|<br>2.750|<br>-0.002|<br>0.282|<br>0.048|<br>2.236|\\n||(0.148)|(0.307)|(0.357)|(3.191)|(0.149)|(0.298)|(0.336)|(3.036)|\\n|Year 1978|<br>0.071|<br>0.107|<br>-0.179|<br>2.577|<br>0.064|<br>0.199|<br>-0.043|<br>2.490|\\n||(0.119)|(0.247)|(0.286)|(2.799)|(0.121)|(0.241)|(0.272)|(2.652)|\\n|Year 1979|<br>0.000|<br>0.050|<br>-0.050|<br>3.389|<br>-0.009|<br>0.170|<br>0.127|<br>3.613|\\n||(0.124)|(0.257)|(0.299)|(2.880)|(0.127)|(0.253)|(0.285)|(2.730)|\\n|Year 1980|<br>0.000|<br>-0.306|<br>-0.417|<br>-0.611|<br>-0.002|<br>-0.285|<br>-0.386|<br>-1.362|\\n||(0.126)|(0.261)|(0.303)|(2.880)|(0.127)|(0.253)|(0.286)|(2.744)|\\n|Year 1981|<br>0.000|<br>0.000|<br>-0.125|<br>-0.333|<br>-0.007|<br>0.084|<br>-0.001|<br>-0.830|\\n||(0.128)|(0.266)|(0.309)|(3.009)|(0.130)|(0.260)|(0.293)|(2.857)|\\n|_N_|106|106|106|86|106|106|106|86|\\n|adj._R_<sup>2</sup>|0.037|0.036|-0.028|0.101|0.017|0.096|0.089|0.193|\\n\\nNotes: Columns 5-8 add birth governorate fixed-effects, household size and gender. Standard errors are clustered at the birth governorate level. *** p<0.01, ** p<0.05, * p<0.1.\\n\\n<mark>There may also be concern that the estimation results may be affected by other factors that are not related to the war including, for example, whether the stated cohorts have disabilities, diseases, or illnesses. To test this hypothesis, we re-run the equation (1) for a variety of measures of health, they are dummy variable for illness or injury in the past 30 days, dummy variable for chronic disease and dummy variable for disability. Table 5 provides the estimation results, which indicate that cohorts of 1969-1981 are not statistically significant for dummy variable for illness or injury in the past 30 days, even when we add birth governorate fixed-effects, household size and gender (Columns 1 and 4). The results are statistically significant for the cohorts of 1972-1981 for dummy variable for chronic disease in both samples with and without birth governorate fixed-effects, household size and gender, however the signs are negative and unexpected (Columns 2 and 5).</mark> 10\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"<mark>Similarly, cohorts of 1971 and 1975-1981 are statistically significant for dummy variable for disability, but the signs are negative and unexpected (Column 3), the result is statistically insignificant when we control for birth governorate fixed-effects, household size and gender. Those findings suggest that exposure to the war during early childhood does not affect our estimation results.</mark>\\n\\n# **Table 5. Robustness checks (continue)**\\n\\n||Dummy<br>variable<br>for illness<br>or injury in<br>the past 30|Dummy<br>variable for<br>chronic<br>disease|Dummy<br>variable<br>for<br>disability|Dummy<br>variable for<br>illness or<br>injury in<br>the past 30|Dummy<br>variable for<br>chronic<br>disease|Dummy<br>variable for<br>disability|\\n|---|---|---|---|---|---|---|\\n||days|||days|||\\n||(1)|(2)|(3)|(4)|(5)|(6)|\\n|Year 1964|<br>-0.007|<br>0.060<sup>***</sup>|<br>-0.004|<br>-0.021|<br>0.069<sup>**</sup>|<br>0.010|\\n||(0.016)|(0.020)|(0.018)|(0.019)|(0.027)|(0.009)|\\n|Year 1965|<br>-0.015|<br>0.018|<br>-0.002|<br>-0.004|<br>0.019|<br>0.001|\\n||(0.012)|(0.021)|(0.016)|(0.017)|(0.022)|(0.007)|\\n|Year 1966|<br>-0.002|<br>0.051<sup>**</sup>|<br>0.002|<br>-0.038<sup>**</sup>|<br>0.025|<br>0.009|\\n||(0.016)|(0.023)|(0.015)|(0.017)|(0.029)|(0.008)|\\n|Year 1967|<br>-0.010|<br>0.030|<br>-0.007|<br>-0.029|<br>0.007|<br>-0.008|\\n||(0.018)|(0.021)|(0.013)|(0.019)|(0.024)|(0.005)|\\n|Year 1969 (Year of change)|<br>0.003|<br>-0.040|<br>-0.014|<br>-0.015|<br>-0.015|<br>-0.004|\\n||(0.018)|(0.024)|(0.011)|(0.014)|(0.017)|(0.007)|\\n|Year 1970|<br>0.004|<br>-0.016|<br>-0.012|<br>-0.014|<br>-0.020|<br>-0.002|\\n||(0.016)|(0.023)|(0.012)|(0.022)|(0.019)|(0.006)|\\n|Year 1971|-0.005|-0.016|-0.030<sup>**</sup>|-0.016|-0.013|-0.004|\\n||(0.016)|(0.020)|(0.012)|(0.020)|(0.023)|(0.008)|\\n|Year 1972|<br>-0.006|<br>-0.042<sup>**</sup>|<br>-0.023|<br>-0.024|<br>-0.032|<br>-0.004|\\n||(0.011)|(0.019)|(0.014)|(0.015)|(0.021)|(0.006)|\\n|Year 1973|<br>-0.011|<br>-0.061<sup>***</sup>|<br>-0.017|<br>-0.012|<br>-0.064<sup>***</sup>|<br>0.001|\\n||(0.015)|(0.018)|(0.011)|(0.019)|(0.018)|(0.007)|\\n|Year 1974|0.006|-0.062<sup>***</sup>|-0.022|0.003|-0.075<sup>***</sup>|-0.006|\\n||(0.012)|(0.019)|(0.018)|(0.013)|(0.021)|(0.006)|\\n|Year 1975|<br>0.007|<br>-0.042<sup>**</sup>|<br>-0.027<sup>**</sup>|<br>-0.011|<br>-0.056<sup>**</sup>|<br>-0.000|\\n||(0.013)|(0.018)|(0.013)|(0.015)|(0.020)|(0.010)|\\n|Year 1976|0.018|-0.066<sup>***</sup>|-0.024<sup>*</sup>|0.004|-0.081<sup>***</sup>|-0.003|\\n||(0.014)|(0.016)|(0.013)|(0.016)|(0.022)|(0.006)|\\n|Year 1977|<br>0.010|<br>-0.074<sup>***</sup>|<br>-0.029<sup>**</sup>|<br>-0.012|<br>-0.083<sup>***</sup>|<br>-0.001|\\n||(0.012)|(0.016)|(0.014)|(0.015)|(0.017)|(0.006)|\\n|Year 1978|<br>0.019|<br>-0.084<sup>***</sup>|<br>-0.028<sup>**</sup>|<br>-0.005|<br>-0.097<sup>***</sup>|<br>-0.010|\\n||(0.012)|(0.019)|(0.012)|(0.012)|(0.020)|(0.006)|\\n|Year 1979|0.019|-0.082<sup>***</sup>|-0.039<sup>**</sup>|-0.005|-0.090<sup>***</sup>|-0.007|\\n||(0.015)|(0.016)|(0.015)|(0.015)|(0.020)|(0.005)|\\n|Year 1980|<br>0.024|<br>-0.090<sup>***</sup>|<br>-0.037<sup>***</sup>|<br>0.004|<br>-0.090<sup>***</sup>|<br>0.001|\\n||(0.015)|(0.016)|(0.012)|(0.013)|(0.020)|(0.006)|\\n|||<sup>***</sup>|<sup>***</sup>||<sup>***</sup>||\\n|Year 1981|0.007|-0.095|-0.040|0.001|-0.106|-0.009|\\n||(0.017)|(0.016)|(0.012)|(0.016)|(0.018)|(0.006)|\\n|_N_|<br>14367|<br>14384|<br>14384|<br>14655|<br>14666|<br>14666|\\n|adj._R_<sup>2</sup>|0.043|0.043|0.004|0.057|0.055|0.001|\\n\\n11\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"Notes: Columns 4-6 add birth governorate fixed-effects, household size and gender. Standard errors are clustered at the birth governorate level. *** p<0.01, ** p<0.05, * p<0.1.\\n\\n# **4.4. Heterogeneity analysis**\\n\\n# **Female versus male**\\n\\nIn this section, we further examine the heterogeneous effects of the Islamic Republic of Iran–Iraq war on gender. Table 6 provides the estimation results on male and female educational attainment. Column 1 shows no evidence of the impact of the Islamic Republic of Iran–Iraq war on literacy for men. The result is negative and statistically significant for male primary education completion or above for cohorts of 1972-1981 (Column 2). Also, secondary education completion or above and years of education decreased for men for the cohort of 1972-1981. In terms of the impact of war on female education, we find that literacy, primary education completion or above, secondary education completion or above decreased for women, however, the result is not statistically significant for years of education (Columns 5-8). We find that the magnitude of the impact of the Islamic Republic of Iran–Iraq war is larger for female education than for male education, for example, the Islamic Republic of Iran–Iraq war decreased the probability of male cohort of 1972 completing primary education or above by 3.1% while this figure is 7.1% female cohort of 1972. Overall, female education was more severely affected by the Islamic Republic of Iran–Iraq war than male education.\\n\\n**Table 6. Educational attainment between males and females**\\n\\n|||M|ale|||Fe|male||\\n|---|---|---|---|---|---|---|---|---|\\n||Literacy|Primary<br>education<br>completion<br>or above|Secondary<br>education<br>completion<br>or above|Years of<br>education|Literacy|Primary<br>education<br>completion<br>or above|Secondary<br>education<br>completion<br>or above|Years of<br>education|\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|\\n|Year 1964|0.035<sup>***</sup>|0.027|0.063<sup>**</sup>|-0.036|0.038<sup>**</sup>|0.019|0.131<sup>***</sup>|0.274|\\n||(0.011)|(0.021)|(0.029)|(0.194)|(0.015)|(0.028)|(0.039)|(0.319)|\\n|Year 1965|0.011|0.024|0.090<sup>***</sup>|0.131|0.018|0.005|0.081<sup>**</sup>|0.082|\\n||(0.006)|(0.019)|(0.027)|(0.244)|(0.013)|(0.021)|(0.029)|(0.234)|\\n|Year 1966|0.016<sup>**</sup>|0.022|0.064|0.061|0.047<sup>***</sup>|0.009|0.087<sup>***</sup>|0.117|\\n||(0.008)|(0.014)|(0.037)|(0.219)|(0.015)|(0.018)|(0.024)|(0.141)|\\n|Year 1967|0.007<sup>*</sup>|-0.000|0.060<sup>**</sup>|0.150|0.009|-0.000|0.048|0.198|\\n||(0.004)|(0.023)|(0.026)|(0.272)|(0.009)|(0.024)|(0.029)|(0.247)|\\n|Year 1969 (Year|0.006|0.004|0.021|-0.169|-0.014|-0.004|0.006|0.226|\\n|of change)|||||||||\\n||(0.005)|(0.018)|(0.026)|(0.208)|(0.013)|(0.015)|(0.025)|(0.213)|\\n|Year 1970|0.003|-0.005|-0.025|-0.391|-0.011|-0.046<sup>**</sup>|-0.029|0.085|\\n||(0.004)|(0.017)|(0.030)|(0.238)|(0.007)|(0.019)|(0.025)|(0.171)|\\n|Year 1971|0.003|0.001|-0.014|-0.288|-0.029<sup>**</sup>|-0.084<sup>***</sup>|-0.066<sup>**</sup>|0.082|\\n||(0.004)|(0.018)|(0.028)|(0.199)|(0.010)|(0.020)|(0.025)|(0.176)|\\n|Year 1972|<br>0.004|<br>-0.031<sup>*</sup>|<br>-0.037|<br>-0.473<sup>**</sup>|<br>-0.026<sup>**</sup>|<br>-0.071<sup>**</sup>|<br>-0.077<sup>***</sup>|<br>-0.193|\\n||(0.004)|(0.018)|(0.023)|(0.205)|(0.010)|(0.026)|(0.026)|(0.161)|\\n|Year 1973|-0.002|-0.053<sup>***</sup>|-0.053<sup>**</sup>|-0.551<sup>***</sup>|-0.019<sup>*</sup>|-0.065<sup>**</sup>|-0.061<sup>**</sup>|-0.315|\\n\\n12\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"||(0.002)|(0.017)|(0.024)|(0.186)|(0.010)|(0.028)|(0.025)|(0.193)|\\n|---|---|---|---|---|---|---|---|---|\\n|Year 1974|0.003|-0.056<sup>***</sup>|-0.033|-0.465<sup>**</sup>|-0.016|-0.058<sup>***</sup>|-0.073<sup>***</sup>|-0.335|\\n||(0.005)|(0.019)|(0.025)|(0.191)|(0.010)|(0.018)|(0.021)|(0.195)|\\n|Year 1975|0.001|-0.032<sup>**</sup>|-0.026|-0.453<sup>*</sup>|-0.027<sup>***</sup>|-0.043<sup>**</sup>|-0.009|0.236|\\n||(0.002)|(0.012)|(0.032)|(0.246)|(0.009)|(0.020)|(0.023)|(0.237)|\\n|Year 1976|0.003|-0.055<sup>**</sup>|-0.055<sup>*</sup>|-0.435<sup>*</sup>|-0.024<sup>***</sup>|-0.047<sup>**</sup>|-0.054<sup>**</sup>|-0.028|\\n||(0.003)|(0.023)|(0.027)|(0.234)|(0.008)|(0.020)|(0.024)|(0.203)|\\n|Year 1977|0.004|-0.052<sup>***</sup>|-0.018|-0.322|-0.025<sup>**</sup>|-0.045<sup>***</sup>|-0.044<sup>*</sup>|0.080|\\n||(0.003)|(0.015)|(0.029)|(0.212)|(0.009)|(0.015)|(0.022)|(0.252)|\\n|Year 1978|0.003|-0.078<sup>***</sup>|-0.073<sup>**</sup>|-0.704<sup>***</sup>|-0.027<sup>**</sup>|-0.056<sup>***</sup>|-0.083<sup>***</sup>|-0.151|\\n||(0.003)|(0.021)|(0.028)|(0.234)|(0.010)|(0.019)|(0.022)|(0.209)|\\n|Year 1979|0.002|-0.066<sup>***</sup>|-0.049<sup>**</sup>|-0.453<sup>***</sup>|-0.029<sup>***</sup>|-0.047<sup>**</sup>|-0.063<sup>**</sup>|0.230|\\n||(0.003)|(0.016)|(0.021)|(0.152)|(0.008)|(0.019)|(0.023)|(0.191)|\\n|Year 1980|-0.001|-0.087<sup>***</sup>|-0.032|-0.449|-0.027<sup>***</sup>|-0.054<sup>***</sup>|-0.058<sup>***</sup>|0.131|\\n||(0.003)|(0.018)|(0.029)|(0.263)|(0.008)|(0.013)|(0.020)|(0.240)|\\n|Year 1981|0.002|-0.080<sup>***</sup>|-0.054<sup>*</sup>|-0.536<sup>**</sup>|-0.038<sup>***</sup>|-0.057<sup>***</sup>|-0.105<sup>***</sup>|-0.102|\\n||(0.003)|(0.018)|(0.029)|(0.221)|(0.009)|(0.016)|(0.023)|(0.225)|\\n|_N_|14384|14384|14384|13452|14666|14666|14666|11672|\\n|adj._R_<sup>2</sup>|0.013|0.045|0.029|0.042|0.023|0.032|0.031|0.062|\\n\\nNotes: All regressions add birth governorate fixed-effects, household size and gender. Standard errors are clustered at the birth governorate level. *** p<0.01, ** p<0.05, * p<0.1.\\n\\nNow, we consider the impact of the Islamic Republic of Iran–Iraq war on female and male labor market outcomes. The results show that most of the male cohorts of 1970-1981 are statistically significant for dummy variables for social security or pension plan, however, the female cohorts of 1970-1981 are not statistically significant. Similarly, we find that the Islamic Republic of Iran– Iraq war decreased the probability of men having health care, paid leave and job permanence for most of cohorts of 1969-1981. While we find little evidence of the impact of the Islamic Republic of Iran–Iraq war on women for the cohorts of 1969-1981. In terms of wage, the results are negative and highly statistically significant for most of the male cohorts of 1969-1981. The Islamic Republic of Iran–Iraq war also has negative impacts on female wages for the cohorts of 1973 - 1981.\\n\\nFemale labor force participation in Iraq is limited due to lack of government backing and social norms. Women's labor force participation in Iraq is low: only 12.3% of working-age women in Iraq were employed or searching for employment in 2018 (Reach, 2019). Table 7 also shows that only 1,715 women participated in the labor market while 8,823 men did in the labor market. In other hand, women participating in the labor market accounted for only 16.3 percent in 2006 in Iraq. Furthermore, of those workers who were exposed to the Islamic Republic of Iran–Iraq war during childhood, about 72 percent of workers with social security, pension, or paid leave were male, 74 percent of workers with health care were male, and 85 percent of workers with job permanence were male in 2006. This may suggest why the effect of the Islamic Republic of Iran– Iraq war on female labor market outcomes is smaller compared with male market outcomes.\\n\\n13\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"# **Table 7. Social welfare of workers between males and females**\\n\\n||||M|ale|||||Fe|male|||\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n||Dummy<br>variable<br>for social<br>security|Dummy<br>variable<br>for<br>pension<br>plan|Dummy<br>variable for<br>health care|Dummy<br>variable<br>for paid<br>leave|Dummy<br>variable for<br>job<br>permanence|Log of<br>wage|Dummy<br>variable<br>for social<br>security|Dummy<br>variable<br>for<br>pension<br>plan|Dummy<br>variable<br>for health<br>care|Dummy<br>variable<br>for paid<br>leave|Dummy<br>variable for<br>job<br>permanence|Log of<br>wage|\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|(9)|(10)|(11)|(12)|\\n|Year 1964|-0.010|0.003|-0.011|0.028|0.029|0.120|0.014|0.037|-0.098<sup>*</sup>|0.028|0.017|0.156<sup>*</sup>|\\n||(0.049)|(0.047)|(0.035)|(0.051)|(0.027)|(0.165)|(0.031)|(0.025)|(0.049)|(0.037)|(0.030)|(0.075)|\\n|Year 1965|-0.021|-0.013|0.002|-0.003|-0.040|-0.106|0.058|0.058<sup>*</sup>|-0.083<sup>*</sup>|0.079<sup>*</sup>|0.029|0.064|\\n||(0.040)|(0.040)|(0.036)|(0.036)|(0.027)|(0.145)|(0.038)|(0.030)|(0.045)|(0.042)|(0.025)|(0.076)|\\n|Year 1966|-0.010|0.004|0.023|0.030|-0.007|-0.016|-0.022|-0.024|-0.131|-0.001|0.036|-0.126|\\n||(0.031)|(0.034)|(0.028)|(0.028)|(0.023)|(0.140)|(0.034)|(0.040)|(0.084)|(0.041)|(0.033)|(0.121)|\\n|Year 1967|-0.048|-0.041|-0.039|-0.008|-0.013|-0.081|-0.027|-0.023|-0.142<sup>**</sup>|0.017|0.040|-0.167<sup>**</sup>|\\n||(0.041)|(0.039)|(0.034)|(0.047)|(0.022)|(0.128)|(0.048)|(0.032)|(0.052)|(0.049)|(0.043)|(0.062)|\\n|Year 1969|-0.069|-0.073|-0.078<sup>**</sup>|-0.083<sup>*</sup>|-0.025|-0.224<sup>*</sup>|-0.008|-0.009|-0.090<sup>*</sup>|0.035|0.012|-0.061|\\n|(Year of<br>|||||||||||||\\n|change)|||||||||||||\\n||(0.040)|(0.044)|(0.032)|(0.044)|(0.025)|(0.127)|(0.045)|(0.044)|(0.050)|(0.053)|(0.034)|(0.105)|\\n|Year 1970|-0.066<sup>*</sup>|-0.065|-0.039|-0.056|-0.036<sup>*</sup>|-0.216<sup>**</sup>|-0.008|-0.029|-0.120<sup>**</sup>|0.053|0.047|-0.180<sup>*</sup>|\\n||(0.037)|(0.039)|(0.030)|(0.036)|(0.020)|(0.095)|(0.044)|(0.047)|(0.051)|(0.054)|(0.036)|(0.091)|\\n|Year 1971|-0.048|-0.049|-0.048<sup>*</sup>|-0.042|-0.002|-0.164|-0.006|0.005|-0.136<sup>*</sup>|0.029|0.031|-0.107|\\n||(0.034)|(0.035)|(0.026)|(0.033)|(0.019)|(0.131)|(0.051)|(0.042)|(0.065)|(0.052)|(0.036)|(0.078)|\\n|Year 1972|-0.085<sup>**</sup>|-0.085<sup>**</sup>|-0.040|-0.072<sup>*</sup>|-0.032<sup>*</sup>|-0.295<sup>***</sup>|0.041|0.047|-0.053|0.078<sup>*</sup>|0.059<sup>**</sup>|-0.062|\\n||(0.032)|(0.035)|(0.025)|(0.038)|(0.017)|(0.090)|(0.037)|(0.028)|(0.039)|(0.039)|(0.025)|(0.069)|\\n|Year 1973|-0.113<sup>***</sup>|-0.116<sup>***</sup>|-0.045|-0.094<sup>**</sup>|-0.036<sup>**</sup>|-0.262<sup>**</sup>|0.024|0.021|-0.087<sup>*</sup>|0.038|0.021|-0.214<sup>**</sup>|\\n||(0.035)|(0.036)|(0.033)|(0.040)|(0.017)|(0.116)|(0.047)|(0.036)|(0.042)|(0.053)|(0.034)|(0.085)|\\n|Year 1974|-0.047|-0.057|-0.047|-0.034|-0.005|-0.115|0.035|0.061<sup>*</sup>|-0.034|0.076|0.056<sup>*</sup>|-0.037|\\n||(0.042)|(0.044)|(0.032)|(0.041)|(0.022)|(0.135)|(0.044)|(0.031)|(0.052)|(0.046)|(0.030)|(0.077)|\\n|Year 1975|-0.050|-0.069<sup>*</sup>|-0.045|-0.040|-0.018|-0.134|0.008|0.003|-0.040|0.051|0.030|-0.300<sup>**</sup>|\\n||(0.036)|(0.034)|(0.033)|(0.034)|(0.016)|(0.119)|(0.039)|(0.035)|(0.071)|(0.046)|(0.036)|(0.116)|\\n|Year 1976|-0.072|-0.060|-0.037|-0.037|-0.012|-0.195<sup>*</sup>|-0.017|-0.017|-0.083|0.021|0.029|-0.237<sup>***</sup>|\\n||(0.044)|(0.044)|(0.033)|(0.039)|(0.020)|(0.106)|(0.038)|(0.038)|(0.058)|(0.040)|(0.024)|(0.064)|\\n|Year 1977|-0.077<sup>*</sup>|-0.086<sup>*</sup>|-0.051<sup>**</sup>|-0.058<sup>*</sup>|-0.041<sup>*</sup>|-0.377<sup>***</sup>|-0.006|0.012|-0.058|0.065|0.086<sup>**</sup>|-0.185<sup>**</sup>|\\n||(0.039)|(0.044)|(0.023)|(0.030)|(0.022)|(0.099)|(0.052)|(0.045)|(0.076)|(0.052)|(0.034)|(0.071)|\\n|Year 1978|-0.087<sup>**</sup>|-0.079<sup>*</sup>|-0.027|-0.066|-0.008|-0.360<sup>***</sup>|0.049|0.046|-0.172<sup>***</sup>|0.086<sup>*</sup>|0.063<sup>**</sup>|-0.178<sup>*</sup>|\\n||(0.041)|(0.043)|(0.031)|(0.039)|(0.023)|(0.116)|(0.044)|(0.035)|(0.053)|(0.049)|(0.026)|(0.096)|\\n|Year 1979|-0.113<sup>***</sup>|-0.111<sup>***</sup>|-0.058<sup>**</sup>|-0.091<sup>**</sup>|-0.065<sup>***</sup>|-0.466<sup>***</sup>|0.026|0.020|-0.031|0.061|0.046<sup>*</sup>|-0.256<sup>**</sup>|\\n||(0.032)|(0.034)|(0.027)|(0.032)|(0.019)|(0.108)|(0.050)|(0.040)|(0.048)|(0.045)|(0.025)|(0.091)|\\n\\n14\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"|Year 1980|-0.102<sup>***</sup>|-0.110<sup>***</sup>|-0.070<sup>**</sup>|-0.082<sup>**</sup>|-0.046<sup>*</sup>|-0.403<sup>***</sup>|-0.024|-0.006|-0.113<sup>**</sup>|0.049|0.068|-0.259<sup>***</sup>|\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n||(0.035)|(0.037)|(0.028)|(0.037)|(0.024)|(0.095)|(0.050)|(0.040)|(0.051)|(0.044)|(0.051)|(0.079)|\\n|Year 1981|-0.152<sup>***</sup>|-0.157<sup>***</sup>|-0.063<sup>**</sup>|-0.109<sup>**</sup>|-0.056<sup>**</sup>|-0.523<sup>***</sup>|0.021|0.040|-0.054|0.078|0.040|-0.370<sup>***</sup>|\\n||(0.034)|(0.038)|(0.028)|(0.038)|(0.020)|(0.115)|(0.038)|(0.029)|(0.045)|(0.046)|(0.024)|(0.098)|\\n|_N_|8823|8823|8823|8823|8823|8577|1715|1715|1715|1715|1715|1661|\\n|adj._R_<sup>2</sup>|0.035|0.038|0.083|0.032|0.020|0.047|0.011|0.029|0.179|0.028|0.066|0.033|\\n\\nNotes: All regressions add birth governorate fixed-effects, household size and gender. Standard errors are clustered at the birth governorate level. *** p<0.01, ** p<0.05, * p<0.1.\\n\\n15\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"Table 8 provides heterogeneous analysis on working conditions of workers. The results show men of cohorts of 1969 to 1981 are more likely to work in dangerous conditions and work without air conditioning while we find no evidence of the impact of the Islamic Republic of Iran–Iraq war on working conditions of female workers like hazardous job and air conditioning. Descriptive statistics shows that female workers work in hazardous conditions or without air conditioning accounted for only 26 percent in the survey. This explains why we find no impacts of the Islamic Republic of Iran– Iraq war on working conditions, including hazardous job and air conditioning.\\n\\nRegarding health impact, health of men of the cohorts of 1972, 1979, and 1980 was adversely affected at the working environment, while we find the evidence of the negative impact for health of women for all cohorts of 1970-1981.\\n\\n**Table 8. Working conditions of workers between males and females**\\n\\n|||Male|||Female||\\n|---|---|---|---|---|---|---|\\n||Hazardous<br>job|Air<br>conditioning|Health<br>impact|Hazardous<br>job|Air<br>conditioning|Health<br>impact|\\n||(1)|(2)|(3)|(4)|(5)|(6)|\\n|Year 1964|0.002|-0.009|0.039|0.048|0.034|0.603<sup>**</sup>|\\n||(0.052)|(0.046)|(0.216)|(0.045)|(0.065)|(0.262)|\\n|Year 1965|-0.026|-0.043|-0.182|0.028|0.029|0.491<sup>**</sup>|\\n||(0.033)|(0.031)|(0.150)|(0.046)|(0.059)|(0.228)|\\n|Year 1966|0.006|-0.012|-0.010|0.002|0.059|0.903<sup>**</sup>|\\n||(0.029)|(0.038)|(0.158)|(0.056)|(0.062)|(0.357)|\\n|Year 1967|0.011|-0.042|0.020|0.001|-0.003|0.413<sup>**</sup>|\\n||(0.038)|(0.039)|(0.135)|(0.047)|(0.065)|(0.171)|\\n|Year 1969 (Year of change)|<br>-0.018|<br>-0.095<sup>**</sup>|<br>0.054|<br>-0.018|<br>0.045|<br>0.376|\\n||(0.036)|(0.037)|(0.122)|(0.051)|(0.049)|(0.277)|\\n|Year 1970|-0.048|-0.097<sup>**</sup>|0.172|0.004|-0.028|0.451<sup>*</sup>|\\n||(0.037)|(0.045)|(0.120)|(0.051)|(0.052)|(0.215)|\\n|Year 1971|-0.061|-0.053<sup>**</sup>|-0.056|-0.036|0.031|0.769<sup>**</sup>|\\n||(0.035)|(0.019)|(0.126)|(0.076)|(0.057)|(0.315)|\\n|Year 1972|<br>-0.068<sup>*</sup>|<br>-0.111<sup>***</sup>|<br>0.265<sup>*</sup>|<br>0.006|<br>0.029|<br>0.452<sup>*</sup>|\\n||(0.034)|(0.031)|(0.153)|(0.034)|(0.055)|(0.231)|\\n|Year 1973|<br>-0.102<sup>**</sup>|<br>-0.089<sup>***</sup>|<br>0.006|<br>-0.100<sup>*</sup>|<br>-0.034|<br>0.536<sup>**</sup>|\\n||(0.036)|(0.026)|(0.143)|(0.050)|(0.068)|(0.227)|\\n|Year 1974|<br>-0.109<sup>***</sup>|<br>-0.069<sup>**</sup>|<br>0.270|<br>-0.015|<br>0.024|<br>0.703<sup>***</sup>|\\n||(0.031)|(0.031)|(0.164)|(0.060)|(0.083)|(0.212)|\\n|Year 1975|<br>-0.049|<br>-0.082<sup>***</sup>|<br>0.030|<br>-0.059|<br>-0.055|<br>0.415<sup>**</sup>|\\n||(0.039)|(0.025)|(0.136)|(0.038)|(0.067)|(0.159)|\\n|Year 1976|-0.058|-0.049|0.126|-0.003|0.050|0.473<sup>*</sup>|\\n||(0.043)|(0.031)|(0.111)|(0.044)|(0.062)|(0.253)|\\n|Year 1977|<br>-0.065<sup>*</sup>|<br>-0.072<sup>**</sup>|<br>0.188|<br>-0.008|<br>0.004|<br>0.616<sup>**</sup>|\\n||(0.034)|(0.032)|(0.142)|(0.056)|(0.075)|(0.288)|\\n|Year 1978|<br>-0.044<sup>*</sup>|<br>-0.085<sup>**</sup>|<br>0.214|<br>0.052|<br>-0.051|<br>0.608<sup>*</sup>|\\n||(0.024)|(0.035)|(0.142)|(0.059)|(0.064)|(0.301)|\\n|Year 1979|<br>-0.047|<br>-0.115<sup>***</sup>|<br>0.303<sup>**</sup>|<br>0.024|<br>0.052|<br>0.876<sup>***</sup>|\\n||(0.038)|(0.025)|(0.114)|(0.042)|(0.066)|(0.221)|\\n|Year 1980|<br>-0.067<sup>**</sup>|<br>-0.089<sup>***</sup>|<br>0.133|<br>-0.042|<br>-0.005|<br>0.553<sup>**</sup>|\\n||(0.025)|(0.028)|(0.137)|(0.051)|(0.065)|(0.218)|\\n|Year 1981|<br>-0.078<sup>**</sup>|<br>-0.149<sup>***</sup>|<br>0.290<sup>**</sup>|<br>-0.012|<br>0.036|<br>0.833<sup>***</sup>|\\n\\n16\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"||(0.032)|(0.032)|(0.111)|(0.039)|(0.043)|(0.287)|\\n|---|---|---|---|---|---|---|\\n|_N_|8823|8823|8823|1715|1715|1715|\\n|adj. _R_<sup>2</sup>|0.047|0.032|0.148|0.078|0.052|0.195|\\n\\nNotes: All regressions add birth governorate fixed-effects, household size and gender. Standard errors are clustered at the birth governorate level. *** p<0.01, ** p<0.05, * p<0.1.\\n\\n# **Severity of the Islamic Republic of Iran–Iraq war**\\n\\n<mark>Educational attainment and labor market outcomes may be different between governorates. The children living in the areas with more intense fighting should do worse in school and later in the labor markets. The conflict concentrated along the countries’ joint border. Therefore, we expect that educational attainment of children born in the governorates near the Islamic Republic of Iran’s border would be worse than educational attainment of children born in other governorates in Iraq. We divide the sample into two sub-sample: a sub-sample with children born in the governorates near the Islamic Republic of Iran’s border and other sub-sample with children born in other governorates in Iraq.</mark> Table 9 reports the heterogeneous results on educational attainment between governorates near the Islamic Republic of Iran’s border and other Iraqi governorates. The Islamic Republic of Iran–Iraq war has larger impacts on literacy of cohorts 1975-1980 who born in governorates near the Islamic Republic of Iran’s border compared with literacy of those born in other governorates in Iraq. The impacts of the Islamic Republic of Iran–Iraq war on primary education completion or above is lower in governorates near the Islamic Republic of Iran’s border than in other governorates in Iraq. However, the impacts of the Islamic Republic of Iran–Iraq war on secondary education completion or above is greater in governorates near the Islamic Republic of Iran’s border than in other governorates in Iraq. Overall, we find that cohorts of 1970, 1972, 1974 who attain lower years of education in governorates near the Islamic Republic of Iran’s border than in other governorates in Iraq.\\n\\n**Table 9. Educational attainment between border governorates and other Iraqi governorates**\\n\\n||Governor<br>|ates near the Is<br>bor<br>|lamic Republi<br>der<br>|c of Iran’s<br>||Other gove<br>|rnorates in Iraq<br>|<br>|\\n|---|---|---|---|---|---|---|---|---|\\n||Literacy|Primary<br>education<br>completion<br>or above|Secondary<br>education<br>completion<br>or above|Years of<br>education|Literacy|Primary<br>education<br>completion<br>or above|Secondary<br>education<br>completion<br>or above|Years of<br>education|\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|\\n|Year 1964|0.027|0.043|0.047<sup>**</sup>|-0.107|0.040<sup>**</sup>|0.013|0.122<sup>***</sup>|0.170|\\n||(0.013)|(0.026)|(0.014)|(0.210)|(0.013)|(0.022)|(0.027)|(0.200)|\\n|Year 1965|0.023<sup>*</sup>|0.050|0.054|0.154|0.012|-0.000|0.101<sup>***</sup>|0.110|\\n||(0.011)|(0.026)|(0.030)|(0.394)|(0.009)|(0.019)|(0.030)|(0.197)|\\n|Year 1966|0.024<sup>*</sup>|0.039|-0.018|-0.239|0.035<sup>**</sup>|0.005|0.115<sup>***</sup>|0.218|\\n||(0.010)|(0.035)|(0.027)|(0.160)|(0.012)|(0.013)|(0.023)|(0.161)|\\n|Year 1967|0.003|0.023|0.009|-0.073|0.009|-0.011|0.074<sup>***</sup>|0.302|\\n||(0.011)|(0.015)|(0.044)|(0.353)|(0.006)|(0.019)|(0.020)|(0.198)|\\n|Year 1969|-0.016|0.013|-0.015|-0.214|0.001|-0.006|0.027|0.086|\\n|(Year of|||||||||\\n\\n17\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"|change)|||||||||\\n|---|---|---|---|---|---|---|---|---|\\n||(0.011)|(0.028)|(0.032)|(0.233)|(0.007)|(0.009)|(0.020)|(0.166)|\\n|Year 1970|<br>-0.005|<br>0.001|<br>-0.110<sup>***</sup>|<br>-0.521<sup>**</sup>|<br>-0.004|<br>-0.038|<br>0.008|<br>-0.043|\\n||(0.005)|(0.012)|(0.023)|(0.168)|(0.006)|(0.021)|(0.020)|(0.141)|\\n|Year 1971|<br>-0.018|<br>-0.025|<br>-0.089<sup>***</sup>|<br>-0.124|<br>-0.011<sup>*</sup>|<br>-0.047<sup>**</sup>|<br>-0.017|<br>-0.124|\\n||(0.011)|(0.014)|(0.021)|(0.221)|(0.005)|(0.015)|(0.029)|(0.177)|\\n|Year 1972|<br>-0.021|<br>-0.029|<br>-0.106<sup>**</sup>|<br>-0.558<sup>*</sup>|<br>-0.008|<br>-0.061<sup>**</sup>|<br>-0.034<sup>*</sup>|<br>-0.285<sup>*</sup>|\\n||(0.014)|(0.026)|(0.029)|(0.268)|(0.006)|(0.023)|(0.017)|(0.147)|\\n|Year 1973|<br>-0.012|<br>-0.020|<br>-0.094<sup>**</sup>|<br>-0.433|<br>-0.010<sup>*</sup>|<br>-0.075<sup>***</sup>|<br>-0.039<sup>**</sup>|<br>-0.439<sup>***</sup>|\\n||(0.012)|(0.032)|(0.027)|(0.267)|(0.005)|(0.019)|(0.016)|(0.122)|\\n|Year 1974|<br>-0.010|<br>-0.042<sup>*</sup>|<br>-0.093<sup>**</sup>|<br>-0.605<sup>*</sup>|<br>-0.005|<br>-0.064<sup>***</sup>|<br>-0.035<sup>**</sup>|<br>-0.322<sup>**</sup>|\\n||(0.008)|(0.020)|(0.032)|(0.261)|(0.008)|(0.014)|(0.013)|(0.124)|\\n|Year 1975|<br>-0.017<sup>*</sup>|<br>-0.032|<br>-0.044|<br>-0.185|<br>-0.011<sup>*</sup>|<br>-0.038<sup>**</sup>|<br>-0.003|<br>-0.119|\\n||(0.007)|(0.027)|(0.041)|(0.529)|(0.006)|(0.016)|(0.023)|(0.155)|\\n|Year 1976|<br>-0.017<sup>*</sup>|<br>-0.030<sup>**</sup>|<br>-0.123<sup>***</sup>|<br>-0.612|<br>-0.010|<br>-0.059<sup>**</sup>|<br>-0.023|<br>-0.088|\\n||(0.008)|(0.009)|(0.028)|(0.309)|(0.005)|(0.020)|(0.020)|(0.133)|\\n|Year 1977|-0.016<sup>*</sup>|-0.017|-0.069|-0.135|-0.009|-0.062<sup>***</sup>|-0.014|-0.142|\\n||(0.007)|(0.021)|(0.035)|(0.233)|(0.007)|(0.012)|(0.012)|(0.140)|\\n|Year 1978|<br>-0.014|<br>-0.045<sup>**</sup>|<br>-0.085<sup>*</sup>|<br>-0.451|<br>-0.012<sup>*</sup>|<br>-0.078<sup>***</sup>|<br>-0.073<sup>***</sup>|<br>-0.445<sup>**</sup>|\\n||(0.014)|(0.015)|(0.036)|(0.501)|(0.006)|(0.022)|(0.019)|(0.163)|\\n|Year 1979|<br>-0.019<sup>**</sup>|<br>-0.026|<br>-0.088|<br>0.027|<br>-0.011|<br>-0.070<sup>***</sup>|<br>-0.042<sup>***</sup>|<br>-0.233<sup>**</sup>|\\n||(0.007)|(0.013)|(0.045)|(0.365)|(0.006)|(0.015)|(0.012)|(0.104)|\\n|Year 1980|<br>-0.020<sup>*</sup>|<br>-0.052<sup>***</sup>|<br>-0.099<sup>*</sup>|<br>-0.265|<br>-0.012<sup>**</sup>|<br>-0.076<sup>***</sup>|<br>-0.020|<br>-0.151|\\n||(0.008)|(0.012)|(0.040)|(0.604)|(0.005)|(0.013)|(0.015)|(0.148)|\\n|Year 1981|<br>-0.027<sup>***</sup>|<br>-0.062<sup>**</sup>|<br>-0.140<sup>***</sup>|<br>-0.556|<br>-0.015<sup>**</sup>|<br>-0.071<sup>***</sup>|<br>-0.054<sup>**</sup>|<br>-0.255|\\n||(0.005)|(0.021)|(0.033)|(0.310)|(0.005)|(0.013)|(0.023)|(0.162)|\\n|_N_|<br>8802|<br>8802|<br>8802|<br>7262|<br>20142|<br>20142|<br>20142|<br>17776|\\n|adj._R_<sup>2</sup>|0.017|0.051|0.027|0.086|0.020|0.032|0.023|0.076|\\n\\nNotes: All regressions add birth governorate fixed-effects, household size and gender. Standard errors are clustered at the birth governorate level. *** p<0.01, ** p<0.05, * p<0.1.\\n\\nTable 10 provides the results on social welfare disparities between border governorates and other Iraqi governorates. Workers born in the governorates near the Islamic Republic of Iran’s border have lower opportunities of accessing to social security and pension plan than those born in other Iraqi governorates.\\n\\n18\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"# **Table 10. Social welfare disparities between border governorates and other Iraqi governorates**\\n\\n||G|overnorates|near the Isla|mic Republi|c of Iran’s bord|er|||Other Iraqi|governorate|s||\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n||Dummy<br>variable<br>for social|Dummy<br>variable<br>for|Dummy<br>variable<br>for health|Dummy<br>variable<br>for paid|Dummy<br>variable for<br>job|Log of<br>wage|Dummy<br>variable<br>for social|Dummy<br>variable<br>for|Dummy<br>variable<br>for health|Dummy<br>variable<br>for paid|Dummy<br>variable for<br>job|Log of<br>wage|\\n||security|pension<br>plan|care|leave|permanence||security|pension<br>plan|care|leave|permanence||\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|(9)|(10)|(11)|(12)|\\n|Year 1964|-0.087<sup>*</sup>|-0.066|-0.013|-0.088<sup>*</sup>|0.017|-0.054|0.021|0.032|-0.036|0.068|0.021|0.198|\\n||(0.037)|(0.044)|(0.043)|(0.035)|(0.054)|(0.148)|(0.057)|(0.053)|(0.041)|(0.053)|(0.021)|(0.185)|\\n|Year 1965|0.032|0.064|0.065|0.042|-0.014|0.067|-0.025|-0.030|-0.044|-0.002|-0.036|-0.124|\\n||(0.050)|(0.049)|(0.054)|(0.047)|(0.046)|(0.168)|(0.045)|(0.044)|(0.034)|(0.042)|(0.027)|(0.160)|\\n|Year 1966|-0.040|-0.025|0.058|-0.006|-0.013|-0.259|-0.009|0.001|-0.036|0.030|0.003|0.033|\\n||(0.059)|(0.058)|(0.044)|(0.051)|(0.036)|(0.247)|(0.028)|(0.031)|(0.035)|(0.022)|(0.024)|(0.128)|\\n|Year 1967|<br>-0.097<sup>*</sup>|<br>-0.087|<br>-0.078<sup>*</sup>|<br>-0.084|<br>-0.042|<br>-0.375<sup>***</sup>|<br>-0.017|<br>-0.011|<br>-0.039|<br>0.038|<br>0.017|<br>0.040|\\n||(0.042)|(0.045)|(0.032)|(0.073)|(0.036)|(0.082)|(0.044)|(0.041)|(0.039)|(0.048)|(0.022)|(0.139)|\\n|Year 1969 (Year|<br>-0.061|<br>-0.069<sup>**</sup>|<br>-0.041|<br>-0.063<sup>*</sup>|<br>-0.023|<br>-0.250<sup>**</sup>|<br>-0.054|<br>-0.055|<br>-0.088<sup>**</sup>|<br>-0.059|<br>-0.017|<br>-0.176|\\n|of change)|||||||||||||\\n||(0.036)|(0.025)|(0.037)|(0.028)|(0.049)|(0.078)|(0.050)|(0.056)|(0.032)|(0.054)|(0.025)|(0.158)|\\n|Year 1970|-0.091<sup>**</sup>|-0.089<sup>*</sup>|-0.048|-0.050|-0.034|-0.216<sup>*</sup>|-0.040|-0.045|-0.053|-0.031|-0.011|-0.215<sup>*</sup>|\\n||(0.023)|(0.038)|(0.038)|(0.050)|(0.022)|(0.092)|(0.040)|(0.043)|(0.035)|(0.037)|(0.022)|(0.115)|\\n|Year 1971|-0.019|-0.022|-0.024|-0.015|-0.007|-0.264|-0.056|-0.052|-0.084<sup>***</sup>|-0.040|0.006|-0.107|\\n||(0.040)|(0.033)|(0.045)|(0.042)|(0.027)|(0.155)|(0.037)|(0.041)|(0.023)|(0.036)|(0.022)|(0.142)|\\n|Year 1972|<br>-0.036|<br>-0.026|<br>-0.001|<br>-0.047|<br>-0.028|<br>-0.229<sup>*</sup>|<br>-0.075<sup>*</sup>|<br>-0.079<sup>*</sup>|<br>-0.059<sup>**</sup>|<br>-0.044|<br>-0.010|<br>-0.269<sup>**</sup>|\\n||(0.038)|(0.050)|(0.028)|(0.068)|(0.027)|(0.098)|(0.037)|(0.039)|(0.024)|(0.037)|(0.016)|(0.106)|\\n|Year 1973|-0.096<sup>*</sup>|-0.092<sup>*</sup>|-0.044|-0.087|-0.022|-0.235|-0.085<sup>*</sup>|-0.092<sup>**</sup>|-0.054|-0.060|-0.029|-0.254<sup>*</sup>|\\n||(0.040)|(0.041)|(0.033)|(0.056)|(0.032)|(0.124)|(0.040)|(0.040)|(0.040)|(0.042)|(0.017)|(0.135)|\\n|Year 1974|<br>-0.018|<br>0.000|<br>-0.007|<br>0.027|<br>0.015<sup>*</sup>|<br>0.043|<br>-0.038|<br>-0.053|<br>-0.060<sup>*</sup>|<br>-0.034|<br>0.000|<br>-0.163|\\n||(0.045)|(0.054)|(0.035)|(0.052)|(0.006)|(0.137)|(0.050)|(0.053)|(0.033)|(0.050)|(0.028)|(0.158)|\\n|Year 1975|<br>-0.053<sup>***</sup>|<br>-0.061<sup>*</sup>|<br>-0.044|<br>-0.060<sup>**</sup>|<br>0.009|<br>-0.202|<br>-0.027|<br>-0.050|<br>-0.034|<br>-0.001|<br>-0.021|<br>-0.145|\\n||(0.013)|(0.024)|(0.076)|(0.019)|(0.030)|(0.191)|(0.045)|(0.043)|(0.035)|(0.040)|(0.017)|(0.134)|\\n|Year 1976|<br>-0.128<sup>***</sup>|<br>-0.092<sup>**</sup>|<br>-0.048|<br>-0.076|<br>0.011|<br>-0.286<sup>**</sup>|<br>-0.033|<br>-0.036|<br>-0.040|<br>-0.006|<br>-0.011|<br>-0.166|\\n||(0.024)|(0.027)|(0.028)|(0.047)|(0.030)|(0.086)|(0.049)|(0.052)|(0.040)|(0.040)|(0.022)|(0.128)|\\n|Year 1977|<br>-0.089<sup>***</sup>|<br>-0.090|<br>-0.021|<br>-0.006|<br>-0.001|<br>-0.352<sup>**</sup>|<br>-0.055|<br>-0.062|<br>-0.061<sup>**</sup>|<br>-0.051|<br>-0.029|<br>-0.355<sup>***</sup>|\\n||(0.022)|(0.054)|(0.022)|(0.057)|(0.037)|(0.124)|(0.044)|(0.052)|(0.025)|(0.032)|(0.022)|(0.114)|\\n|Year 1978|<br>-0.070<sup>**</sup>|<br>-0.028|<br>-0.006|<br>-0.037|<br>-0.021|<br>-0.289<sup>*</sup>|<br>-0.068|<br>-0.077|<br>-0.068<sup>**</sup>|<br>-0.046|<br>0.013|<br>-0.360<sup>**</sup>|\\n||(0.021)|(0.048)|(0.048)|(0.055)|(0.045)|(0.118)|(0.053)|(0.053)|(0.030)|(0.047)|(0.020)|(0.144)|\\n|Year 1979|<br>-0.068<sup>*</sup>|<br>-0.068|<br>0.002|<br>-0.074|<br>-0.036|<br>-0.459<sup>***</sup>|<br>-0.108<sup>**</sup>|<br>-0.107<sup>**</sup>|<br>-0.081<sup>**</sup>|<br>-0.067<sup>**</sup>|<br>-0.053<sup>*</sup>|<br>-0.437<sup>***</sup>|\\n||(0.030)|(0.039)|(0.027)|(0.056)|(0.032)|(0.095)|(0.039)|(0.040)|(0.031)|(0.029)|(0.025)|(0.129)|\\n|Year 1980|<br>-0.124<sup>**</sup>|<br>-0.135<sup>**</sup>|<br>-0.086<sup>**</sup>|<br>-0.103<sup>*</sup>|<br>0.010|<br>-0.464<sup>***</sup>|<br>-0.075<sup>*</sup>|<br>-0.076<sup>*</sup>|<br>-0.065<sup>*</sup>|<br>-0.041|<br>-0.050|<br>-0.350<sup>**</sup>|\\n\\n19\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"||(0.033)|(0.036)|(0.032)|(0.044)|(0.040)|(0.059)|(0.038)|(0.040)|(0.033)|(0.039)|(0.028)|(0.118)|\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n|Year 1981|-0.127<sup>**</sup>|-0.117<sup>**</sup>|-0.062|-0.072|-0.077|-0.595<sup>***</sup>|-0.128<sup>***</sup>|-0.135<sup>**</sup>|-0.060<sup>**</sup>|-0.084<sup>*</sup>|-0.027|-0.478<sup>***</sup>|\\n||(0.037)|(0.035)|(0.046)|(0.037)|(0.040)|(0.101)|(0.039)|(0.045)|(0.027)|(0.044)|(0.018)|(0.141)|\\n|_N_|3351|3351|3351|3351|3351|3169|7159|7159|7159|7159|7159|7043|\\n|adj._R_<sup>2</sup>|0.125|0.141|0.110|0.130|0.033|0.074|0.145|0.156|0.131|0.145|0.036|0.075|\\n\\nNotes: All regressions add birth governorate fixed-effects, household size and gender. Standard errors are clustered at the birth governorate level. *** p<0.01, ** p<0.05, * p<0.1.\\n\\n20\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"Table 11 shows that workers born in the governorates near the Islamic Republic of Iran’s border work in more hazardous conditions and have lower probability of accessing to air conditioning than those born in other governorates in Iraq. Also, we find that workers born in the governorates near the Islamic Republic of Iran’s border are more adversely affected at the working environment than those born in other governorates in Iraq.\\n\\n**Table 11. Disparities in working conditions of workers between border governorates and other Iraqi governorates**\\n\\n||Governorat|es near the Islam<br>Iran’s border|ic Republic of|Oth|er governorates|in Iraq|\\n|---|---|---|---|---|---|---|\\n||Hazardous<br>job|Air<br>conditioning|Health impact|Hazardous<br>job|Air<br>conditioning|Health impact|\\n||(1)|(2)|(3)|(4)|(5)|(6)|\\n|Year 1964|-0.012|-0.076|0.411|0.027|0.038|-0.007|\\n||(0.054)|(0.042)|(0.354)|(0.059)|(0.046)|(0.227)|\\n|Year 1965|<br>-0.015|<br>-0.052|<br>-0.061|<br>-0.019|<br>-0.014|<br>-0.113|\\n||(0.049)|(0.052)|(0.366)|(0.033)|(0.033)|(0.122)|\\n|Year 1966|<br>-0.012|<br>-0.009|<br>0.386|<br>0.010|<br>0.007|<br>0.051|\\n||(0.057)|(0.039)|(0.272)|(0.030)|(0.045)|(0.228)|\\n|Year 1967|<br>-0.007|<br>-0.092|<br>0.164|<br>0.019|<br>0.003|<br>0.030|\\n||(0.053)|(0.060)|(0.274)|(0.043)|(0.038)|(0.132)|\\n|Year 1969|<br>-0.007|<br>-0.106|<br>0.049|<br>-0.021|<br>-0.046|<br>0.093|\\n|(Year of|||||||\\n|change)|||||||\\n||(0.043)|(0.074)|(0.193)|(0.040)|(0.033)|(0.149)|\\n|Year 1970|<br>-0.133<sup>*</sup>|<br>-0.114<sup>*</sup>|<br>0.364<sup>*</sup>|<br>0.008|<br>-0.068|<br>0.127|\\n||(0.063)|(0.052)|(0.157)|(0.035)|(0.051)|(0.118)|\\n|Year 1971|<br>-0.052|<br>-0.045|<br>0.157|<br>-0.061|<br>-0.033|<br>0.031|\\n||(0.051)|(0.042)|(0.136)|(0.039)|(0.021)|(0.160)|\\n|Year 1972|<br>-0.076|<br>-0.040|<br>0.298|<br>-0.046|<br>-0.107<sup>**</sup>|<br>0.270|\\n||(0.049)|(0.056)|(0.160)|(0.036)|(0.036)|(0.159)|\\n|Year 1973|<br>-0.172<sup>**</sup>|<br>-0.094|<br>0.169|<br>-0.071<sup>*</sup>|<br>-0.071<sup>**</sup>|<br>0.062|\\n||(0.052)|(0.050)|(0.210)|(0.038)|(0.027)|(0.147)|\\n|Year 1974|<br>-0.164<sup>**</sup>|<br>-0.116|<br>0.468<sup>*</sup>|<br>-0.066<sup>*</sup>|<br>-0.025|<br>0.254|\\n||(0.045)|(0.062)|(0.206)|(0.035)|(0.030)|(0.162)|\\n|Year 1975|<br>-0.116|<br>-0.138<sup>**</sup>|<br>0.254|<br>-0.020|<br>-0.047<sup>*</sup>|<br>0.002|\\n||(0.058)|(0.044)|(0.221)|(0.037)|(0.026)|(0.119)|\\n|Year 1976|<br>-0.105|<br>-0.093|<br>0.534<sup>**</sup>|<br>-0.020|<br>-0.002|<br>-0.019|\\n||(0.089)|(0.051)|(0.200)|(0.037)|(0.026)|(0.125)|\\n|Year 1977|<br>-0.109|<br>-0.093|<br>0.512<sup>*</sup>|<br>-0.030|<br>-0.039|<br>0.138|\\n||(0.073)|(0.050)|(0.206)|(0.032)|(0.039)|(0.166)|\\n|Year 1978|<br>-0.088|<br>-0.120|<br>0.636<sup>*</sup>|<br>-0.007|<br>-0.056|<br>0.099|\\n||(0.051)|(0.066)|(0.277)|(0.021)|(0.038)|(0.143)|\\n|Year 1979|<br>-0.076|<br>-0.071|<br>0.672<sup>**</sup>|<br>-0.017|<br>-0.094<sup>***</sup>|<br>0.251<sup>*</sup>|\\n||(0.058)|(0.061)|(0.180)|(0.038)|(0.028)|(0.125)|\\n|Year 1980|<br>-0.115<sup>*</sup>|<br>-0.142<sup>**</sup>|<br>0.470<sup>*</sup>|<br>-0.041|<br>-0.038|<br>0.050|\\n||(0.048)|(0.040)|(0.206)|(0.025)|(0.028)|(0.142)|\\n|Year 1981|<br>-0.111<sup>***</sup>|<br>-0.143<sup>*</sup>|<br>0.628<sup>**</sup>|<br>-0.049|<br>-0.105<sup>**</sup>|<br>0.242<sup>**</sup>|\\n||(0.024)|(0.064)|(0.230)|(0.037)|(0.034)|(0.104)|\\n|_N_|<br>3351|<br>3351|<br>3351|<br>7159|<br>7159|<br>7159|\\n|adj._R_<sup>2</sup>|0.112|0.083|0.166|0.130|0.089|0.127|\\n\\n21\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"Notes: All regressions add birth governorate fixed-effects, household size and gender. Standard errors are clustered at the birth governorate level. *** p<0.01, ** p<0.05, * p<0.1.\\n\\n# **5. Conclusions**\\n\\nA large body of literature examines the long-lasting impacts of early exposure to the war on educational attainment and labor market outcomes in adulthood. Nonetheless, there is little known about the long-term impact on children, who were exposed to the war at the ages of 6 to 12 in Iraq. The Islamic Republic of Iran–Iraq war in 1980-1988 provides an excellent setting to study the longterm impact of war. Apart from all other—human, financial, and emotional—costs, the Islamic Republic of Iran–Iraq war led to a significant drop in educational attainment for both men and women who were of elementary school age during the conflict. The Islamic Republic of Iran–Iraq war had a large impact on men’s education level and women’s education level.\\n\\nIn addition to this significant educational loss, the Islamic Republic of Iran–Iraq war appears to have caused an adverse impact on social welfare of men, such as accessing social security, pension plan, health care, paid leave and job permanence. This impact is not statistically significant for women. We also find that men who were exposed to the war during elementary school age have higher probability of working in hazardous jobs and without air conditioning; however, we find no evidence for women. Similarly, we find evidence of the impact of war intensity on educational attainment and labor market outcomes. We conduct a variety of robustness checks and falsification tests, and our estimation results are still robust.\\n\\nWe believe that the most likely channel through which the Islamic Republic of Iran–Iraq war affects the labor market outcomes is the educational loss induced by the war. The educational loss of men is larger than the educational loss of women, and the Islamic Republic of Iran–Iraq war has more adverse impacts on labor market outcomes for men than for women.\\n\\n22\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099349102012434554/pdf/IDU1f87e964d10cb614d3219bdd1759b52b0527b.pdf\", \"pages\": [23]}}]}"}}},{"rowIdx":96,"cells":{"doc_id":{"kind":"string","value":"34249852"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34249852\", \"last_modified_date\": \"2024-03-08T00:00:00Z\", \"authors\": {\"0\": {\"author\": \"Goldstein,Markus P.\"}, \"1\": {\"author\": \"Gonzalez Martinez,Paula Lorena\"}, \"2\": {\"author\": \"Kilic,Talip\"}, \"3\": {\"author\": \"Papineni,Sreelakshmi\"}, \"4\": {\"author\": \"Wollburg,Philip Randolph\"}}, \"count\": \"World\", \"count_key\": \"517191\", \"docna\": {\"0\": {\"docna\": \"Breadwinners and Caregivers : Examining the\\n Global Relationship between Gender Norms and Economic Behavior\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"Gender Impact Evaluation (AFEGI),Development Data Group (DECDG)\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34249852\"}, \"repnb\": \"WPS10692\", \"docdt\": \"2024-02-01T00:00:00Z\", \"datestored\": \"2024-02-01T06:21:48Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"Gender norms are often emphasized to\\n help explain gender gaps in the labor market. This paper\\n examines global patterns of gender attitudes and norms\\n toward the stereotypical gender roles of the male\\n breadwinner and female caregiver, and broad support for\\n gender equality in opportunities, and studies their\\n relationship with economic behavior. Using data collected\\n via Facebook from 150,000 individuals across 111 countries\\n the paper explores how gender beliefs and norms are related\\n to labor supply, household production, and intrahousehold\\n decision-making power within a country. The paper provides\\n descriptive evidence that the more gender equitable or\\n counter-stereotypical are beliefs and norms, the more likely\\n women are to work, the more time men spend on household\\n chores, and the higher the likelihood of joint\\n decision-making among couples. The findings suggest an\\n underestimation of the support for gender equality globally\\n and the extent of underestimation varies by gender and\\n region. The paper concludes with a discussion of potential\\n entry points for policy to help address gender norms.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10692; MNACE labor and gender research programs; PEOPLE; RRR\", \"display_title\": \"Breadwinners and Caregivers :\\n Examining the Global Relationship between Gender Norms and\\n Economic Behavior\", \"disclosure_date\": \"2024-02-01T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099821402012437374/text/IDU1bf06cd191699f1462f18236168ed5e5295bf.txt\", \"listing_relative_url\": \"/research/2024/02/34249852/breadwinners-caregivers-examining-global-relationship-between-gender-norms-economic-behavior\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099821402012437374/IDU1bf06cd191699f1462f18236168ed5e5295bf\", \"new_url\": \"2024/02/34249852/IDU1bf06cd191699f1462f18236168ed5e5295bf\", \"ext_pub_date\": \"2024-02-01T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Labor Markets m1326488 855,Gender m1327621 615\", \"docm_id\": \"IDU-bf06cd19-699f-462f-8236-68ed5e5295bf\", \"chronical_docm_id\": \"IDU-bf06cd19-699f-462f-8236-68ed5e5295bf\", \"publishtoextweb_dt\": \"2024-02-01T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10692\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"guid\": \"099821402012437374\", \"no_of_pages\": \"72\", \"author_keys\": {\"0\": {\"author_key\": \"769522\"}, \"1\": {\"author_key\": \"1250774\"}, \"2\": {\"author_key\": \"886702\"}, \"3\": {\"author_key\": \"1343136\"}, \"4\": {\"author_key\": \"1464872\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099821402012437374/IDU1bf06cd191699f1462f18236168ed5e5295bf\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10692\\n\\n# Breadwinners and Caregivers\\n\\nExamining the Global Relationship between Gender Norms and Economic Behavior _Markus Goldstein Paula Gonzalez Talip Kilic Sreelakshmi Papineni Philip Wollburg_ Gender Innovation Lab & Development Data Group February 2024 A verified reproducibility package for this paper is available at http://reproducibility.worldbank.org, click **here** for direct access.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10692\\n\\n# **Abstract**\\n\\nGender norms are often emphasized to help explain gender gaps in the labor market. This paper examines global patterns of gender attitudes and norms toward the stereotypical gender roles of the _male breadwinner_ and _female caregiver_ , and broad support for _gender equality in opportunities_ , and studies their relationship with economic behavior. Using data collected via Facebook from 150,000 individuals across 111 countries the paper explores how gender beliefs and norms are related to labor supply, household production, and intrahousehold decision-making power within a country. The paper provides descriptive evidence that the more gender equitable or counter-stereotypical are beliefs and norms, the more likely women are to work, the more time men spend on household chores, and the higher the likelihood of joint decision-making among couples. The findings suggest an underestimation of the support for gender equality globally and the extent of underestimation varies by gender and region. The paper concludes with a discussion of potential entry points for policy to help address gender norms.\\n\\nThis paper is a product of the Gender Innovation Lab, Africa Region and the Development Data Group, Development Economics. I. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http:// www.worldbank.org/prwp. The authors may be contacted at spapineni@worldbank.org and tkilic@worldbank.org. A verified reproducibility package for this paper is available at http://reproducibility.worldbank.org, click **here** for direct access.\\n\\n<!-- Start of picture text --> TRANSPARENT<br>ANALYSIS<br>G<br>R E<br>N<br>S<br>I P<br>Y E<br>K A<br>C A<br>R P<br>LI R<br>O E<br>O C<br>W R<br>H<br>S<br>P<br><!-- End of picture text --> _The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [{\"mention_name\": {\"text\": \"data collected via Facebook\", \"confidence\": 0.663912832736969, \"start\": 411, \"end\": 438, \"clean_text\": \"data collected via Facebook\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 411, \"end\": 438, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 411, \"end\": 438, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 411, \"end\": 438}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 411, \"end\": 438, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8302798867225647, \"start\": 411, \"end\": 438}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.663912832736969, \"start\": 411, \"end\": 438}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.826272189617157, \"start\": 411, \"end\": 438}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.826272189617157, \"start\": 411, \"end\": 438}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"Breadwinners and Caregivers: Examining the Global Relationship between Gender Norms and Economic Behavior<sup>∗</sup> Markus Goldstein<sup>†</sup> Paula Gonzalez<sup>‡</sup> Talip Kilic<sup>§</sup> Sreelakshmi Papineni<sup>¶</sup> Philip Wollburg<sup>‖</sup> Gender, Attitudes, Norms, Labor, Global [JEL] J16, J20, O12\\n\\n> ∗We thank Rachael Pierotti, S Anukriti, Ha Nguyen and Lili Mottaghi for their helpful comments and guidance. This paper uses data from the Gender Equality at Home surveys - a collaborative effort between Facebook Data For Good, the World Bank and other development partners. We thank Samhir Vasdev, Safaa Amer, Bronwen Raff and Anna Lerner at Facebook Data For Good. We gratefully acknowledge the financial support from the World Bank Group’s Umbrella Facility for Gender Equality (UFGE) and the Office of the Chief Economist for the Middle East and North Africa (MNACE) under the regional Labor and Gender Research Programs. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.\\n\\n> †World Bank (on leave) `mgoldstein@worldbank.org`\\n\\n> ‡World Bank and The Graduate Institute, Geneva `pgonzalezmartine@worldbank.org`\\n\\n> §World Bank Development Economics `tkilic@worldbank.org`\\n\\n- ¶World Bank Africa Gender Innovation Lab `spapineni@worldbank.org (Corresponding Author)`\\n\\n- ‖World Bank Development Economics `pwollburg@worldbank.org`\", \"datasets\": [{\"mention_name\": {\"text\": \"Gender Equality at Home surveys\", \"confidence\": 0.922354519367218, \"start\": 461, \"end\": 492, \"clean_text\": \"Gender Equality at Home surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 461, \"end\": 492, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 461, \"end\": 492, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 461, \"end\": 492}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 461, \"end\": 492, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9955286383628845, \"start\": 461, \"end\": 492}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.922354519367218, \"start\": 461, \"end\": 492}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9867857098579407, \"start\": 461, \"end\": 492}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9867857098579407, \"start\": 461, \"end\": 492}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **1 Introduction**\\n\\nGender norms are often emphasized as key to help explain remaining gender gaps in the labor market (Bertrand, 2020; Goldin, 2021).<sup>1</sup> For example, the stereotypical gender roles that men should be the main income earners or “breadwinners” and women should primarily be responsible for the care of children and household chores or “caregivers” in the household are widely acknowledged normative constraints to women’s economic empowerment (Jayachandran, 2021). Despite the recognition of the importance of social norms for gender equality among economists and policy makers, very little development-funding is currently devoted to programs that directly influence norms-related change.<sup>2</sup> In this paper we collect and analyze data from 111 countries to characterize how gender beliefs and norms relate to economic behavior across the world. We use the findings to help highlight some potential entry points for policy to address gender norms.\\n\\nWomen’s economic decisions may be constrained by a range of social norms that influence the types of roles and responsibilities that are acceptable for men and women and uphold widely shared conceptions of masculinity and femininity (Bicchieri, 2016; Marcus and Harper, 2014). Akerlof and Kranton (2000, 2002, 2010) translated theories of social identity into an economics framework and propose that social identity influences economic outcomes because deviating from the prescribed behavior is inherently costly. Norms (often subconsciously) encourage behaviors that are socially valued and discourage behaviors that elicit social sanctions and stigma (Bernhardt et al., 2018). Gender norms are likely to significantly constrain women’s choices about whether and which types of work to pursue, and equally may prevent men from engaging in care and domestic chore activities.\\n\\nOur interest in this paper is to contribute to a growing literature that examines societal norms as a barrier to female labor market outcomes (e.g. Bernhardt et al. (2018); Bursztyn and Yang (2022); Giuliano (2021); Jayachandran (2021)). While there is general acceptance that gender norms play a role in explaining gender differences in labor market outcomes, there is currently limited empirical evidence on the relationship between norms and economic outcomes (Field et al., 2021). Existing research has been concentrated in specific regions and countries, and persistent gaps in obtaining gender-disaggregated data have prevented research at scale. Studies also have relied on relatively small sample sizes and the majority of research has focused on how aggregate, country-level attitudes relate to rates of female labor supply. More recently studies have emphasized the importance of complement-\\n\\n> 1 Gender norms, a subset of social norms, are defined as shared expectations about how women and men should behave in a particular social group or culture.\\n\\n> 2For example, at the World Bank social norms are not an operational focus and are often simply considered the “enabling environment” under which capital- or skills-related programs operate.\\n\\n2\", \"datasets\": [{\"mention_name\": {\"text\": \"data from 111 countries\", \"confidence\": 0.846656322479248, \"start\": 765, \"end\": 788, \"clean_text\": \"data from 111 countries\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 765, \"end\": 788, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 765, \"end\": 788, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 765, \"end\": 788}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 765, \"end\": 788, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.4070735573768616, \"start\": 765, \"end\": 788}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.846656322479248, \"start\": 765, \"end\": 788}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6712522506713867, \"start\": 765, \"end\": 788}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6712522506713867, \"start\": 765, \"end\": 788}}, {\"mention_name\": {\"text\": \"gender-disaggregated data\", \"confidence\": 0.8486813306808472, \"start\": 2453, \"end\": 2478, \"clean_text\": \"gender-disaggregated data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2453, \"end\": 2478, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2453, \"end\": 2478, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2453, \"end\": 2478}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2453, \"end\": 2478, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.2398366779088974, \"start\": 784, \"end\": 809}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8486813306808472, \"start\": 784, \"end\": 809}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.45953255891799927, \"start\": 784, \"end\": 809}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.45953255891799927, \"start\": 784, \"end\": 809}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"ing attitudes with measures of perceived norms (i.e. what individuals think others might approve or disapprove of) since perceptions are likely to matter for decision-making (see for example, Bernhardt et al., 2018; Bicchieri, 2016; Bursztyn and Yang, 2022).<sup>3</sup> An influential study by Bursztyn, Gonzalez and Yanagizawa-Drott (2020) documented that perceptions of peers’ opinions toward female employment outside the home influenced behavior beyond an individual’s own opinions.\\n\\nIn this paper, we fill data and knowledge gaps by leveraging a unique dataset that includes data both on personal attitudes and perceived norms and link them to individual-level employment outcomes on a global scale. We use data from over 150,000 individuals across 111 countries from all regions of the world collected through a survey implemented online via the Facebook platform, namely the Survey on Gender Equality at Home.<sup>4</sup> The survey has a greater coverage of low- and middle-income countries than alternative global databases that measure gender attitudes.<sup>5</sup> We designed questions for the survey to capture an individual’s personal belief and perception about others’ beliefs around traditional gender roles to study how gender norms relate to specific behaviors for women and men. We complement two levels of beliefs - what individuals think and what they think their community thinks - and correlate them with labor market, time use and decision-making behaviors. Specifically, this paper asks how do personal attitudes and perceived community norms towards a _broad norm on rights to equal opportunities_ , and the stereotypical gender roles of the _male breadwinner_ and _female caregiver_ correlate with labor force participation, intrahousehold decision-making power, and time spent on paid work, care and domestic activities for both men and women globally?\\n\\nThe stereotypical roles of the male breadwinner and female caregiver are hypothesized to have been born between the mid-19th and mid-20th century in many countries as industrialization gave rise to the separation of home and work. In this model of the family men are responsible for economic provision through employment while women are responsible for home and family. However, Alesina, Giuliano and Nunn (2013) examine the historical origins of gender norms based on a theory proposed by Boserup (1970), and show that gender attitudes are more unequal among descendants of societies that practiced plow agriculture. Since plow agriculture was much more capital-intensive and required greater upper-body strength, this led to a greater gender-based division of labor. The authors find historical\\n\\n> 3Beliefs about what others do are referred to as _descriptive_ norms and beliefs about what others approve of are referred to as _injunctive_ norms in social norms theory (Cialdini, Kallgren and Reno, 1991; Heinicke, Konig-Kersting and Schmidt, 2022). Here we are measuring the injunctive norm.\\n\\n> 4The survey was administered on Facebook in July 2020 to a sample of Facebook general population users. 5 See World Values Survey (WVS) for global data on gender beliefs and values and Bursztyn et al. (2023 _a_ ) for analysis of data on the rights of women to work outside the home and support for women in leadership positions from 60 countries in a 2020 Gallup World Poll survey.\\n\\n3\", \"datasets\": [{\"mention_name\": {\"text\": \"data from over 150,000 individuals\", \"confidence\": 0.48421162366867065, \"start\": 713, \"end\": 747, \"clean_text\": \"data from over 150,000 individuals\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 713, \"end\": 747, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 713, \"end\": 747, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 713, \"end\": 747}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 713, \"end\": 747, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.99798583984375, \"start\": 713, \"end\": 747}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.48421162366867065, \"start\": 713, \"end\": 747}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.973773181438446, \"start\": 713, \"end\": 747}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.973773181438446, \"start\": 713, \"end\": 747}}, {\"mention_name\": {\"text\": \"Survey on Gender Equality at Home\", \"confidence\": 0.8731344938278198, \"start\": 883, \"end\": 916, \"clean_text\": \"Survey on Gender Equality at Home\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 883, \"end\": 916, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 883, \"end\": 916, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 883, \"end\": 916}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 883, \"end\": 916, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.99798583984375, \"start\": 883, \"end\": 916}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8731344938278198, \"start\": 883, \"end\": 916}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.973773181438446, \"start\": 883, \"end\": 916}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.973773181438446, \"start\": 883, \"end\": 916}}, {\"mention_name\": {\"text\": \"global databases\", \"confidence\": 0.39821645617485046, \"start\": 1017, \"end\": 1033, \"clean_text\": \"global databases\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1017, \"end\": 1033, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1017, \"end\": 1033, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1017, \"end\": 1033}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1017, \"end\": 1033, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9984394907951355, \"start\": 224, \"end\": 240}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.39821645617485046, \"start\": 224, \"end\": 240}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7772397398948669, \"start\": 224, \"end\": 240}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7772397398948669, \"start\": 224, \"end\": 240}}, {\"mention_name\": {\"text\": \"World Values Survey\", \"confidence\": 0.7599176168441772, \"start\": 3092, \"end\": 3111, \"clean_text\": \"World Values Survey\"}, \"acronym\": {\"text\": \"WVS\", \"confidence\": 0.5118392109870911, \"start\": 3113, \"end\": 3116, \"clean_text\": \"WVS\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3092, \"end\": 3111, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3092, \"end\": 3111}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3092, \"end\": 3111, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9994545578956604, \"start\": 558, \"end\": 577}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7599176168441772, \"start\": 558, \"end\": 577}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9496158361434937, \"start\": 558, \"end\": 577}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9496158361434937, \"start\": 558, \"end\": 577}}, {\"mention_name\": {\"text\": \"Gallup World Poll survey\", \"confidence\": 0.7614596486091614, \"start\": 3338, \"end\": 3362, \"clean_text\": \"Gallup World Poll survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3338, \"end\": 3362, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3338, \"end\": 3362, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3338, \"end\": 3362}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3338, \"end\": 3362, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.996772825717926, \"start\": 804, \"end\": 828}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7614596486091614, \"start\": 804, \"end\": 828}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.834029495716095, \"start\": 804, \"end\": 828}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.834029495716095, \"start\": 804, \"end\": 828}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"plow use to be negatively associated with current attitudes towards gender equality, and female participation in labor, politics, and entrepreneurial activities. Therefore, the work of Alesina, Giuliano and Nunn (2013) suggests that norms around gender roles may have been established even earlier in the pre-industrial period.\\n\\nRestrictive or conservative gender norms may translate into behaviors that disadvantage female labor supply and earnings, and/or discourage male engagement in childcare and household chores. Norms are expected to influence behavior through expectations of what others in your reference group do or approve of (Gauri, Rahman and Sen, 2019). In addition, when measuring gender beliefs, there may be expectations that men and women should play different roles in society (gender role beliefs) and that men and women are essentially different (gender category beliefs). The questions included in the Survey on Gender Equality at Home mainly capture measures of gender role beliefs rather than measures of beliefs in gender essentialism. Both genders may face norms costs imposed by those community members who disapprove of certain behaviors that contradict the expectations of the roles of men and women in their society (Bernhardt et al., 2018). Women or men who defy stereotypical norms potentially do so at a personal cost. For example, Bertrand, Kamenica and Pan (2015) show in the US that it is relatively unlikely that a woman will earn more than her husband, and when she does, marital satisfaction is lower and divorce is more likely. Friedson-Ridenour and Pierotti (2019) find that some women in Ghana explicitly limit their business growth in order to reinforce their husband’s role as a primary provider. In addition, for many men, engaging in care and housework is inconsistent with male gender roles and indicates weakness; and when men feel threatened in their role as providers, they may be even less inclined to engage in behavior associated with female gender roles (Munoz-Boudet et al., 2013).<sup>6</sup> A first contribution of the paper is to provide global evidence on attitudes and norms concerning gender stereotypical roles reported by both men and women. In the paper we begin by conducting cross-country comparisons of personal attitudes and perceived community norms with respect to a broad norm on gender equality in opportunities, and the male breadwinner norm and female caregiver norm. Descriptive findings suggest that gender attitudes and norms vary widely across countries and within countries. For instance, the proportion of respondents that agree with the male breadwinner norm statement ranges from 11% (in Denmark) to 78% (in the Arab Republic of Egypt).<sup>7</sup> This compares to a country average in the perceived community norm of 28% (in Denmark) and 69% (in Egypt).<sup>8</sup>\\n\\n> 6For instance, Bernhardt et al. (2018) find that the majority (70%) of married men in rural India perceive themselves to be the main recipient of social punishment if their wife was working for pay; whereas married women think the social cost is more evenly shared between spouses.\\n\\n> 7A lower average percentage represents more gender liberal beliefs in that country.\\n\\n> 8This translates to Egyptian men and women in the sample perceiving that 69% of their community would agree with the male breadwinner norm.\\n\\n4\", \"datasets\": [{\"mention_name\": {\"text\": \"Survey on Gender Equality at Home\", \"confidence\": 0.793177604675293, \"start\": 925, \"end\": 958, \"clean_text\": \"Survey on Gender Equality at Home\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 925, \"end\": 958, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 925, \"end\": 958, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 925, \"end\": 958}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 925, \"end\": 958, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9953889846801758, \"start\": 925, \"end\": 958}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.793177604675293, \"start\": 925, \"end\": 958}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9949504137039185, \"start\": 925, \"end\": 958}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9949504137039185, \"start\": 925, \"end\": 958}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"Next, we explore whether there is a general misperception of norms within the sampled countries. Misperceptions are defined by comparing perceived norms with aggregate attitudes of the sample that implies a degree of _pluralistic ignorance_ as described in Bursztyn and Yang (2022).<sup>9</sup> We show that misperceptions of gender norms are widespread around the world but the extent of misperception varies by region-of-the-world and gender. Descriptive findings suggest that globally there is an underestimation of the support for gender equality. We explore demographic characteristics that are predictive of an individual over- or underestimating community norms. Women are more likely to underestimate support for gender equality than men. More highly educated, younger and urban men are less likely to underestimate support for gender equality. Conversely, more highly educated and younger women are more likely to underestimate the support for gender equality in their community.\\n\\nThe main contribution of our paper is to provide novel evidence on the link between gender attitudes and norms to economic behaviors measured across the world. We use multivariate regression analysis to explore how beliefs and norms correlate with observed economic variables at the individual level. Overall, we find that women’s own personal belief in a norm of equal opportunities and their perception of general community support for gender equality are independently correlated with whether women work. Individual beliefs do not have much of a relationship with male labor force participation; whereas for women, the more gender progressive her own beliefs on gender equality, the more likely she is to be working. Community norms show a similar story to individual beliefs for women’s labor force participation: the more liberal she thinks her community is, the more she works. For the more proximal norms (caregiver, breadwinner) her individual beliefs seem to matter more than her perceptions of the community norm. That is, women’s perception of community support for the male breadwinner and female caregiver norms is not independently correlated with whether and how much they work, net of their own personal beliefs. In terms of intrahousehold bargaining power, we find that decisions are more likely to be made jointly when there is greater support for gender equality. Greater support for the female caregiver norm in the community is strongly associated with greater involvement in household activities by both men and women. Even after taking into account men’s own beliefs, the perceived beliefs of those in the community matter for male engagement in household chores. We use these findings to highlight potential entry points for policy.\\n\\nThe remainder of this paper is organized as follows. Section 2 describes the data, sample and variables, while Section 3 details the empirical strategy used for the analysis. Section 4 presents the results; and section 5 concludes where we discuss policy implications.\\n\\n> 9Pluralistic ignorance is the term used by psychologists to describe when people are inaccurate when estimating the prevailing attitude in their community.\\n\\n5\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"# **2 Data and Sample**\\n\\n## **2.1 The Gender Equality at Home Survey**\\n\\nOur analysis uses the individual level data from the Gender Equality at Home survey that was administered on the Facebook platform in July 2020.<sup>10</sup> The Gender Equality at Home survey is a collaboration between Facebook, the World Bank and other development partners to survey individuals on Facebook on issues related to gender equality and women’s empowerment. The partners developed a short survey questionnaire to collect data from Facebook’s general population of users. The questionnaire was designed to measure employment, beliefs and norms on gender, plus a number of key demographic questions (for example, gender, age, and marital status of the respondent) as well as time spent on work and domestic and care responsibilities, decision making and resource allocation across household members.\\n\\nSince the survey was administered in July 2020, additional questions on the impact of the coronavirus (COVID-19) pandemic were also included. Since the survey was still early in the coronavirus pandemic it is less of a concern that survey responses were directly affected by the pandemic.<sup>11</sup> Certain survey modules were randomized across participants in order to mitigate survey fatigue among respondents such that no more than 30 questions were asked in total.\\n\\nThe survey was administered on the Facebook platform across 208 countries, territories, and islands. The sampling frame for this survey is the individual database of Facebook and it was administered to 461,748 respondents sampled across the globe from the target population. While 208 economies were surveyed, the sample considered in this paper is the 111 countries with sufficient sample size to conduct gender-disaggregated analysis as described further in section 2.2. The dataset is designed to reflect the Facebook user base rather than any specific national population, focusing on countries where internet access is widely available.<sup>12</sup>\\n\\n## **2.2 Sample**\\n\\nIn this paper, we report responses from the primary individual targeted by Facebook from a sample of general Facebook users. The analysis in this paper restricts the sample to the 111 countries with sufficient sample size to conduct gender-disaggregated analysis. Sample\\n\\n> 10See https://dataforgood.fb.com/tools/gendersurvey/. The survey was collected over 3 weeks in July 2020.\\n\\n> 11The hours worked variable that has a short recall period of “in the past 7 days” might be more affected by the pandemic.\\n\\n> 12 To account for the sampling design, non-response, and online presence, advanced statistical weights were used. These weights reflect the number of persons each person in the sample represents and help calibrate the results to the target population which is in this case is the online population within each country.\\n\\n6\", \"datasets\": [{\"mention_name\": {\"text\": \"Gender Equality at Home Survey\", \"confidence\": 0.961403489112854, \"start\": 38, \"end\": 68, \"clean_text\": \"Gender Equality at Home Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 38, \"end\": 68, \"clean_text\": \"\"}, \"producer\": {\"text\": \"Facebook\", \"confidence\": 0.3323855698108673, \"start\": 292, \"end\": 300, \"clean_text\": \"Facebook\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 38, \"end\": 68}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 38, \"end\": 68, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9991129040718079, \"start\": 38, \"end\": 68}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.961403489112854, \"start\": 38, \"end\": 68}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.767959475517273, \"start\": 38, \"end\": 68}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.767959475517273, \"start\": 38, \"end\": 68}}, {\"mention_name\": {\"text\": \"individual database of Facebook\", \"confidence\": 0.7851943373680115, \"start\": 1501, \"end\": 1532, \"clean_text\": \"individual database of Facebook\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1501, \"end\": 1532, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1501, \"end\": 1532, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1501, \"end\": 1532}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1501, \"end\": 1532, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8953220248222351, \"start\": 745, \"end\": 776}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.7851943373680115, \"start\": 745, \"end\": 776}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8644018173217773, \"start\": 745, \"end\": 776}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8644018173217773, \"start\": 745, \"end\": 776}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"sizes by country, region, and gender are shown in Table A1.<sup>13</sup> The paper’s dataset comprises 157,483 observations, encompassing all respondents who provided complete information on individual characteristics. However, the sample size varies when analyzing different outcomes, as detailed in A.7. Questions posed later in the survey experienced higher rates of nonresponse.\\n\\nThe sample for the Gender Equality at Home survey is comprised of individuals who have access to the internet and personal Facebook accounts. As such, the results should not be viewed as representative of a general population in each country. Nevertheless, the survey’s extensive geographic reach and substantial sample size provide a unique opportunity to explore the research questions on a global scale. Given the significant increase in internet bandwidth and Facebook’s expanding user base, which stands at approximately 2.9 billion monthly active users at the time of this writing, this survey offers valuable insights despite its limitations.<sup>14</sup> Table 1 outlines the demographic characteristics of our sample, split by gender of the respondent. The total sample comprises of 152,555 individuals, with 52% of the sample being women. The demographic characteristics are broadly similar across female and male respondents. On average, 58% of the individuals in our sample are in a long-term partnership or married. The majority, 73%, are over 25 years old. Education levels are high, with 60% possessing an education beyond secondary level. Additionally, 60% of respondents reside in urban areas. Notable gender-based differences in the sample include a 19% lower likelihood of women identifying as the head of the household compared to men, and a 7% lower likelihood of women owning land.\\n\\n# **2.3 Description of Key Variables**\\n\\n## **2.3.1 Economic Behaviors**\\n\\nThe dependent variables in our study are categorized as follows: labor supply (employment status in the previous year and weekly hours worked), intrahousehold decision-making power (categorized as either solely female, solely male, jointly made, or none), and household production (identification as the primary caregiver for children, and the allocation of time to caregiving and domestic responsibilities). Details about these variables, including their con-\\n\\n> 13The sample size calculation achieved a 95% confidence level for estimating regional as well as country level proportions with an average 3% error rate and an 80% power to detect differences across regions and by gender for the online population. Finite population correction was used for countries with smaller online populations. The targeted sample sizes per country ranged from 600 respondents to 5,000 respondents.\\n\\n> 14Total number of individuals on Facebook are estimated to be equal to the number of monthly active users worldwide as of 4th quarter 2018 that were using Facebook Source: https://www.statista.com/statistics/264810/number-of-monthly-active-facebook-users-worldwide/.\\n\\n7\", \"datasets\": [{\"mention_name\": {\"text\": \"Gender Equality at Home survey\", \"confidence\": 0.7943848371505737, \"start\": 403, \"end\": 433, \"clean_text\": \"Gender Equality at Home survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 403, \"end\": 433, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 403, \"end\": 433, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 403, \"end\": 433}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 403, \"end\": 433, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9957203269004822, \"start\": 403, \"end\": 433}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7943848371505737, \"start\": 403, \"end\": 433}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6785478591918945, \"start\": 403, \"end\": 433}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6785478591918945, \"start\": 403, \"end\": 433}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"struction, are elaborated in Appendix A, Table A2. Additionally, Table A3 in the appendix provides a gender-based breakdown of the descriptive characteristics of the respondents.\\n\\n# **2.3.2 Gender Attitudes and Norms**\\n\\nThe survey collected data on individual personal beliefs and perceived community norms towards gender equality in opportunities, and the stereotypical roles of men and women. Each norm theme is asked as a set of personal beliefs (using a Likert scale) and perceived community norms (out of 10 of your neighbors how many do you think agree...). Below we define how the beliefs and norms are constructed in more detail:\\n\\n1. **Broad Norm on Gender Equality** The survey elicited individual-level personal beliefs on gender equality by asking how much the respondent agrees with the following statement: “men and women should have equal opportunities” using a 5-point Likert scale from strongly disagree to strongly agree. The personal belief variable is coded as a dummy variable equal to 1 if the respondent agrees or strongly agrees that men and women should have equal opportunities; and 0 otherwise. Second, the survey measures perceptions of community norms, where respondents are asked to indicate out of ten neighbors in their community, how many they think would agree that men and women should have equal opportunities.<sup>15</sup> These norm constructs give us a measure of individuals’ perceptions of what others around them think, i.e. the share of the community the respondent believes would agree with equal opportunities between the sexes. In the regression analysis the perceived community norm variable is rescaled to be a value between 0 and 1 in increments of 0.1 where 0 indicates the most conservative views towards gender equality in their community and 1 indicates the most liberal views towards gender equality perceived in their community.\\n\\n2. **Male Role as a Breadwinner** The individual personal belief for the male breadwinner norm asks the respondent how much they agree with the following statement: “household expenses are the responsibility of the man, even if his wife can help him”, again using the 5-point Likert scale. The personal belief male breadwinner variable is coded as a dummy variable equal to 1 if the respondent agrees or strongly agrees that household expenses are the responsibility of the man, even if his wife can help him; and 0 otherwise. For the perceived community male breadwinner norm, respondents are asked to indicate out of ten neighbors in their community, how many they think would agree that household expenses are the responsibility of the man, even if his wife can help\\n\\n- 15To simplify the question for respondents, the survey asked them to report a number out of 10. We then\\n\\n- convert this number to a proportion i.e. out of 100%.\\n\\n8\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"him. In the regression analysis the perceived community norm variable is rescaled to be a value between 0 and 1 in increments of 0.1 where 0 indicates the most progressive or counter-stereotypical views and 1 indicates the most conservative or stereotypical views towards the male breadwinner norm.\\n\\n3. **Female Role as a Caregiver** The individual personal belief for the female caregiver norm asks the respondent how much they agree with the following statement: “a woman’s most important role is to take care of her home and children” again using the 5-point Likert scale. The personal belief variable is coded as a dummy variable equal to 1 if the respondent agrees or strongly agrees that a woman’s most important role is to take care of her home and children and 0 otherwise. For the community norm, respondents are asked to indicate out of ten neighbors in their community, how many they think would agree that a woman’s most important role is to take care of her home and children. In the analysis the community norm variable is rescaled to be a value between 0 and 1 in 0.1 increments where 0 indicates the most progressive or counter-stereotypical views and 1 indicates the most conservative or stereotypical views towards the female caregiver norm.\\n\\n# **2.3.3 Misperceived Norms**\\n\\nIn the paper we analyze misperceptions of norms, i.e. we quantify the gap between actual beliefs within our study sample (aggregate of individual beliefs in a country) and the perceptions of others’ beliefs (community perceived norms). We present country-level aggregates of beliefs and individual responses as defined in Bursztyn and Yang (2022). A greater understanding of misperceived norms helps to identify scope for correcting pluralistic ignorance.\\n\\nTo evaluate misperceptions at the country level, we:\\n\\n1. Calculate the average perceived community belief as reported by individuals within a country (expressed as “out of 10 of your neighbors” by respondents) and convert this to a percentage scale (0-100%).\\n\\n2. Ascertain the actual percentage of participants (both men and women) in the sample who affirm agreement with the norm at the country level.\\n\\n3. Determine the disparity between the average perceived community belief (1) and the actual agreement percentage (2), without gender specification due to the community norm’s non-gendered reference: Community Belief minus Aggregate Individual Belief.\\n\\nThe patterns of misperceptions by region-of-the-world and country-level are described in further detail in the results section 4.1.3. In addition, following the method described in 9\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"Bursztyn and Yang (2022), we establish an individual level measure of misperception based on how individuals’ perceptions compare to actual beliefs. The “true value” is determined by the average belief among respondents in a country who agree with a given norm. We assess the perceived community norm against this true value to calculate the proportion of “correct beliefs” among respondents. At the individual level, we can discern the fraction of the population that accurately assesses, overestimates, or underestimates the prevalence of a norm. Consequently, we report the proportion of respondents with accurate perceptions (those within 0.5 standard deviations of the truth); those who overestimate (respondents who hold beliefs that are at least 0.5 standard deviation greater than the truth); and those who underestimate (respondents who hold beliefs that are at least 0.5 standard deviation less than the truth).\\n\\nWe define _OverestimateNormij_ as a binary indicator that equals 1 for individuals whose perceptions exceed the true average belief by at least 0.5 standard deviations. Similarly, _UnderestimatesNormij_ equals 1 for individuals whose perceptions fall at least 0.5 standard deviations below the true average belief. The baseline category, _AroundNormij_ , includes individuals whose perceptions are within 0.5 standard deviations of the actual beliefs.\\n\\nSince the reference group in our norm constructs is deliberately broad (i.e. neighbors), this may not directly correlate with individuals’ most influential social circles. The extent to which our behaviors are shaped by close personal connections versus the broader community remains an empirical question, hence our decision to keep a general reference group. Also since our norm constructs refer to “neighbors”, a group not synonymous with the Facebook user base itself. Therefore, while our analysis is useful for establishing broad patterns, we caution that any observed misperceptions might be attributable to sample selection rather than pluralistic ignorance.\\n\\n# **2.4 Comparison of the Gender Equality at Home Survey with Other Data Sources**\\n\\nThe Gender Equality at Home Survey represents a general population of Facebook users and should not be considered representative of a country’s population. For example, demographic characteristics of the respondents on Facebook might differ from the average characteristics of a representative individual in the same country; or they may have different gender attitudes.\\n\\nWhile our sample is not nationally representative, we examine how closely the gender attitude statements from our data correlate with comparable attitude statements collected in nationally representative global data sources: the World Values Survey (WVS)<sup>16</sup> and\\n\\n> 16For our analysis, we utilized data from Wave 7 of the World Value Survey, spanning from 2017 to 2021.\\n\\n10\", \"datasets\": [{\"mention_name\": {\"text\": \"Gender Equality at Home Survey\", \"confidence\": 0.8866772651672363, \"start\": 2132, \"end\": 2162, \"clean_text\": \"Gender Equality at Home Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2132, \"end\": 2162, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2132, \"end\": 2162, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2132, \"end\": 2162}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2132, \"end\": 2162, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9990286827087402, \"start\": 422, \"end\": 452}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8866772651672363, \"start\": 422, \"end\": 452}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.560336172580719, \"start\": 422, \"end\": 452}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.560336172580719, \"start\": 422, \"end\": 452}}, {\"mention_name\": {\"text\": \"World Values Survey\", \"confidence\": 0.7894322276115417, \"start\": 2729, \"end\": 2748, \"clean_text\": \"World Values Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2729, \"end\": 2748, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2729, \"end\": 2748, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2729, \"end\": 2748}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2729, \"end\": 2748, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9926403164863586, \"start\": 1019, \"end\": 1038}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7894322276115417, \"start\": 1019, \"end\": 1038}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5194879770278931, \"start\": 1019, \"end\": 1038}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5194879770278931, \"start\": 1019, \"end\": 1038}}, {\"mention_name\": {\"text\": \"World Value Survey\", \"confidence\": 0.48588207364082336, \"start\": 2831, \"end\": 2849, \"clean_text\": \"World Value Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2831, \"end\": 2849, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2831, \"end\": 2849, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2831, \"end\": 2849}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2831, \"end\": 2849, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9985215067863464, \"start\": 237, \"end\": 255}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.48588207364082336, \"start\": 237, \"end\": 255}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9884020686149597, \"start\": 237, \"end\": 255}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9884020686149597, \"start\": 237, \"end\": 255}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"Gallup World Poll 2020.<sup>17</sup> Overall, the correlation with our data is strong, assuring us that, at least in this vein, our sample is not unrepresentative. The WVS and Gallup World Poll have a lower coverage globally and include fewer countries in Sub-Saharan Africa than the Gender Equality at Home survey on Facebook.<sup>18</sup> The WVS also does not include measures of community norms. However, the Gallup World Poll 2020 data captures both attitudes and norms.\\n\\nWe consider attitudes that address the male breadwinner-female caregiver theme in the WVS and a broad norm using Gallup World Poll 2020 data. We compare the attitudes on the male breadwinner norm with the WVS statement: “If a woman earns more money than her husband, it’s almost certain to cause problems” and the female caregiver norm with the WVS statement: “When a mother works for pay, the children suffer”; and whether women should be allowed to work outside of the home from the Gallup World Poll. In the comparison, we use attitudinal responses disaggregated by gender, and compute averages at the country-level. We standardize the averages and assign a rank to each country within our sample and split into quintiles (i.e. from most progressive=1 to least progressive=5). We compare how the countries differ in rank across the two surveys. For those countries where the difference in rank is positive this indicates that the Facebook sample in that country is potentially more conservative or stereotypical than a respondent in the WVS. Conversely, if the rank is negative this indicates that the sample in the Facebook population is more gender progressive or counter-stereotypical than the more representative WVS. In Appendix A, tables A4 and A5, we show the country level differences in quintile rank. The results vary depending on the norm in question and gender. For example, in Spain (ESP) men are equally gender progressive in both the Gender Equality at Home survey and the WVS (difference in rank is 0); whereas Spanish women are equally progressive in the male breadwinner norm (difference in rank is 0) but are ranked more progressive in our survey measure than the WVS (difference in rank 1). While in most countries there is little difference in the rank across the two surveys for the male breadwinner norm; for the female caregiver norm there are some discrepancies, e.g. Tunisia and Uzbekistan have a difference of 4. Using the Gallup World Poll we present comparisons in Table A6 and show that in the majority of countries there is a less than 10 percentage point difference with our survey measure.\\n\\n> Specifically, we selected the dataset year that most closely aligned with our collection of the Gender at Home data in 2020. This data is referenced as Haerpfer et al. (2022)\\n\\n> 17Gallup world data was subtracted at the country level using the Appendix of Bursztyn et al. (2023 _b_ )\\n\\n> 18The overlap in country coverage is 68% between the WVS and the Gender Equality at Home survey; and approximately 50% with the Gallup World Poll 2020.\\n\\n11\", \"datasets\": [{\"mention_name\": {\"text\": \"Gallup World Poll 2020\", \"confidence\": 0.5896294713020325, \"start\": 0, \"end\": 22, \"clean_text\": \"Gallup World Poll 2020\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 0, \"end\": 22, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 0, \"end\": 22, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"2020\", \"confidence\": 0.34397321939468384, \"start\": 431, \"end\": 435}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 0, \"end\": 22, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9969497323036194, \"start\": 0, \"end\": 22}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5896294713020325, \"start\": 0, \"end\": 22}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7705980539321899, \"start\": 0, \"end\": 22}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7705980539321899, \"start\": 0, \"end\": 22}}, {\"mention_name\": {\"text\": \"Gender Equality at Home survey\", \"confidence\": 0.7998489141464233, \"start\": 284, \"end\": 314, \"clean_text\": \"Gender Equality at Home survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 284, \"end\": 314, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 284, \"end\": 314, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 284, \"end\": 314}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 284, \"end\": 314, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9991375207901001, \"start\": 284, \"end\": 314}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6648504137992859, \"start\": 284, \"end\": 314}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5757646560668945, \"start\": 284, \"end\": 314}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5757646560668945, \"start\": 284, \"end\": 314}}, {\"mention_name\": {\"text\": \"Gallup World Poll 2020 data\", \"confidence\": 0.6391947865486145, \"start\": 590, \"end\": 617, \"clean_text\": \"Gallup World Poll 2020 data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 590, \"end\": 617, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 590, \"end\": 617, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 590, \"end\": 617}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 590, \"end\": 617, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9696362018585205, \"start\": 590, \"end\": 617}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.6391947865486145, \"start\": 590, \"end\": 617}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9301276206970215, \"start\": 590, \"end\": 617}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9301276206970215, \"start\": 590, \"end\": 617}}, {\"mention_name\": {\"text\": \"Gallup World Poll\", \"confidence\": 0.8365246653556824, \"start\": 962, \"end\": 979, \"clean_text\": \"Gallup World Poll\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 962, \"end\": 979, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 962, \"end\": 979, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 962, \"end\": 979}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 962, \"end\": 979, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9717731475830078, \"start\": 962, \"end\": 979}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6023865938186646, \"start\": 962, \"end\": 979}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6811279058456421, \"start\": 962, \"end\": 979}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6811279058456421, \"start\": 962, \"end\": 979}}, {\"mention_name\": {\"text\": \"WVS\", \"confidence\": 0.6466302871704102, \"start\": 1968, \"end\": 1971, \"clean_text\": \"WVS\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1968, \"end\": 1971, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1968, \"end\": 1971, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1968, \"end\": 1971}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1968, \"end\": 1971, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9982684850692749, \"start\": 434, \"end\": 437}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6466302871704102, \"start\": 434, \"end\": 437}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8911818265914917, \"start\": 434, \"end\": 437}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8911818265914917, \"start\": 434, \"end\": 437}}, {\"mention_name\": {\"text\": \"Gender at Home data\", \"confidence\": 0.4036842882633209, \"start\": 2702, \"end\": 2721, \"clean_text\": \"Gender at Home data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2702, \"end\": 2721, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2702, \"end\": 2721, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2702, \"end\": 2721}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2702, \"end\": 2721, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.36753419041633606, \"start\": 379, \"end\": 398}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.4036842882633209, \"start\": 379, \"end\": 398}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5265955924987793, \"start\": 379, \"end\": 398}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5265955924987793, \"start\": 379, \"end\": 398}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"# **3 Empirical Strategy**\\n\\nIn this paper we are interested in the relationship between gender norms and economic behaviors disaggregated by the gender of the respondent. To study the relationship between gender norms and observed economic behaviors we conduct multivariate regression analysis and run the following analyses.\\n\\n## **3.1 Correlation of Attitudes and Norms with Economic Behaviors**\\n\\nWhere _Yij_ is the outcome of interest for individual i in country j. Female is a dummy variable equal to 1 if the respondent is female; 0 if male. Equation 1 gives the correlation of an individual’s own attitudes and the outcome of interest. Where _Beliefij_ is a dummy variable equal to 1 if the respondent i personally agrees with a gender statement; and 0 if they disagree. Similarly, equation 2 gives the correlation of the perceived community norm and the outcome of interest. _Normij_ is a continuous variable that captures the perceived community norm (i.e. how many out of 10 neighbors) with respect to the gender role statement that has been standardized between 0-1. _Xij_ is a set of demographic controls of individual i, and _γc_ indicates country fixed effects. Robust Eicker-Huber-White standard errors are used throughout. Equation 1 and 2 are estimated using Ordinary Least Squares (OLS) estimation.\\n\\nIn equations 1 and 2 the interpretation of the coefficients _β_ 0 through _β_ 3 is as follows. _β_ 0 is the intercept, _β_ 1 is the marginal effect of being female on the outcome of interest, _β_ 2 is the marginal effect of the belief/norm on the outcome of interest for men (that is at Female equals 0), _β_ 3 is the differential effect of the belief/norm on the outcomes for women relative to men.\\n\\n## **3.2 Correlation of Norms, Conditional on Individuals’ Attitudes**\\n\\nEquation 3 analyzes the question, conditional on one’s own beliefs, how does perception of what the community thinks correlate with an individual’s economic behaviors. We include individual belief as well as perceived community beliefs in the regressions.\\n\\n12\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"Equation 3 is estimated using Ordinary Least Squares (OLS) estimation. The interpretation of the coefficients _β_ 2 through _β_ 5 is as follows. _β_ 2 gives the marginal effect of the individual agreeing with the gender norm in question on the outcomes for men. _β_ 3 represents the differential effect of agreeing with the norm in question for women relative to men, with _β_ 2+ _β_ 3 being the composite effect of agreeing with the norm in question for women. _β_ 4 is the marginal effect of the perceived community norm on the outcome of interest for men (that is at Female = 0), conditional on personal beliefs. _β_ 5 is the differential effect of the community norms on the outcomes for women relative to men. As such, the sum of _β_ 4+ _β_ 5 gives the marginal effect of community norms on outcomes for women, conditional on personal beliefs.\\n\\n# **4 Results**\\n\\nIn this section, we delve into the analysis of our findings. We begin by conducting descriptive analysis to provide cross-country and cross-region comparisons. The analysis examines personal gender beliefs, perceived community norms, and misperceptions of these norms among both men and women. The results are presented graphically to understand patterns and variations across countries. We then proceed to examine the relationship between gender norms, as delineated by three specific norm-related questions detailed in section 2.3.2, and a range of socio-economic factors. These factors include: (1) labor market dynamics, capturing both labor force participation in the past year and the hours committed to paid work in the week preceding the survey (applicable only to those who were employed during that period); (2) the extent of decision-making autonomy regarding major financial expenditures; and (3) the distribution of time dedicated to caregiving duties and household chores.\\n\\n## **4.1 Global Patterns of Gender Norms and Misperceptions**\\n\\nIn this section we begin by analyzing gender attitudes and norms by region-of-the-world, to observe how norms differ from one part of the world to another. We then examine the relationship between individual beliefs and community norms and examine correlations across countries, as well as their relationship with key macroeconomic indicators. Following this, we assess the nature and extent of misperceptions associated with these gender norms.\\n\\n### **4.1.1 Gender Norms around the World**\\n\\nFigures 1 to 4 outline comparisons of aggregate attitudes and perceived community norms on a regional and country level. Figure 1 details the gender gaps in personal beliefs and norms by region-of-the-world. Personal beliefs are indicated by purple dots for females and 13\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"green dots for males, in contrast to community norms indicated by red and yellow dots. A prominent finding is the global tendency of personal beliefs to be more progressive and counter-stereotypical than what individuals perceive as prevailing community norms, indicating a disconnect between personal beliefs and societal perceptions.\\n\\nThe measures for norms (red and yellow) differ from those for attitudes (purple and green). Notably, female personal beliefs (purple dots) are consistently more progressive than male beliefs (green dots) across all regions. This pattern holds true across East Asia and Pacific (EAP), Europe and Central Asia (ECA), Latin America and the Caribbean (LAC), the Middle East and North Africa (MENA), North America, South Asia, and Sub-Saharan Africa (SSA). Despite this, the perceived community norms for men and women within these regions do not differ greatly. For instance, in Sub-Saharan Africa, 87% of women and 78% of men support gender equality in opportunities personally, but they perceive only 55% of their community does. With regard to the belief that expenses are a man’s responsibility, 39% of women and 52% of men agree, while 60% of the community is perceived to agree. Personal beliefs and perceived community norms about the female caregiver role are closely matched at 69%.\\n\\nFigures 2 to 4 present an analysis of gender attitudes and norms across countries.<sup>19</sup> These figures compare aggregate personal beliefs, represented by green dots, with the average perceived community norms, shown by purple dots, at the country level. The horizontal line within these figures quantifies the gap between personal beliefs and average perceived community norms, thereby illustrating the extent of norm misperception in each country. Countries in the figures are arranged based on the magnitude of this gap.\\n\\nFigure 2 demonstrates the variance in the agreement with the norm that “men and women should have equal opportunities”, which spans from 80.84% in Algeria to 97.69% in Portugal for personal beliefs. The perceived community norms for this statement range from a low of 41.25% in Iraq to a high of 81.44% in Denmark. Meanwhile, Figure 3 depicts the belief that “household expenses should be the man’s responsibility, even if his wife can help him,” ranging from 10.99% in Denmark to 78.06% in Egypt, suggesting a wide spectrum of beliefs about the male breadwinner norm. The perceived community norm ranges from 29% in Denmark to 74.97% in Mali. Notably, Chile shows the largest misperception, with 20% of respondents personally subscribing to the male breadwinner norm, while 57% are perceived to do so by the community. In contrast, Nigeria displays close alignment between personal beliefs and perceived community norms at around 60%. Finally, in terms of the female caregiver norm, as shown in Figure 4, agreement levels vary significantly, from 13.63% in Denmark, indicating more progressive views, to 87.25% in Pakistan. Perceived community norms on this issue also\\n\\n> 19Data by region and country is also provided in Tables A7 to A9.\\n\\n14\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"exhibit a wide range, from 35.91% in Denmark to 80.75% in Bangladesh. Mexico shows the largest gap, with only 20% of respondents personally endorsing the norm that “a woman’s primary role is to care for her home and children,” compared to a perceived community agreement of 66%. On the other hand, Nigeria presents an interesting contrast, where 78% of respondents personally agree with the norm, higher than the 68% they perceive as the community norm, reflecting a more conservative personal belief relative to the perceived community stance.\\n\\n# **4.1.2 Relationship between Beliefs and Perceived Community Norms**\\n\\nIn Appendix B in Figures B1 to B3 we provide further data visualizations of the correlation between beliefs and norms at the country level. We split the analysis by gender and analyze the three norms: the broad norm of gender equality in opportunities, the male breadwinner norm, and the female caregiver norm separately.\\n\\nFor the broad norm regarding gender equality in opportunities, there is a significant difference by gender. Male respondents exhibit a very strong positive correlation, with a Pearson’s correlation coefficient of 0.92, indicating that their personal beliefs are highly consistent with what they perceive their community thinks. Female respondents, on the other hand, show a much lower correlation coefficient of 0.41, suggesting a significant difference between their personal beliefs in support of gender equality and their perception of societal expectations. For the breadwinner norm, there is a strong relationship between individual beliefs and perceived community norms for both genders. Male respondents show a particularly robust correlation, with a coefficient of 0.92, indicating that their own beliefs about financial responsibility in marriage are almost identical to what they view as societal standards. Female respondents also demonstrate a positive correlation, with a coefficient of 0.74. Lastly, when examining the caregiver norm, both female and male respondents show strong positive correlations between their individual beliefs and perceived community norms. Females display a correlation coefficient of 0.73, while males a coefficient of 0.92. This similarity in correlation strength, especially among males, reflects a broad acknowledgment of traditional caregiver roles as the norm within communities.\\n\\nOur findings indicate that for the broad norm of gender equality in opportunities, there is a noticeable gender disparity in how individual beliefs correlate with perceived community norms, with women showing greater support for gender equality than men. For the breadwinner norm, there is a strong correlation for both genders. Across all norms, men consistently demonstrate a stronger correlation, suggesting their own personal beliefs are more closely aligned with societal expectations. The caregiver norm analysis reinforces this pattern.\\n\\n15\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"# **4.1.3 Degree of Norm Misperception across the World**\\n\\nFigures 5 to 7 present maps of the degree of misperception across the globe with respect to the broad norm, the breadwinner norm, and the caregiver norm. These maps are color-coded to represent the varying degrees of misperception across different countries. Where darker shades indicate a greater disparity between personal beliefs and perceived community norms; and lighter shades suggest that norms are more closely aligned with actual beliefs.\\n\\nIn Figure 5, the map quantifies the misperception of the broad norm at the country level and presents the absolute value of the difference. We show a universal underestimation of support for gender equality. Figures 6 and 7 explore misperceptions around the gender role specific norms. Here, positive values signify a perception that society is more traditional or gender-stereotypical than personal beliefs, while negative values indicate the opposite. Darker shades of blue highlight countries where the breadwinner norm is believed to be more conservative or stereotypical in society than actual beliefs held by the Facebook sample. Light blue and yellow-green shades demonstrate a closer match between perceived norms and beliefs or a slight underestimation of the norm i.e. they think their community is more gender progressive or counter-stereotypical than actual beliefs.\\n\\nCollectively, these figures reveal regional patterns where certain areas consistently show misperceptions across all norm constructs, while others display varying degrees of misperception depending on the norm in question. To discern whether misperceptions regarding gender norms are more prevalent among men or women, we delve into a comprehensive analysis in Appendix D. This examination leverages individual responses from our dataset to evaluate the discrepancies between perceived community norms and actual beliefs, dissected by gender and geographic region.\\n\\nAccording to Table D1, the extent of misperception is substantial, with the frequency of under- and over-estimation eclipsing the instances of accurate assessments of community beliefs. It is important to note that when evaluating the level of misperception (i.e., whether an individual overestimates or underestimates a norm), we juxtapose an individual’s perception of their community norm against the aggregate beliefs within their country. The findings indicate that, on average, respondents perceive gender norms within their communities to be more traditional than what national averages indicate. Particularly, individuals in the Latin America and the Caribbean (LAC) and the Middle East and North Africa (MENA) regions exhibit the highest levels of misperception regarding the gender equality norm. Furthermore, the LAC region stands out with the most pronounced misperception concerning gender role norms.\\n\\nSome of what we are labelling as “misperceptions” could in fact reflect sample selection whereby the respondents’ neighbors are not necessarily the same as the average Facebook 16\", \"datasets\": [{\"mention_name\": {\"text\": \"Facebook sample\", \"confidence\": 0.3934740126132965, \"start\": 1127, \"end\": 1142, \"clean_text\": \"Facebook sample\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1127, \"end\": 1142, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1127, \"end\": 1142, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1127, \"end\": 1142}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1127, \"end\": 1142, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.7341388463973999, \"start\": 1127, \"end\": 1142}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.3934740126132965, \"start\": 1127, \"end\": 1142}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5709908604621887, \"start\": 1127, \"end\": 1142}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5709908604621887, \"start\": 1127, \"end\": 1142}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"user. For instance, while 84% of the sample from Iraq personally support gender equality in opportunities, they think that, on average, only 41% of their community would be supportive. Respondents from Iraq on Facebook could be considered a group who are more gender liberal than the general population, or that there is an overall misperception of the norm, where respondents think their community is more conservative than what is actually true. In Table D2 we present analysis to examine whether certain demographic characteristics are correlated with an individual over- or under-estimating his or her community’s norms, e.g. are more educated individuals more aware or informed? In Table D2 column 1 we show that, globally, underestimation of the support for gender equality norm varies by gender and specific characteristics. More highly educated men are less likely to underestimate support for gender equality, and men over 25 years and located in rural areas are more likely to underestimate support for gender equality. More highly educated women are more likely to underestimate the support for gender equality in their community. Women who are older than 25 are less likely to underestimate the gender equality norm. In columns 3 and 5, for the gender roles norms, we show men who are more highly educated are more likely to overestimate the gender roles norms (i.e. think their community is more conservative than it actually is). The age of the female is predictive of the male breadwinner norm where women older than 25 are more likely to underestimate and less likely to overestimate the breadwinner norm. Married women are more likely to overestimate the caregiver norm.\\n\\n# **4.2 Gender Beliefs, Norms and Economic Behaviors**\\n\\nIn this section we present the results of multivariate regression analysis to examine the relationship between gender norms and observed economic behaviors. In Tables 2a to 7 the explanatory variables stay consistent throughout where _Personal Belief:Agree_ is a dummy variable for an individual’s personal attitude to the gender statement in question. _Community Norm_ is a continuous variable between 0 and 1 in increments of 0.1 that indicates the proportion of the community that the respondent thinks agrees with the gender statement in question. We begin with the results for labor, then turn to decision-making power, and finally time spent on childcare and domestic responsibilities.\\n\\n## **4.2.1 Labor**\\n\\nTables 2a, 2b and 2c examine the relationship between the three norm constructs (broad norm on gender equality, male role as breadwinner, and female role as caregiver) and labor market behaviors. The dependent variables are a dummy variable for whether the respondent’s main status is either working (=1) or not working (=0) in the past year; and the hours spent 17\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"working for pay in the previous week, conditional on working in the past 7 days.<sup>20</sup> Tables 2a-2c present the results for the three norm constructs in separate tables for ease for the reader. In Tables 2a to 2c, columns (1) through (4) show the relationship between personal beliefs and the outcomes of interest (equation 1), while columns (5) through (8) show the relationship between perceived community norms and the outcomes of interest (equation 2). We estimate the model in turn without and with country-fixed effects (indicated at the bottom row of the table). In the models with country-fixed effects, the identified variation comes only from within-country variation.\\n\\nThere are clearly variations across countries in how societal beliefs correlate with labor force decisions. For example, in Table 2a men who personally believe in gender equality ( _Personal Belief:Agree_ ) are 2 percentage points more likely to have worked in the past year in the model without country-fixed effects, though this effect disappears once we control for country fixed effects. That is, once we account for cross-country differences, men’s personal beliefs towards gender equality has no correlation with men’s participation in the labor market as we might expect. Interestingly, the results on labor supply of women are relatively similar for the models with and without country-fixed effects. In order to abstract from societal differences, which are likely to introduce an additional level of omitted variables and endogeneity, we focus the rest of our discussion on estimates with country-fixed effects.\\n\\nTables 2a-2c show a number of patterns. Starting with the individual beliefs, we can see that they do not have much of a relationship with male labor force participation. For women, they matter a lot: the more progressive her own beliefs, the more likely she is to be working.<sup>21</sup> Table 2a column 2 suggests that when a woman personally agrees with gender equality in opportunities, she is 5 percentage points more likely to have worked in the past year (see p-value for test _Norms+Fem_ at the bottom of Tables 2a-2c to read the composite effect for women). Similar results are found for the gender role specific norms in Tables 2b and 2c: women who personally agree with the male breadwinner norm are 7 percentage points less likely to be working; and women who agree with the female caregiver norm are 9 percentage points less likely to be working. We find no significant correlation between women’s personal beliefs towards gender equality and hours worked.\\n\\n> 20At the extensive margin participation in the labor market is given by “main status is work in the past year” and at the intensive margin “hours spent on paid work in the past week”. Since the survey was conducted in July 2020 which was when COVID-19 lockdowns started being mandated in some countries, the hours worked variable might have been affected by cases of temporary or recent unemployment during the COVID-19 pandemic. Descriptive statistics suggest that 40% of women and 52% of men in the sample were engaged in work in the past week (compared to 56% of women and 71% of men who report work over the past year). Responses to the main status of work in the past year are less likely to have been significantly influenced by the COVID-19 shock since the survey was conducted relatively early in the pandemic.\\n\\n> 21The coefficient on _Female_ shows that, in general, women are less likely to have worked in the past year than men and spend around one hour less at work conditional on having worked for pay in the last week.\\n\\n18\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"For men, individual beliefs are more important at the intensive margin, where for both the male breadwinner norm and equal opportunities, a more liberal belief held by him is associated with him working fewer hours. For example, men with more egalitarian personal beliefs towards gender equality spend around 0.22 hours less hours at work, conditional on working in the last week (Table 2a and 2b column 4). On the other hand, for women the only belief that is associated with her hours worked is around the female caregiver norm, which goes in the expected direction of 0.32 fewer hours spent working. That is, women who personally agree with the female caregiver norm are less likely to participate in the labor market at both the extensive and intensive margin.\\n\\nTaken alone, community norms show a similar story to individual beliefs for women’s labor force participation: the more liberal she thinks her community is, the more she works. For example, her perception of an additional 10 percent of persons showing support for gender equality in opportunities in the community translates into a 4 percentage point greater likelihood of a woman working over the past year, and an extra 0.49 hours spent working per week. Overall, this pattern of results suggests that women’s perception of greater support for gender equality in opportunities in their community is associated with a higher likelihood of women participating in the labor market at both the extensive and intensive margin. For the gender role specific norms (breadwinner and caregiver), community perceptions are similarly important for a woman’s participation in the labor market, but there is no net significant effect for hours worked. In Tables 2a and 2c, among women, the perception of an additional 10 percent of persons agreeing with the male breadwinner/female caregiver norm in the community translates into a 3 and 4 percentage point lower likelihood of a woman working over the past year, respectively.\\n\\nFor men, again community beliefs are similar to individual beliefs for labor force participation, i.e. what he perceives the community thinks has no relationship on whether he works or not. For the hours he works, gender equal opportunities does not matter, but the gender role norms indicate that men work more hours (conditional on working at all) when they think their community is more gender conservative.\\n\\nThese two levels of beliefs - what individuals think and what they think their community thinks, could be highly related and operating on the same decision simultaneously. In Table 3, we control for both types of beliefs in the same regression (as described for equation 3). For labor force participation, we can see that for the more gender role specific norms (caregiver, breadwinner) controlling for the individual beliefs causes the community beliefs for women to be no longer significant. This suggests that for these more proximal norms, her individual beliefs matter more than those of her community - in contrast to the more distal gender equality norm. For male labor force participation, we see the earlier counterintuitive 19\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"negative result on the female caregiver norms is offset by positive community effects, which suggests that any negative impacts on male labor supply may be driven by men who think they are at odds with their community.<sup>22</sup> At the intensive margin, controlling for the two levels of beliefs (individual or community) does not change our results at all.\\n\\nTaken together, we show that women’s own personal belief in a norm of equal opportunities and their perception of general community support for gender equality are independently correlated with whether women work. On the other hand, women’s perception of community support for male breadwinner and female caregiver norms is not independently correlated with whether and how much they work, net of their own personal beliefs.\\n\\n# **4.2.2 Decision-Making Power**\\n\\nIn the following we analyze intrahousehold decision-making power among married men and women. In Tables 4a, 4b, and 5 we examine correlations between beliefs and norms of broad gender equality and the male breadwinner on power within the household.<sup>23</sup> The dependent variables in Tables 4a-5 present four categories of decision making power related to large purchases in the household: ‘female in the household has all the power’, ‘male in the household has all the power’, ‘joint decision making with spouse’, and ‘no power at all’. The regression models shown include country fixed effects.<sup>24</sup> Table 4a suggests that among men, a higher personal belief in support of gender equality ( _Personal Belief:Agree_ ) is associated with a significant shift away from sole decision making power (Male in HH has all the power), towards joint decision making where men are 10 percentage points less likely to report sole decision power. This is the pattern we might expect when an individual agrees with gender equality in opportunities - that decisions made within the marital relationship are made jointly. Similarly, in Table 4b, the pattern for the male breadwinner norm is in the direction we would expect - personally agreeing with the male breadwinner norm is associated with higher male sole decision making power (9 percentage points) and commensurate lower joint decision making in the household.\\n\\nSimilarly, for women personally agreeing with the gender equality norm is associated with a shift from sole decision making (Female in the HH has all the power) to joint decision making power by 4 percentage points. A similar pattern is found for the male breadwinner norm - agreeing with the breadwinner norm is associated with a lower likelihood of joint power and\\n\\n> 22For instance, men who personally agree with the female caregiver norm may reflect men who place greater weight on family values themselves or men who revert to more traditional masculinity roles when out of work. 23Separate survey modules were randomized and administered to a random subsample in order to minimize overall survey length. Therefore, some behaviors were paired with fewer norms questions.\\n\\n> 24While the two models with and without county fixed effects differ slightly in magnitude of the coefficients, the level of statistical significance is consistent so we only show the specification with country fixed effects.\\n\\n20\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"higher likelihood of sole power (+4 percentage points for the female; +2 percentage points for the male in the household) and 4 percentage points higher likelihood of ‘no power at all’. While the individual beliefs results for men and women follow similar patterns, the magnitudes of the associations are larger for men.\\n\\nTurning to perceived community norms, Table 4a and 4b columns 5 to 8 ( _Community Norms_ ) suggest a similar pattern to own beliefs. When men perceive that their community are more gender liberal then they are more likely to practice joint decision making with their wife and move away from husband-centered decision making. For example, an additional 10 percent in persons who support the gender equality norm is associated with 8 percentage points higher likelihood of men reporting joint decision making; and an additional 10 percent in persons agreeing with the male breadwinner norm is associated with a decrease in joint decision making of 3 percentage points. For women, greater perceived support for gender equality is associated with lower reports of ‘no power at all’ by 5 percentage points and greater joint decision making power. Higher perceived male breadwinner norms by women are associated with lower likelihood of joint decision making, as well as higher sole decision making power for women themselves and a higher probability of reporting ‘no power at all’.\\n\\nTaken together these results suggest that men and women appear to be interacting with the norms differently. Men with more liberal beliefs report a transfer from sole decisionmaking power to joint decision-making power with his spouse. Whereas more liberal women report a transfer from spousal power towards both joint and sole.\\n\\nIn Table 5 we combine personal beliefs and perceived community beliefs in the same regression (as per equation 3). Even after controlling for their personal beliefs, men’s perception of greater acceptance of gender equality in their community continues to be associated with a shift away from husband-centric decision making towards joint decision making (perception of an additional 10 percent in persons agreeing with the norm is associated with -5 percentage points in sole husband and +6 percentage points in joint). For the male breadwinner norm, controlling for the individual beliefs causes the community beliefs to be no longer significant.\\n\\nFor women, after controlling for her personal beliefs, a higher perceived community support for gender equality is associated with a lower likelihood of her having ‘no power at all’ (-5 percentage points), and a higher likelihood of joint decision making in her household. However, we find a more counter-intuitive result for the male breadwinner norm where the community norm is positively correlated with greater sole decision making power by the female. One possible explanation could be that women who are already engaging in counterstereotypical behaviors such as making sole decisions around large purchases may face more social stigma from their community which could reinforce the existence of the stereotypical norm that men are breadwinners for these women.\\n\\n21\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"# **4.2.3 Childcare and Domestic Responsibilities**\\n\\nIn this final set of behaviors, in Tables 6a, 6b, and 7 we examine the relationship between the gender equality norm and female caregiver norm with childcare and domestic responsibilities. There are three dependent variables presented: dummy variable for being the self-reported main caregiver for pre-school or children 6-18 years in the household, hours spent on care activities and hours spent on household chores, conditional on being the main caregiver.\\n\\nIn Tables 6a and 6b again we present the regression model with country fixed effects as shown in equations 1 and 2, first assessing personal beliefs and perceived norms separately. Across the board, women are more likely to report to be the main caregiver of children and spend more time on both care work and household chores (see positive coefficient on _Female_ ).\\n\\nIn Table 6a, we show that personal support for gender equality is not significantly associated with any of the care and household chores outcomes for men (see coefficient on _Personal Belief:Agree_ ). Whereas in Table 6b for men who personally agree with the female caregiver norm we find a higher likelihood of him being the main caregiver of children (+ 3 percentage points), as well as an increase in hours spent on chores (+0.16 hours), and care (+0.41 hours). While this norm is intended to capture beliefs on _females_ as caregivers it may also represent greater family values in general that are held by some men. If a man thinks a woman’s most important role is to take care of her home and children then he may be putting a premium on family and care.<sup>25</sup> Women, in contrast, who personally agree with gender equality in opportunities report spending marginally fewer hours on household chores (-0.24 hours) with no observed relationship with care. For women, personal agreement with the female caregiver norm is strongly positively correlated with both the likelihood she is the main caregiver and the time she spends on care and chores. The composite effect suggests that women who personally agree with the female caregiver norm are 6 percentage points more likely to be main caregivers, and spend an additional 1.05 hours on care work and 0.36 hours on chores.\\n\\nIn terms of the perceived community norm, for men, the perception of an additional 10 percent in persons agreeing with the gender equality norm is associated with spending more time on household chores (+0.35 hours). For women, greater perceptions of agreement with gender equality within the community is strongly associated with time spent on care activities, with an additional 1.36 additional hours on care work for a unit increase in the community norm. This result is surprising since perhaps one might expect redistribution of care work between the husband and wife as the community becomes more gender progressive and therefore a reduction in care activities by women. Interestingly, the perception of the\\n\\n> 25An alternative hypothesis might be a reverse causality explanation that the more men work on carerelated activities, the more they are convinced it is their wife’s job.\\n\\n22\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"female caregiver norm is strongly positively associated with the amount of care and domestic work done by both women and men. Men who perceive that an additional 10 percent of persons in their community agree with the female caregiver norm report spending 1.06 additional hours on care work and 0.43 additional hours on chores. This association is even more pronounced for women, who are 4 percentage points more likely to be main caregivers, spend 1.69 hours more on care work, and 0.51 additional hours on chores for every unit increase in the female caregiver norm.\\n\\nIn Table 7 we combine personal and perceived community norms in the same regression (as per equation 3) and examine the relationship with domestic and care responsibilities, conditioning on the personally held belief. After controlling for personal beliefs, the correlations with the perceived community norms hold as was described earlier for Tables 6a and 6b,. Greater support for the female caregiver norm in the community is strongly associated with greater involvement in care related activities by both men and women (for men, the result is significant on time spent on chores for the gender equality norm, and both time on care and chores for the female caregiver norm). This suggests that even after taking into account their own beliefs, the perceived beliefs of those in the community matter for time spent on activities within the household by men.\\n\\n# **5 Discussion and Concluding Thoughts**\\n\\nGiven the large and persistent gender inequality in labor force participation and earnings in many countries, it is important to better understand how social norms might constrain women’s labor market choices and outcomes. In this paper we provide empirical evidence on the relationship between gender attitudes, norms and economic behavior on a global scale. We examine norms with respect to gender equality in opportunities, and the stereotypical gender roles of the male breadwinner and female caregiver.\\n\\nPersistent gaps in the accurate measurement of norms and obtaining gender-disaggregated data have prevented research at scale. For example, previous research has often relied on country-level aggregates of gender attitudes to proxy for norms. In this paper we leverage a unique dataset that collected data on both personal attitudes and perceived norms on gender across 111 countries and link them to individual-level employment, decision-making and time allocation variables. Our measures shed light on how norms are internalized and acted upon. Measurement of both social expectations and personal beliefs can reveal phenomena, such as pluralistic ignorance, which is an important insight for policies. Our findings suggest a difference between the aggregate country-level measure of personal beliefs and perceived norms across many countries of the world. The extent of the difference varies by region of 23\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"the world, gender and the particular norm in question.\\n\\nResults suggest a general underestimation of the support for gender equality globally (men and women in general think their community is more conservative). We show that education and age are predictive of the degree of underestimation of the support for gender equality. In terms of linking attitudes and norms with behaviors, we show that men’s beliefs and perceived norms about support for gender equality are not correlated with whether he works or not. However, the perceived beliefs of those in the community matter for male engagement in household productive activities. Greater perceived community support for gender equality is associated with greater involvement in care and household chores by married men and a greater likelihood of joint decision-making power with his wife. Women’s own personal beliefs about support for gender equality and gender roles matter a lot for their decision to work. The more progressive her own beliefs, the more likely she is to participate in the labor force. Higher perceived community support for gender equality is also associated with married women having a higher likelihood of joint decision-making power.\\n\\nOur findings, while descriptive in nature, suggest that our measures have important informational content and existing tools to measure social norms need to be improved and refined. For policy, additional data may be needed to be able to diagnose the specific norms at play that bind on behaviors we look to influence. Our findings highlight the entrenched nature of traditional gender roles in the collective consciousness and the varying degrees to which they are personally endorsed by men and women. In addition, identifying the reasons why people comply will help unpack the ‘black box’ of how norms operate. Perhaps people comply because norms are hidden, or because they have a strong desire to conform, or because they gain benefits or fear sanctions for going against a norm.\\n\\nPolicy may encourage more gender progressive personal beliefs, support an updating of misperceived beliefs around gender equality, or encourage more gender progressive or counterstereotypical behaviors, irrespective of norms. Existing gender norms programs in low- and middle- income contexts have typically focused on interventions among youth, communitylevel training programs and low-touch behavioral or information campaigns. For example, recent gender norms programs in India (Dhar, Jain and Jayachandran, 2022) and in Somalia (Brar et al., 2023) find an egalitarian shift in gender attitudes among young adolescents. Bursztyn, Gonzalez and Yanagizawa-Drott (2020) finds that Saudi men systematically overestimate their peers’ disapproval of women’s work and a simple information intervention that corrects misperceptions increased men’s willingness to help their wife search for a job. In the Democratic Republic of Congo (DRC) training that engaged fathers in sessions of critical reflection on fatherhood and caregiving led to higher levels of male participation in childcare and household tasks, relative to a control group (Vaillant et al., 2020). In Nigeria, Banerjee, 24\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"Ferrara and Orozco (2019) find that exposure to an edutainment intervention improved television viewers’ attitudes towards violence against women. Bertrand (2020) proposes that direct and ongoing exposure to a proscribed counter-stereotypical behavior, such as women’s work outside the home, may eventually reduce norm-related costs associated with the behavior and eventually help erode the norm.\\n\\nOften the settings where we want to use policy to influence a norms change are in lowresource settings with low levels of human capital. In that sense, social norms interventions should not be considered standalone solutions but complementary to capital- or skills-related programming. There are a variety of channels through which policy could operate that may be norm-sensitive or norm-transformative. Below we provide ideas for a variety of potential entry points for policy interventions to address social norms either directly or indirectly (Munoz-Boudet, Pierotti and Rahman, 2023).\\n\\n- **Circumvent** : work around the norm with behaviors that are more accepted (e.g. girls completing education before marriage, or women working from the homestead) or conduct interventions in locations/places where women are more likely to be present (e.g. in health centers, or collection of cash transfers).\\n\\n- **Prevent** : take actions for the norm to not be triggered by separating women (e.g. women’s only transport, female-only workplaces, or self-help groups).\\n\\n- **Eliminate sanctions ‘myth’** : create default options that can bypass normative choices or use role models and social proof activities that show sanctions are not enacted in reality.\\n\\n- **Tackle** : engage in collective discussions on a norm, change or enact legislation, or change aspirations and influence younger generations prior to path dependence.\\n\\n- **Create a new norm** : promote early child development and involved parenting (behavioral discourse change). Introduce incentives (financial, legal) for new behaviors that break with the norm.\\n\\nGovernment policy that embeds gender norms programming in school curricula can be a pathway to scale. To effect a real and lasting shift in norms may require a targeted policy approach that encourages and applauds deviation from the stereotypical gender roles and promotes gender equality. Policy makers that promote gender equality need to ensure greater investment is devoted towards programs that could affect norms. Governments and media can also play a more active role in challenging the social dialogue around the male breadwinner and female caregiver gender roles in the quest to support gender equality and promote economic growth.\\n\\n25\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"**6 All tables and figures**\\n\\n26\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [27]}}, {\"page\": 28, \"chunk\": 28, \"input_text\": \"Table 1: Sample Characteristics by Respondent Gender\\n\\n||(1)|(2)|t-test|\\n|---|---|---|---|\\n||Male|Female|Difference|\\n|Variable|Mea|n/SE|(1)-(2)|\\n|Marital Status: Having a long-term partner or a spouse|0.59|0.57|0.02***|\\n||[0.00]|[0.00]||\\n|Relationship with head of household||||\\n|Head of Household|0.44|0.25|0.19***|\\n||[0.00]|[0.00]||\\n|Spouse or partner|0.16|0.37|-0.22***|\\n||[0.00]|[0.00]||\\n|Child or Grandchild|0.26|0.24|0.02***|\\n||[0.00]|[0.00]||\\n|Other|0.15|0.13|0.01***|\\n||[0.00]|[0.00]||\\n|Number of people under same roof (excluding respondent)||||\\n|0, I live alone|0.09|0.07|0.02***|\\n||[0.00]|[0.00]||\\n|1 person|0.14|0.17|-0.03***|\\n||[0.00]|[0.00]||\\n|2-5 people|0.61|0.62|-0.01***|\\n||[0.00]|[0.00]||\\n|6-10 people|0.13|0.12|0.01***|\\n||[0.00]|[0.00]||\\n|11 or more|0.03|0.02|0.01***|\\n||[0.00]|[0.00]||\\n|Education||||\\n|Secondary or less|0.42|0.39|0.02***|\\n||[0.00]|[0.00]||\\n|More than secondary|0.58|0.61|-0.02***|\\n||[0.00]|[0.00]||\\n|Age||||\\n|24 or younger|0.26|0.28|-0.01***|\\n||[0.00]|[0.00]||\\n|25-64 years old|0.74|0.72|0.01***|\\n||[0.00]|[0.00]||\\n|Urbanicity||||\\n|City|0.60|0.61|-0.01***|\\n||[0.00]|[0.00]||\\n|Village, rural area, or other|0.40|0.39|0.01***|\\n||[0.00]|[0.00]||\\n|Own land|0.23|0.16|0.07***|\\n||[0.00]|[0.00]||\\n|Observations<br>***, **, and * indicate significance at the 1, 5, and 10 percent criti<br>The value displayed for t-tests are the differences in the means acr|73535<br>cal level.<br>oss the gro|79020<br>ups.||\\n\\n27\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [28]}}, {\"page\": 29, \"chunk\": 29, \"input_text\": \"<!-- Start of picture text --> Broad Norm: Men and women should have equal opportunities<br>Gender Gaps in Personal Beliefs and Perceived Community Norms<br>100<br>95<br>“1185 l : ‘<br>80<br>7065 ! 8<br>55 |<br>50 I<br>a a i i i, a |<br>EAP ECA LAC MENA North America South Asia SSA<br>Region of the world<br>@ Female - Personal Beliefs @ Female - Perceived Community Norms<br>@ = Male - Personal Beliefs @ = Male - Perceived Community Norms<br>Personal beliefs are an aggregate of the percentage of the sample who agree with the statement.<br>Perceived community norms are an aggregate of the proportion of neighbors who they believe agree with the statement.<br>Aggregates are taken at the country level and then averaged by region of the world.<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> Breadwinner Norm: Expenses are a man's responsibility even if his wife can help him<br>Gender Gaps in Personal Beliefs and Perceived Community Norms<br>80<br>75<br>70<br>65 t<br>60 3 e<br>55 ®<br>50 3<br>s& 4035 |<br>30<br>25<br>20<br>15<br>10<br>5<br>0<br>EAP ECA LAC MENA North America South Asia SSA<br>Region of the world<br>@ Female - Personal Beliefs @ Female - Perceived Community Norms<br>@ = Male - Personal Beliefs @ Male - Perceived Community Norms<br>Personal beliefs are an aggregate of the percentage of the sample who agree with the statement.<br>Perceived community norms are an aggregate of the proportion of neighbors who they believe agree with the statement.<br>Aggregates are taken at the country level and then averaged by region of the world.<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> Caregiver Norm: Women's most important role is to take care of her home and children<br>Gender Gaps in Personal Beliefs and Perceived Community Norms<br>80<br>75<br>65-370 :<br>60 e ?<br>5550 !<br>45<br>s& 40<br>35<br>30 8 i<br>25<br>20<br>15<br>10<br>5<br>0<br>EAP ECA LAC MENA North America South Asia SSA<br>Region of the world<br>@ Female - Personal Beliefs @ Female - Perceived Community Norms<br>@ = Male - Personal Beliefs @ Male - Perceived Community Norms<br><!-- End of picture text --> Personal beliefs are an aggregate of the percentage of the sample who agree with the statement. Perceived community norms are an aggregate of the proportion of neighbors who they believe agree with the statement. Aggregates are taken at the country level and then averaged by region of the world.\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [29]}}, {\"page\": 30, \"chunk\": 30, \"input_text\": \"<!-- Start of picture text --> Broad Norm: Men and women should have equal opportunities<br>Personal Belief and Perceived Community Norms<br>D : : : : : : : : : : : : : : : : ; : :<br>United States of America . . . - - . . . . . . . . . : =? .<br>Visa : : : : : : : : : : : : : -——_— # : :<br>Netherlands ee<br>United Kinedam : : : : . : . : : : : : : ee<br>Bangladesh : : : : : : : : : : : : OOOO : :<br>Laos —_—_—<br>Belgium. : : : : : : : : : : : : : : 14——————<br>Jamaica : : : : : : : : : : : : : ———_——____—-* :<br>= : : : : : : : : : : : : : 4 : : : 5 :<br>Coun 2: : : : : : : : : : : : . ° : : : : : :<br>Japan |<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> Getmany ee<br>Benublic —— ui__ i»<br>a_F<br>Turkey : : : : : : : : : é : : —_—<br>: : : : : : : : : : : : oo<br>ne © ,? OF:<br>Tunisia ee...0 0s<br>France AVX i<br>Tada _—<br>Spe : : : : : : : : : : : : 3<br>Mazambare Ss<br>Georgia : : : : : : : : : : —oO_ToTY——\\\"*ao i : :<br>Moldova oe _-<br>Anzola : : : : : : : : : UH: : :<br>Malai —_—<br>Burkina Faso Ot<br>Croatia —_—<br>Lebanon HH<br>oneBartazalBrazil : : : : : : : : i- : eS____OOes '’=<br>Albania o_O<br>CT TT &,T TTT TT TT TT TCS STS CST<br>0 3 10 15 20 23 30 33 40 45 30 35 60 65 70 rh) 80 85 90 95 100<br>« Community ® Individual<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [30]}}, {\"page\": 31, \"chunk\": 31, \"input_text\": \"<!-- Start of picture text --> Breadwinner Norm: Expenses are a man's responsibility even if his wife can help him<br>Personal Beliefs and Perceived Community Norms<br>Chis : : : : : : : : : : - : : : : : : : :<br>Bl Saar : : : WOOO : : : : : : :<br>Dominican Republic : : : : —_— 1 : : : : : : :<br>Huneary : : a : : : : : : - - . : : : : : : :<br>Sn : : : : : : : : : : - : : : : : : :<br>Serbia : : : a is : : : : : : : :<br>Switzerland : : ee : : : : : : : : :<br>United Statesof America : : a : - : : : : : . 7 .<br>Jamaica : : : : | : : : : . - : :<br>Sqariand : : : : : : : : : : : : : : : : : :<br>(Czech Republic WRee<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> Philippines —<br>i<br>Bulear: : : . . ——————_, - : : : : . . : :<br>Pauenon : : : ° : : : ee : : : . : :<br>ZambabweGemany : : : : :2: : So S— : : : : : :<br>guier : : : : : : : : : . - : : : : : : : :<br>United Amb Pee : : : : : : : : —o— : : : : : : :<br>Latvia : : : : : al : : : : : : : : :<br>dapsnaeF Moldova“ineals; : : : : : : : | EEE : : _——_.....~C—___—__—-+*_ : : : : :<br>hee : : : : : : : : :—__———_+ : : : : : : : :<br>a : : : : : : : : F : . : : : : : : : :<br>Malaysia : : : : : : : : : : : «+ : : : : :<br>Thailand e+<br>1Bahrain :: :: :: :: :: :: :: :: :: /: ges:: : : :: :: :: :: :: ::<br>Laos : : : : : : : : : s———_+ . : : : : : : :<br>Belarus. 4<br>Saudi Arabia -——<br>Morocco : : : : : : : : : >: -——-: : : : : : : :<br>Haiti #——*<br>Kuwait : : : : : : : : : : : _ -——es) : : : : :<br>Supe ease : : : : - : : : : : -— +: : : : - : :<br>Mali : : : : : : : : : : : : : > : : : :<br>Jordan 4<br>Kazakhstan. -_e<br>igeria<br>2sBrne :: :: :: :: :: :: :: :: :: :: .: nea ai :: :: :: .: :: ::<br>Libya : : : : : : : : : : : ee : : : : :<br>0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100<br>« Community ® Individual<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [31]}}, {\"page\": 32, \"chunk\": 32, \"input_text\": \"<!-- Start of picture text --> Caregiver Norm: Women's most important role is to take care of her home and children<br>Personal Beliefs and Perceived Community Norms<br>Mexico : : : . : : : : : : - : : : : : : : :<br>Guatemala : : : : : a : : : : : :<br>El SalvadorSpain : : -—_S_: : 010,00Ss a OSS: ————s: : : : : .: : :<br>Argentina : : : : : ——_—Wa WH, * : : : : : : :<br>Slovakia : : : : ee| : : : : : : : :<br>Eelsiom : : : ° - : : : —— : : : : : : :<br>Poland : : : : : [ee : : : : : : :<br>United Kingdom . . . - a . . . . . . . . .<br>Sorbia : : : : : : : ‘° ° : : : : : :<br>Slovenia ‘ ‘ : : : i2| : ‘ : ‘ : : : ‘<br>Switzerland : : : : : a> * : : : : : : : : :<br>Czech HondurasRepublic : : : : : : : asesa : : : : : : :<br>United States of America : : : : : : _ : : : : : : : :<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> aRole :: :: :: :- :: :_E :: :ee: : : : :: :: :: :- :: ::<br>BelarusMS :: :: :: :: :: :: :: :: :SC;- : : :: :: :: :: :: ::<br>wazzea : : : : : : : : : : : : : ‘ : : : : :<br>tabaE : : : : : : : : : : re_ : : : : :<br>Malaya : : : : : : : : : : : —_———*o—_—____—__- : : : : :<br>MoroccoLatvia : : : : : : : : : —_—*“+4 : : : : : : :<br>aaa2 : : : : : : : : : —— : : : : : : :<br>SmbamePalestine205 :: :: :: :. :- :: :: :: :: :: /-oe :-4 —_—s: :: :: :. :: ::<br>has : : : : : : : : : : : : _-— : : : : :<br>Piulpaines : : : : : : : : : : - : : eo : : : :<br>Uanet ampSuratSS :: :: :: :: :: :: :: :: :: :: :- :: [es, ou :: :: :. :: ::<br>oedageacar :: :: :: :- :- :: :: :: :: :: :- :: aLegaoe * :: :: :- :: ::<br>‘aaa : : : : : : : : : : - : ‘ee T* : : . : :<br>Uganda : : : - - : : : : : - : : eo : : - : :<br>Suan : : ; : : : : : : ; : : . - ; : : :<br>Myanmar : : : : : : : : : : : : .: : : : : : :<br>Afghanistan_, mene : : : : : : : : : : : : : 7 s : : : : :<br>Unbelastm -—<br>Maunitania —<br>Jordan : : : : : : : : : : : : _ : : : : :<br>Algeria eo<br>Lesotho : : : : : : : : : : : : : : —e - : : :<br>Buin : : : : : : : : : : : : es : : : : :<br>Libya : : : : : : : : : : : : [-_— : ; : : :<br>Banzladeth «#——<br>Pakistan : : : : : : : : : : : : : : eo : :<br>Ezypt e+<br>sorwnne : : : : : : : : : : . : : +: ———— : :<br>oo|<br>0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 TS 80 85 90 95 100<br>e Community © Individual<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [32]}}, {\"page\": 33, \"chunk\": 33, \"input_text\": \"IBRD 47747 | MARCH 2024\\n\\n<!-- Start of picture text --> Broad Norm: men and women should have equal opportunities<br>Misperception ofthe norm: |Perceived Community Norms - Personal Beliefs (%)|<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> Z<br>-<br>; a<br>ng \\\\ vn :<br>F * tee ns<br>Ss eG, \\\\<br>‘@ — a<br>Sin, a 7 ! ; i Ly<br>A | } , /<br>45-50 - yy Ni@eis<br>Mi 35-40 neae. sy “4<br>WM 30-35 4<br>WmMm 25-3020-25 Sy ae<br>my 15-20<br>my 10-15<br>fy 5-10<br>0-5<br>No data<br>-<br><!-- End of picture text -->\\n\\nMisperception is determined by the disparity between the Community perceived norm (on a 0-100 scale) and the Aggregate Level ofAgreement with the norm within the country (on a 0-100 scale). Here we present the absolute values ofthe misperception as all countries have a negative value suggesting a general misperception of the support for gender equality globally. Darker shades ofthe map represent a greater misperception of the norm i.e. a greater underestimation in the level of support for gender equality in opportunities in the country. Source: Gender Equality at Home Data, 2020.\", \"datasets\": [{\"mention_name\": {\"text\": \"Gender Equality at Home Data\", \"confidence\": 0.5936136245727539, \"start\": 1081, \"end\": 1109, \"clean_text\": \"Gender Equality at Home Data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1081, \"end\": 1109, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1081, \"end\": 1109, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1081, \"end\": 1109}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1081, \"end\": 1109, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.444389671087265, \"start\": 668, \"end\": 696}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5936136245727539, \"start\": 668, \"end\": 696}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4873109757900238, \"start\": 668, \"end\": 696}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4873109757900238, \"start\": 668, \"end\": 696}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [33]}}, {\"page\": 34, \"chunk\": 34, \"input_text\": \"IBRD 47748 | MARCH 2024\\n\\n<!-- Start of picture text --> Breadwinner Norm: Expenses are a man’s responsibility even if his wife can help him<br>Misperception of the norm: Perceived Community Norms - Personal Beliefs (9)<br>F a<br>rey € aa \\\\<br>| | rte ><br>t bees é 8<br>we } ire:|<br>—<br>Wi 20-25 aor: wy 4<br>Mi 15-20 eel<br>mmmm 5-1010-15 at rs.4<br>my 0-5<br>my -5-0<br>mm -10--5<br>-15--10 4<br>No data<br>a.<br><!-- End of picture text -->\\n\\nMisperception is determined by the disparity between the Community perceived norm (on a 0-100 scale) and the Aggregate Level ofAgreement with the norm within the country (on a 0-100 scale). Here we present the actual values of the misperception — some countries havea positive value (overestimation) andsome a negative value (underestimation) suggesting varying degrees of misperception. Darker shades ofthe map suggest an overestimation of the male breadwinnernorm — individuals think their community is more gender conservative or stereotypical than actual beliefs in that country. Lighter shades ofthe map suggest an underestimation of the male breadwinner norm — individuals think their community is more gender progressive or counter-stereotypical than actual beliefs in that country. Source: Gender Equality at Home Data, 2020.\", \"datasets\": [{\"mention_name\": {\"text\": \"Gender Equality at Home Data\", \"confidence\": 0.5727502703666687, \"start\": 1245, \"end\": 1273, \"clean_text\": \"Gender Equality at Home Data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1245, \"end\": 1273, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1245, \"end\": 1273, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1245, \"end\": 1273}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1245, \"end\": 1273, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.444389671087265, \"start\": 833, \"end\": 861}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5727502703666687, \"start\": 833, \"end\": 861}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4873109757900238, \"start\": 833, \"end\": 861}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4873109757900238, \"start\": 833, \"end\": 861}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [34]}}, {\"page\": 35, \"chunk\": 35, \"input_text\": \"IBRD 47749 | MARCH 2024\\n\\n<!-- Start of picture text --> Caregiver Norm: Women’s most important role is to take care of her home and children<br>Misperception ofthe norm: Perceived Community Norms - Personal Beliefs (%)<br>a eK, \\\\<br>Jaunt, ~.<br>~ ee a<br>Mi 40-50 ak \\\" i yw t | ‘4 |aas<br>mi 35-40 . ay Ni@eis<br>WM 25-30 aeae ay, “4<br>WH 20-25 2<br>mmW@ 15-2010-15 Sy pe)<br>my 0-10 :<br>my -5-0<br>mm -10--5<br>! -20--10 4<br>No data<br>a<br><!-- End of picture text -->\\n\\nMisperception is determined by the disparity between the Community perceived norm (on a 0-100 scale) and the Aggregate Level ofAgreement with the norm within the country (on a 0-100 scale). Here we present the actual values ofthe misperception — some countries havea positive value (overestimation) andsome a negative value (underestimation) suggesting varying degrees of misperception. Darker shades ofthe map suggest an overestimation of the female caregiver norm — individuals think their community is more gender conservative or stereotypical than actual beliefs in that country. Lighter shades ofthe map suggest an underestimation of the female caregiver norm — individuals think their community is more gender progressive or counter-stereotypical than actual beliefs in that country. Source: Gender Equality at Home Data, 2020.\", \"datasets\": [{\"mention_name\": {\"text\": \"Gender Equality at Home Data\", \"confidence\": 0.4390663802623749, \"start\": 1274, \"end\": 1302, \"clean_text\": \"Gender Equality at Home Data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1274, \"end\": 1302, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1274, \"end\": 1302, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1274, \"end\": 1302}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1274, \"end\": 1302, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.444389671087265, \"start\": 848, \"end\": 876}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4390663802623749, \"start\": 848, \"end\": 876}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4873109757900238, \"start\": 848, \"end\": 876}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4873109757900238, \"start\": 848, \"end\": 876}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [35]}}, {\"page\": 36, \"chunk\": 36, \"input_text\": \"Table 2a: Labor Outcomes - Correlation with Gender Equality in Opportunities Attitudes and Perceived Norms\\n\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|\\n|---|---|---|---|---|---|---|---|---|\\n||Main<br>Work(Work<br>=0|Status<br>=1/Nowork<br>)|Hours spe<br>work la<br>(Condi|nt at paid<br>st week<br>tional)|Main<br>Work(Work<br>=|Status<br>=1/Nowork<br>0)|Hours sp<br>work la<br>(Cond|ent at paid<br>st week<br>itional)|\\n||**_G_**|**_ender Equal_**|**_ity Norm:_**|**_Men and wo_**|**_men should_**|**_have equal o_**|**_pportuniti_**|**_es_**|\\n|Female|-0.18***|-0.20***|-0.86***|-0.96***|-0.13***|-0.16***|-0.91***|-1.06***|\\n||[0.01]|[0.01]|[0.13]|[0.13]|[0.01]|[0.01]|[0.12]|[0.12]|\\n|Personal Belief:Agree(No=0/Yes=1)|0.02***<br>[0.01]|-0.01<br>[0.01]|-0.15**<br>[0.07]|-0.22***<br>[0.07]|||||\\n|Personal Belief:Agree*Female|0.05***<br>[0.01]|0.06***<br>[0.01]|0.28**<br>[0.13]|0.30**<br>[0.13]|||||\\n|Community Norms (0/1)|||||0.05***<br>[0.01]|0.00<br>[0.01]|0.10<br>[0.11]|-0.07<br>[0.12]|\\n|Community Norms (0/1)*Female|||||0.02<br>[0.01]|0.04***<br>[0.01]|0.48***<br>[0.17]|0.56***<br>[0.17]|\\n|Constant|0.83***|0.60***|8.73***|8.46***|0.85***|0.63***|8.46***|8.06***|\\n||[0.01]|[0.02]|[0.10]|[0.32]|[0.01]|[0.03]|[0.12]|[0.40]|\\n|Observations|100659|100659|55094|55094|61319|61319|37555|37555|\\n|Adjusted R-squared|0.10|0.14|0.01|0.03|0.11|0.15|0.02|0.03|\\n|Mean Female|0.56|0.56|7.54|7.54|0.59|0.59|7.51|7.51|\\n|Mean Male|0.71|0.71|8.18|8.18|0.73|0.73|8.14|8.14|\\n|Norms+Fem(Coeff)|0.07|0.05|0.13|0.07|0.07|0.04|0.58|0.49|\\n|Norms+Fem(SE)|0.01|0.01|0.11|0.11|0.01|0.01|0.13|0.13|\\n|Norms+Fem(P-value)|0.00|0.00|0.25|0.52|0.00|0.00|0.00|0.00|\\n|Sample|ALL|ALL|ALL|ALL|ALL|ALL|ALL|ALL|\\n|CountryFE|NO|YES|NO|YES|NO|YES|NO|YES|\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [36]}}, {\"page\": 37, \"chunk\": 37, \"input_text\": \"Table 2b: Labor Outcomes - Correlation with Male Breadwinner Norm Attitudes and Perceived Norms\\n\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|\\n|---|---|---|---|---|---|---|---|---|\\n||Main<br>Work(Work<br>=|Status<br>=1/Nowork<br>0)|Hours spe<br>work la<br>(Condi|nt at paid<br>st week<br>tional)|Main<br>Work(Work<br>=|Status<br>=1/Nowork<br>0)|Hours sp<br>work la<br>(Cond|ent at paid<br>st week<br>itional)|\\n||**_Male Brea_**|**_dwinner Nor_**|**_m: Expens_**|**_es are a ma_**|**_n's responsib_**|**_ility even if _**|**_his wife ca_**|**_n help him_**|\\n|Female|-0.12***|-0.13***|-0.60***|-0.67***|-0.10***|-0.12***|-0.39***|-0.49***|\\n||[0.00]|[0.00]|[0.07]|[0.07]|[0.01]|[0.01]|[0.12]|[0.12]|\\n|Personal Belief:Agree(No=0/Yes=1)|-0.02***|-0.00|0.19**|0.28***|||||\\n||[0.01]|[0.01]|[0.07]|[0.08]|||||\\n|Personal Belief:Agree*Female|-0.07***|-0.07***|-0.34***|-0.30**|||||\\n||[0.01]|[0.01]|[0.12]|[0.12]|||||\\n|Community Norms (0/1)|||||-0.02*|0.01|0.39***|0.43***|\\n||||||[0.01]|[0.01]|[0.13]|[0.13]|\\n|Community Norms (0/1)*Female|||||-0.05***|-0.04***|-0.55***|-0.52***|\\n||||||[0.01]|[0.01]|[0.19]|[0.19]|\\n|Constant|0.74***|0.62***|8.57***|7.95***|0.87***|0.61***|8.42***|7.93***|\\n||[0.01]|[0.03]|[0.12]|[0.43]|[0.01]|[0.03]|[0.14]|[0.46]|\\n|Observations|51369|51369|27286|27286|44173|44173|25682|25682|\\n|Adjusted R-squared|0.11|0.15|0.02|0.03|0.11|0.15|0.02|0.03|\\n|Mean Female|0.57|0.57|7.50|7.50|0.58|0.58|7.49|7.49|\\n|Mean Male|0.72|0.72|8.24|8.24|0.73|0.73|8.22|8.22|\\n|Norms+Fem(Coeff)|-0.09|-0.07|-0.15|-0.01|-0.06|-0.03|-0.16|-0.09|\\n|Norms+Fem(SE)|0.01|0.01|0.09|0.10|0.01|0.01|0.14|0.14|\\n|Norms+Fem(P-value)|0.00|0.00|0.11|0.90|0.00|0.00|0.25|0.51|\\n|Sample|ALL|ALL|ALL|ALL|ALL|ALL|ALL|ALL|\\n|CountryFE|NO|YES|NO|YES|NO|YES|NO|YES|\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [37]}}, {\"page\": 38, \"chunk\": 38, \"input_text\": \"Table 2c: Labor Outcomes - Correlation with Female Caregiver Norm Attitudes and Perceived Norms\\n\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|\\n|---|---|---|---|---|---|---|---|---|\\n||Main<br>Work(Work<br>=0|Status<br>=1/Nowork<br>)|Hours spe<br>work la<br>(Condi|nt at paid<br>st week<br>tional)|Main<br>Work(Work<br>=|Status<br>=1/Nowork<br>0)|Hours sp<br>work la<br>(Cond|ent at paid<br>st week<br>itional)|\\n||**_Female Ca_**|**_regiver Nor_**|**_m: Woman'_**|**_s most impo_**|**_rtant role is_**|**_to take care_**|**_of home an_**|**_d children_**|\\n|Female|-0.11***|-0.12***|-0.35***|-0.46***|-0.10***|-0.11***|-0.22|-0.36***|\\n||[0.01]|[0.01]|[0.07]|[0.07]|[0.01]|[0.01]|[0.14]|[0.14]|\\n|Personal Belief:Agree(No=0/Yes=1)|-0.05***<br>[0.01]|-0.02***<br>[0.01]|-0.03<br>[0.07]|0.09<br>[0.08]|||||\\n|Personal Belief:Agree*Female|-0.07***<br>[0.01]|-0.07***<br>[0.01]|-0.46***<br>[0.11]|-0.41***<br>[0.11]|||||\\n|Community Norms (0/1)|||||-0.03**|0.01|0.20|0.33**|\\n||||||[0.01]|[0.01]|[0.13]|[0.13]|\\n|Community Norms (0/1)*Female|||||-0.04***<br>[0.01]|-0.05***<br>[0.01]|-0.45**<br>[0.19]|-0.40**<br>[0.19]|\\n|Constant|0.86***|0.60***|8.58***|8.44***|0.87***|0.56***|8.46***|7.78***|\\n||[0.01]|[0.03]|[0.12]|[0.46]|[0.01]|[0.03]|[0.14]|[0.45]|\\n|Observations|50048|50048|28020|28020|44275|44275|26230|26230|\\n|Adjusted R-squared|0.11|0.15|0.01|0.03|0.10|0.14|0.01|0.03|\\n|Mean Female|0.56|0.56|7.58|7.58|0.56|0.56|7.57|7.57|\\n|Mean Male|0.70|0.70|8.14|8.14|0.71|0.71|8.11|8.11|\\n|Norms+Fem(Coeff)|-0.12|-0.09|-0.49|-0.32|-0.07|-0.04|-0.25|-0.07|\\n|Norms+Fem(SE)|0.01|0.01|0.08|0.08|0.01|0.01|0.15|0.15|\\n|Norms+Fem(P-value)|0.00|0.00|0.00|0.00|0.00|0.00|0.09|0.63|\\n|Sample|ALL|ALL|ALL|ALL|ALL|ALL|ALL|ALL|\\n|CountryFE|NO|YES|NO|YES|NO|YES|NO|YES|\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [38]}}, {\"page\": 39, \"chunk\": 39, \"input_text\": \"Table 3: Labor Outcomes - Correlation with Perceived Community Norms, conditional on Personal Beliefs\\n\\n||(1)|(2)|(3)|(4)|(5)|(6)|\\n|---|---|---|---|---|---|---|\\n||Main Status<br>Work(Work=1/N<br>owork=0)|Hours spent at<br>paid work last<br>week<br>(Conditional)|Main Status<br>Work(Work=1/N<br>owork=0)|Hours spent at<br>paid work last<br>week<br>(Conditional)|Main Status<br>Work(Work=1/N<br>owork=0)|Hours spent at<br>paid work last<br>week<br>(Conditional)|\\n||**_Gender Equ_**|**_ality Norm_**|**_Male Breadw_**|**_inner Norm_**|**_Female Care_**|**_giver Norm_**|\\n|Female|-0.21***|-1.26***|-0.11***|-0.45***|-0.10***|-0.32**|\\n||[0.01]|[0.18]|[0.01]|[0.12]|[0.01]|[0.14]|\\n|Personal Belief:Agree(No=0/Yes=1)|-0.00|-0.18*|0.01|0.25***|-0.02***|0.01|\\n||[0.01]|[0.09]|[0.01]|[0.08]|[0.01]|[0.08]|\\n|Personal Belief:Agree*Female|0.05***|0.28*|-0.08***|-0.27**|-0.07***|-0.35***|\\n||[0.01]|[0.17]|[0.01]|[0.13]|[0.01]|[0.12]|\\n|Community Norms (0/1)|0.00|-0.01|0.01|0.34**|0.02*|0.31**|\\n||[0.01]|[0.12]|[0.01]|[0.13]|[0.01]|[0.14]|\\n|Community Norms (0/1)*Female|0.04***|0.49***|-0.02|-0.41**|-0.03*|-0.25|\\n||[0.01]|[0.17]|[0.01]|[0.19]|[0.01]|[0.21]|\\n|Constant|0.63***|8.18***|0.61***|7.86***|0.58***|7.80***|\\n||[0.03]|[0.40]|[0.03]|[0.46]|[0.03]|[0.45]|\\n|Observations|61319|37555|44173|25682|44275|26230|\\n|Adjusted R-squared|0.15|0.03|0.15|0.03|0.15|0.03|\\n|Mean Female|0.59|7.51|0.58|7.49|0.56|7.57|\\n|Mean Male|0.73|8.14|0.73|8.22|0.71|8.11|\\n|PersBelief+Fem(P-value)|0.00|0.47|0.00|0.84|0.00|0.00|\\n|ComNorms+Fem(P-value)|0.00|0.00|0.39|0.60|0.56|0.67|\\n|Mean VIF|3.51|3.62|2.26|2.29|2.51|2.56|\\n|Sample|ALL|ALL|ALL|ALL|ALL|ALL|\\n|CountryFE|YES|YES|YES|YES|YES|YES|\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [39]}}, {\"page\": 40, \"chunk\": 40, \"input_text\": \"Table 4a: Decision Making Power over Large Purchases - Correlation with Gender Equality Norm: Attitudes and Norms\\n\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|\\n|---|---|---|---|---|---|---|---|---|\\n||Female in<br>HH has all<br>the power<br>(Yes=1/No<br>=0)|Male in<br>HH has all<br>the power<br>(Yes=1/No<br>=0)|Joint<br>decision<br>making<br>with<br>spouse<br>(Yes=1/No|No power<br>at all<br>(Yes=1/No<br>=0)|Female in<br>HH has all<br>the power<br>(Yes=1/No<br>=0)|Male in<br>HH has all<br>the power<br>(Yes=1/No<br>=0)|Joint<br>decision<br>making<br>with<br>spouse<br>(Yes=1/No|No power<br>at all<br>(Yes=1/No<br>=0)|\\n||**_G_**|**_ender Equa_**|=0)<br><br>**_lity Norm:_**|**_Men and wo_**|**_men should_**|**_have equal_**|=0)<br><br>**_opportuniti_**|**_es_**|\\n|Female|0.17***|-0.18***|0.03**|0.02|0.13***|-0.12***|0.00|0.03**|\\n||[0.01]|[0.01]|[0.01]|[0.01]|[0.01]|[0.01]|[0.01]|[0.01]|\\n|Personal Belief:Agree(No=0/Yes=1)|-0.00|-0.09***|0.10***|-0.00|||||\\n||[0.01]|[0.01]|[0.01]|[0.01]|||||\\n|Personal Belief:Agree*Female|-0.04***|0.09***|-0.06***|-0.02|||||\\n||[0.01]|[0.01]|[0.01]|[0.01]|||||\\n|Community Norms (0/1)|||||0.01|-0.07***|0.08***|-0.01|\\n||||||[0.01]|[0.01]|[0.01]|[0.01]|\\n|Community Norms (0/1)*Female|||||-0.01|0.05***|-0.05***|-0.05***|\\n||||||[0.01]|[0.02]|[0.02]|[0.02]|\\n|Constant|0.24***|0.59***|0.12***|0.02|0.26***|0.48***|0.17***|0.00|\\n||[0.02]|[0.03]|[0.03]|[0.03]|[0.03]|[0.04]|[0.05]|[0.04]|\\n|Observations|42669|42669|42669|42669|25183|25183|25183|25183|\\n|Adjusted R-squared|0.06|0.11|0.20|0.20|0.06|0.09|0.22|0.25|\\n|Mean Female|0.17|0.09|0.56|0.19|0.15|0.07|0.57|0.19|\\n|Mean Male|0.07|0.22|0.51|0.18|0.06|0.18|0.54|0.18|\\n|Norms+Fem(Coeff)|-0.04|-0.01|0.03|-0.02|0.00|-0.02|0.03|-0.05|\\n|Norms+Fem(SE)|0.01|0.01|0.01|0.01|0.01|0.01|0.01|0.01|\\n|Norms+Fem(P-value)|0.00|0.34|0.00|0.03|0.91|0.04|0.05|0.00|\\n|Sample|ALL|ALL|ALL|ALL|ALL|ALL|ALL|ALL|\\n|CountryFE|YES|YES|YES|YES|YES|YES|YES|YES|\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [40]}}, {\"page\": 41, \"chunk\": 41, \"input_text\": \"Table 4b: Decision Making Power over Large Purchases - Correlation with Male Breadwinner Norm: Attitudes and Norms\\n\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|\\n|---|---|---|---|---|---|---|---|---|\\n||Female in HH<br>has all the<br>power<br>(Yes=1/No=0)|Male in HH has<br>all the power<br>(Yes=1/No=0)|Joint decision<br>making with<br>spouse<br>(Yes=1/No=0)|No power at all<br>(Yes=1/No=0)|Female in HH<br>has all the<br>power<br>(Yes=1/No=0)|Male in HH has<br>all the power<br>(Yes=1/No=0)|Joint decision<br>making with<br>spouse<br>(Yes=1/No=0)|No power at all<br>(Yes=1/No=0)|\\n|||**_Male Breadw_**|**_inner Norm: Ex_**|**_penses are a ma_**|**_n's responsibility _**|**_even if his wife ca_**|**_n help him_**||\\n|Female|0.12***|-0.08***|-0.03***|-0.01***|0.11***|-0.09***|-0.02**|-0.02***|\\n||[0.00]|[0.00]|[0.01]|[0.00]|[0.01]|[0.01]|[0.01]|[0.01]|\\n|Personal Belief:Agree(No=0/Yes=1)|0.01***|0.09***|-0.08***|-0.01**|||||\\n||[0.00]|[0.01]|[0.01]|[0.01]|||||\\n|Personal Belief:Agree*Female|0.05***|-0.07***|0.02**|0.05***|||||\\n||[0.01]|[0.01]|[0.01]|[0.01]|||||\\n|Community Norms (0/1)|||||-0.01|0.03***|-0.03***|-0.01|\\n||||||[0.01]|[0.01]|[0.01]|[0.01]|\\n|Community Norms (0/1)*Female|||||0.04***|-0.03**|0.00|0.03**|\\n||||||[0.01]|[0.01]|[0.02]|[0.01]|\\n|Constant|0.24***|0.46***|0.25***|0.03|0.22***|0.44***|0.26***|0.03|\\n||[0.02]|[0.03]|[0.03]|[0.03]|[0.02]|[0.03]|[0.04]|[0.04]|\\n|Observations|43053|43053|43053|43053|36433|36433|36433|36433|\\n|Adjusted R-squared|0.06|0.11|0.20|0.20|0.06|0.09|0.20|0.21|\\n|Mean Female|0.17|0.09|0.55|0.19|0.17|0.08|0.56|0.19|\\n|Mean Male|0.07|0.22|0.50|0.18|0.07|0.21|0.52|0.18|\\n|Norms+Fem(Coeff)|0.06|0.02|-0.06|0.04|0.03|0.01|-0.03|0.02|\\n|Norms+Fem(SE)|0.01|0.01|0.01|0.01|0.01|0.01|0.01|0.01|\\n|Norms+Fem(P-value)|0.00|0.00|0.00|0.00|0.00|0.20|0.01|0.02|\\n|Sample|ALL|ALL|ALL|ALL|ALL|ALL|ALL|ALL|\\n|CountryFE|YES|YES|YES|YES|YES|YES|YES|YES|\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [41]}}, {\"page\": 42, \"chunk\": 42, \"input_text\": \"Table 5: Decision Making Power over Large Purchases - Correlation with Perceived Norms, conditional on Personal Beliefs\\n\\n||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|\\n|---|---|---|---|---|---|---|---|---|\\n||Female in HH<br>has all the<br>power<br>(Yes=1/No=0)|Male in HH has<br>all the power<br>(Yes=1/No=0)|Joint decision<br>making with<br>spouse<br>(Yes=1/No=0)|No power at all<br>(Yes=1/No=0)|Female in HH<br>has all the<br>power<br>(Yes=1/No=0)|Male in HH has<br>all the power<br>(Yes=1/No=0)|Joint decision<br>making with<br>spouse<br>(Yes=1/No=0)|No power at all<br>(Yes=1/No=0)|\\n|||**_Gender Equ_**|**_ality Norm_**|||**_Male Breadw_**|**_inner Norm_**||\\n|Female|0.15***|-0.15***|0.03|0.05***|0.11***|-0.08***|-0.03***|-0.02***|\\n||[0.02]|[0.02]|[0.02]|[0.02]|[0.01]|[0.01]|[0.01]|[0.01]|\\n|Personal Belief:Agree(No=0/Yes=1)|-0.01|-0.08***|0.08***|-0.00|0.02***|0.09***|-0.08***|-0.01**|\\n||[0.01]|[0.01]|[0.01]|[0.01]|[0.00]|[0.01]|[0.01]|[0.01]|\\n|Personal Belief:Agree*Female|-0.03*|0.06***|-0.05**|-0.03**|0.04***|-0.07***|0.02*|0.05***|\\n||[0.01]|[0.02]|[0.02]|[0.02]|[0.01]|[0.01]|[0.01]|[0.01]|\\n|Community Norms (0/1)|0.01|-0.05***|0.06***|-0.01|-0.01|0.00|-0.01|-0.00|\\n||[0.01]|[0.01]|[0.01]|[0.01]|[0.01]|[0.01]|[0.01]|[0.01]|\\n|Community Norms (0/1)*Female|-0.00|0.03**|-0.03|-0.04***|0.02**|0.00|-0.01|0.01|\\n||[0.01]|[0.02]|[0.02]|[0.02]|[0.01]|[0.01]|[0.02]|[0.01]|\\n|Constant|0.26***|0.54***|0.11**|0.00|0.22***|0.42***|0.28***|0.04|\\n||[0.03]|[0.04]|[0.05]|[0.04]|[0.02]|[0.03]|[0.04]|[0.04]|\\n|Observations|25183|25183|25183|25183|36433|36433|36433|36433|\\n|Adjusted R-squared|0.06|0.09|0.22|0.25|0.06|0.10|0.21|0.21|\\n|Mean Female|0.15|0.07|0.57|0.19|0.17|0.08|0.56|0.19|\\n|Mean Male|0.06|0.18|0.54|0.18|0.07|0.21|0.52|0.18|\\n|PersBelief+Fem(P-value)|0.01|0.03|0.02|0.01|0.00|0.00|0.00|0.00|\\n|ComNorms+Fem(P-value)|0.68|0.06|0.08|0.00|0.10|0.34|0.20|0.42|\\n|Mean VIF|4.50|4.50|4.50|4.50|3.21|3.21|3.21|3.21|\\n|Sample|ALL|ALL|ALL|ALL|ALL|ALL|ALL|ALL|\\n|CountryFE|YES|YES|YES|YES|YES|YES|YES|YES|\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [42]}}, {\"page\": 43, \"chunk\": 43, \"input_text\": \"Table 6a: Care and Domestic Responsibilities - Correlation with Gender Equality Norm: Attitudes and Norms\\n\\n||(1)|(2)|(3)|(4)|(5)|(6)|\\n|---|---|---|---|---|---|---|\\n||Main<br>Caregiver:<br>Preschool<br>or School<br>Children|Hours<br>spend on<br>care<br>activities<br>(Main<br>Caregiver<br>Children)|Hours<br>spend on<br>household<br>chores<br>(Main<br>Caregiver<br>Children)|Main<br>Caregiver:<br>Preschool<br>or School<br>Children|Hours<br>spend on<br>care<br>activities<br>(Main<br>Caregiver<br>Children)|Hours spend<br>on household<br>chores (Main<br>Caregiver<br>Children)|\\n||**_Gender Eq_**|**_uality Norm_**|**_: Men and w_**|**_omen shoul_**|**_d have equa_**|**_l opportunities_**|\\n|Female|0.07***|2.95***|1.52***|0.11***|1.89***|1.53***|\\n||[0.02]|[0.43]|[0.15]|[0.02]|[0.38]|[0.13]|\\n|Personal Belief:Agree(No=0/Yes=1)|0.00|-0.12|0.13||||\\n||[0.01]|[0.21]|[0.08]||||\\n|Personal Belief:Agree*Female|0.01|-0.24|-0.32**||||\\n||[0.02]|[0.44]|[0.15]||||\\n|Community Norms (0/1)||||0.02|-0.07|0.35***|\\n|||||[0.02]|[0.36]|[0.12]|\\n|Community Norms (0/1)*Female||||-0.04|1.44***|-0.44**|\\n|||||[0.03]|[0.55]|[0.19]|\\n|Constant|0.31***|5.00***|3.84***|0.29***|4.09***|4.11***|\\n||[0.05]|[0.95]|[0.35]|[0.07]|[0.98]|[0.50]|\\n|Observations|22327|9564|10131|15506|6820|7162|\\n|Adjusted R-squared|0.21|0.15|0.17|0.23|0.17|0.18|\\n|Mean Female|0.47|9.41|4.26|0.48|9.52|4.26|\\n|Mean Male|0.41|6.20|2.94|0.42|6.17|2.93|\\n|Norms+Fem(Coeff)|0.02|-0.36|-0.20|-0.03|1.36|-0.08|\\n|Norms+Fem(SE)|0.02|0.39|0.13|0.02|0.43|0.15|\\n|Norms+Fem(P-value)|0.31|0.36|0.14|0.17|0.00|0.56|\\n|Sample|ALL|ALL|ALL|ALL|ALL|ALL|\\n|CountryFE|YES|YES|YES|YES|YES|YES|\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [43]}}, {\"page\": 44, \"chunk\": 44, \"input_text\": \"Table 6b: Care and Domestic Responsibilities - Correlation with Female Caregiver Norm: Attitudes and Norms\\n\\n||(1)|(2)|(3)|(4)|(5)|(6)|\\n|---|---|---|---|---|---|---|\\n||Main<br>Caregiver:Presch<br>ool or School<br>Children|Hours spend on<br>care activities<br>(Main Caregiver<br>Children)|<br>Hours spend on<br>household chores<br>(Main Caregiver<br>Children)|<br>Main<br>Caregiver:Presch<br>ool or School<br>Children|Hours spend on<br>care activities<br>(Main Caregiver<br>Children)|<br>Hours spend on<br>household chores<br>(Main Caregiver<br>Children)|\\n||**_Female Ca_**|**_regiver Norm: W_**|**_oman's most imp_**|**_ortant role is to ta_**|**_ke care of home_**|**_& children_**|\\n|Female|0.07***|2.40***|1.12***|0.05***|2.27***|1.16***|\\n||[0.01]|[0.18]|[0.06]|[0.01]|[0.33]|[0.11]|\\n|Personal Belief:Agree(No=0/Yes=1)|0.03***|0.41**|0.14**||||\\n||[0.01]|[0.16]|[0.06]||||\\n|Personal Belief:Agree*Female|0.03**|0.64**|0.22**||||\\n||[0.01]|[0.25]|[0.09]||||\\n|Community Norms (0/1)||||0.00|1.06***|0.43***|\\n|||||[0.02]|[0.29]|[0.10]|\\n|Community Norms (0/1)*Female||||0.04**|0.63|0.07|\\n|||||[0.02]|[0.47]|[0.16]|\\n|Constant|0.29***|4.59***|3.79***|0.28***|3.41***|3.54***|\\n||[0.05]|[0.94]|[0.34]|[0.05]|[0.84]|[0.35]|\\n|Observations|22381|9592|10165|21211|9180|9670|\\n|Adjusted R-squared|0.22|0.15|0.17|0.22|0.16|0.17|\\n|Mean Female|0.47|9.40|4.26|0.47|9.42|4.27|\\n|Mean Male|0.41|6.22|2.94|0.42|6.20|2.94|\\n|Norms+Fem(Coeff)|0.06|1.05|0.36|0.05|1.69|0.51|\\n|Norms+Fem(SE)|0.01|0.20|0.07|0.02|0.37|0.12|\\n|Norms+Fem(P-value)|0.00|0.00|0.00|0.00|0.00|0.00|\\n|Sample|ALL|ALL|ALL|ALL|ALL|ALL|\\n|CountryFE|YES|YES|YES|YES|YES|YES|\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [44]}}, {\"page\": 45, \"chunk\": 45, \"input_text\": \"Table 7: Care and Domestic Responsibilities - Correlation with Perceived Norms, conditional on Personal Beliefs\\n\\n||(1)|(2)|(3)|(4)|(5)|(6)|\\n|---|---|---|---|---|---|---|\\n||Main<br>Caregiver:<br>Preschool<br>or School<br>Children|Hours<br>spend on<br>care<br>activities<br>(Main<br>Caregiver<br>Children)|Hours<br>spend on<br>household<br>chores<br>(Main<br>Caregiver<br>Children)|Main<br>Caregiver:<br>Preschool<br>or School<br>Children|Hours<br>spend on<br>care<br>activities<br>(Main<br>Caregiver<br>Children)|Hours<br>spend on<br>household<br>chores<br>(Main<br>Caregiver<br>Children)|\\n||**_Gend_**|**_er Equality _**|**_Norm_**|**_Femal_**|**_e Caregiver_**|**_Norm_**|\\n|Female|0.11***|2.21***|1.77***|0.05***|2.17***|1.13***|\\n||[0.03]|[0.58]|[0.21]|[0.01]|[0.33]|[0.11]|\\n|Personal Belief:Agree(No=0/Yes=1)|0.01|-0.21|0.07|0.03***|0.34*|0.09|\\n||[0.02]|[0.27]|[0.10]|[0.01]|[0.18]|[0.06]|\\n|Personal Belief:Agree*Female|-0.00|-0.31|-0.30|0.02*|0.49*|0.21**|\\n||[0.02]|[0.53]|[0.19]|[0.01]|[0.27]|[0.09]|\\n|Community Norms (0/1)|0.01|0.00|0.33**|-0.01|0.91***|0.39***|\\n||[0.02]|[0.38]|[0.13]|[0.02]|[0.30]|[0.11]|\\n|Community Norms (0/1)*Female|-0.04|1.41**|-0.39**|0.04*|0.46|-0.01|\\n||[0.03]|[0.56]|[0.19]|[0.02]|[0.49]|[0.17]|\\n|Constant|0.29***|4.24***|4.07***|0.27***|3.23***|3.49***|\\n||[0.07]|[0.99]|[0.51]|[0.05]|[0.85]|[0.35]|\\n|Observations|15506|6820|7162|21211|9180|9670|\\n|Adjusted R-squared|0.23|0.17|0.18|0.22|0.16|0.17|\\n|Mean Female|0.48|9.52|4.26|0.47|9.42|4.27|\\n|Mean Male|0.42|6.17|2.93|0.42|6.20|2.94|\\n|PersBelief+Fem(P-value)|0.80|0.26|0.16|0.00|0.00|0.00|\\n|ComNorms+Fem(P-value)|0.17|0.00|0.66|0.07|0.00|0.00|\\n|Mean VIF|3.65|4.79|4.62|2.60|3.80|3.65|\\n|Sample|ALL|ALL|ALL|ALL|ALL|ALL|\\n|CountryFE|YES|YES|YES|YES|YES|YES|\", \"datasets\": [{\"mention_name\": {\"text\": \"Observations\", \"confidence\": 0.60918128490448, \"start\": 1332, \"end\": 1344, \"clean_text\": \"Observations\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1332, \"end\": 1344, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1332, \"end\": 1344, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1332, \"end\": 1344}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1332, \"end\": 1344, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.2574390769004822, \"start\": 198, \"end\": 210}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.60918128490448, \"start\": 198, \"end\": 210}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.585983395576477, \"start\": 198, \"end\": 210}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.585983395576477, \"start\": 198, \"end\": 210}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099821402012437374/pdf/IDU1bf06cd191699f1462f18236168ed5e5295bf.pdf\", \"pages\": [45]}}]}"}}},{"rowIdx":97,"cells":{"doc_id":{"kind":"string","value":"34252495"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34252495\", \"last_modified_date\": \"2024-04-14T00:00:00Z\", \"authors\": {\"0\": {\"author\": \"Hauser,Christina Sarah\"}}, \"count\": \"Tunisia\", \"count_key\": \"82575\", \"docna\": {\"0\": {\"docna\": \"Tackling Gender Discriminatory Inheritance\\n Law Privately : Lessons from a Survey Experiment in Tunisia\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"Office of the Chief Economist (MNACE)\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34252495\"}, \"repnb\": \"WPS10693\", \"docdt\": \"2024-02-06T00:00:00Z\", \"datestored\": \"2024-02-05T06:04:43Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"When reform of gender discriminatory law\\n fails, individual action can offer a second-best solution.\\n As most Muslim-majority countries, Tunisia applies Islamic\\n inheritance law, systematically favoring sons over\\n daughters. By making gifts to their daughter, parents can\\n privately attenuate gender discrimination in inheritance.\\n This study investigates to what extent gifting can represent\\n an alternative to legal reform and for whom. Within a\\n randomized experiment, this study tests whether providing\\n information on public support for inheritance law reform\\n and/or the possibility to make a gift to one’s\\n daughter has a causal impact on individual attitudes towards\\n women’s right to inheritance. The overall\\n evidence on the effectiveness of the proposed informational\\n treatments to encourage gifting is mixed. However, approval\\n of gifting daughters is high—especially among the\\n wealthy. Men are more likely to gift than women. By\\n contrast, demand for legal reform is significantly higher\\n among women and individuals with low educational attainment.\\n The findings thus suggest that gifting indeed represents an\\n alternative to legal reform; but mostly for a relatively\\n well-off subset of the population, leaving the agency to the\\n traditionally male head of the family.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10693; MNACE labor and gender research programs; PEOPLE\", \"display_title\": \"Tackling Gender Discriminatory\\n Inheritance Law Privately : Lessons from a Survey Experiment\\n in Tunisia\", \"disclosure_date\": \"2024-02-05T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099804402052432900/text/IDU16715ac0d1906814580188631d44bf194b848.txt\", \"listing_relative_url\": \"/research/2024/02/34252495/tackling-gender-discriminatory-inheritance-law-privately-lessons-survey-experiment-tunisia\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099804402052432900/IDU16715ac0d1906814580188631d44bf194b848\", \"new_url\": \"2024/02/34252495/IDU16715ac0d1906814580188631d44bf194b848\", \"ext_pub_date\": \"2024-02-05T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Political Economy of Reform m1326263 836,Islamic Law m1326357 2070,Gender and Economic Empowerment m1327622 911\", \"docm_id\": \"IDU-6715ac0d-9068-4580-8863-d44bf194b848\", \"chronical_docm_id\": \"IDU-6715ac0d-9068-4580-8863-d44bf194b848\", \"publishtoextweb_dt\": \"2024-02-05T00:00:00Z\", \"dois\": \"10.1596/41010\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"guid\": \"099804402052432900\", \"no_of_pages\": \"66\", \"author_keys\": {\"0\": {\"author_key\": \"2205624\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099804402052432900/IDU16715ac0d1906814580188631d44bf194b848\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10693 Tackling Gender Discriminatory Inheritance Law Privately Lessons from a Survey Experiment in Tunisia _Christina Sarah Hauser_ Middle East and North Africa Region Office of the Chief Economist February 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10693\\n\\n# **Abstract**\\n\\nWhen reform of gender discriminatory law fails, individual action can offer a second-best solution. As most Muslim-majority countries, Tunisia applies Islamic inheritance law, systematically favoring sons over daughters. By making gifts to their daughter, parents can privately attenuate gender discrimination in inheritance. This study investigates to what extent gifting can represent an alternative to legal reform and for whom. Within a randomized experiment, this study tests whether providing information on public support for inheritance law reform and/or the possibility to make a gift to one’s daughter has a causal impact on individual attitudes towards women’s right to inheritance. The overall evidence on the effectiveness of the proposed informational treatments to encourage gifting is mixed. However, approval of gifting daughters is high—especially among the wealthy. Men are more likely to gift than women. By contrast, demand for legal reform is significantly higher among women and individuals with low educational attainment. The findings thus suggest that gifting indeed represents an alternative to legal reform; but mostly for a relatively well-off subset of the population, leaving the agency to the traditionally male head of the family.\\n\\nThis paper is a product of the Office of the Chief Economist, Middle East and North Africa Region. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The author may be contacted at christina.hauser@eui.eu or christina.sarah.hauser@gmail.com.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"Tackling Gender Discriminatory Inheritance Law Privately: Lessons from a Survey Experiment in Tunisia Christina Sarah Hauser<sup>∗</sup> JEL classification: J16, N37, D31, K37 Keywords: gender discrimination, inheritance, Islamic law, information experiment\\n\\n> ∗Department of Economics, European University Institute (EUI), Via delle Fontanelle 18, 50014 Fiesole, Italy. Email: christina.hauser@eui.eu. I gratefully acknowledge the financial and analytical support from the Office of the Chief Economist for the Middle East and North Africa (MNACE; TTLs: Nelly Elmallakh and Nazmul Chaudhury) under the regional Labor and Gender Research Programs of the World Bank. This research also benefited from a LEAP (Laboratory for Effective Anti-Policy Action) Student Grant and an Early Stage Research Grant of the European University Institute. IRB approval was obtained from the Ethics Committee of the European University Institute in December 2022. The study was preregistered at the AEA RCT registry on 9 January 2023 (AEARCTR-0010343). I am indebted to One-to-One for Research and Polling for superb data collection and to Friedrich Ebert Stiftung for hosting me during my research stay in Tunis. I wish to thank Jihen Abbes, Sule Alan, Haroun Ben Youssef, Tom Crossley, Souleima Dhaya, Leyla Hassen, Zeineb Mazouz, Youssef Meddeb, Imen Mezlini, Dietlind Stolle, Alessandro Tarozzi, Abir Tayari and Weiwei Zheng for helpful comments and advice. Any mistakes are my own.\", \"datasets\": [{\"mention_name\": {\"text\": \"Survey Experiment in Tunisia\", \"confidence\": 0.3281811475753784, \"start\": 73, \"end\": 101, \"clean_text\": \"Survey Experiment in Tunisia\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 73, \"end\": 101, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 73, \"end\": 101, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 73, \"end\": 101}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 73, \"end\": 101, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9937441945075989, \"start\": 73, \"end\": 101}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3281811475753784, \"start\": 73, \"end\": 101}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5099883675575256, \"start\": 73, \"end\": 101}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5099883675575256, \"start\": 73, \"end\": 101}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **1 Introduction**\\n\\nThere is a large consensus in the literature that female empowerment improves women’s and children’s outcomes and plays a positive role in economic development overall (e.g. Duflo, 2003 and Allendorf, 2007). Women living in the Middle East and North Africa (MENA) region score lowest worldwide on most measures of economic empowerment, including access to property ownership (World Economic Forum, 2023). In the region, inheritance has historically been the primary way through which women access property and it plays a decisive role in allowing women to participate in the economy, e.g. by enabling them to start their own businesses (IFC, 2017). Property ownership can insure women against unforeseen misfortunes and increase their bargaining power within the household. However, most MENA countries apply Islamic inheritance law, which systematically favors sons over daughters. The bequest foreseen for an individual’s children is split unevenly by gender: a son always receives twice the share of a daughter. This arrangement is commonly justified with a man’s obligation to take care of his wife, daughter or sister. Social reality in the MENA region, however, is changing. In Tunisia, women represent more than half of the college student population (World Bank, 2022b) and marry increasingly late or not at all (Assaad et al., 2017). The “gender gap” in education has reversed: in the most recent panel survey, 71 percent of Tunisian women were as educated as or even more highly educated than their husbands (Assaad et al., 2017). Even though female labor market participation remains relatively low, household surveys suggest that women contribute a significant share to household budgets (ERF, 2014). The persistence of gender discriminatory inheritance law has been associated with women’s precarization not only in Tunisia, but also in Morocco and Egypt, and particularly when it comes to land bequests: if not entirely left out, women tend to receive the least fertile lands (Tanner, 2020; Conseil National des Droits de l’Homme, 2015; Khodary, 2018). Gender discrimination in inheritance is thus an urgent but so far understudied policy issue in the MENA region and beyond.\\n\\nTunisia was the first MENA country where a large social movement demanded gender equal inheritance law. The reform proposal received presidential support in 2018, but it never reached a parliamentary vote. The legal status quo remains unchanged and the introduction of a gender equal inheritance law seems increasingly unlikely. But legal reform may not be the only way to weaken gender discrimination in inheritance. The Tunisian Code on Personal Status allows parents to make gifts to a person of their choice. If parents gift assets, property or land to their daughter while alive, the remaining wealth is split according 2\", \"datasets\": [{\"mention_name\": {\"text\": \"panel survey\", \"confidence\": 0.4846443235874176, \"start\": 1428, \"end\": 1440, \"clean_text\": \"panel survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1428, \"end\": 1440, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1428, \"end\": 1440, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1428, \"end\": 1440}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1428, \"end\": 1440, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9954159259796143, \"start\": 643, \"end\": 655}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4846443235874176, \"start\": 643, \"end\": 655}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7808733582496643, \"start\": 643, \"end\": 655}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7808733582496643, \"start\": 643, \"end\": 655}}, {\"mention_name\": {\"text\": \"household surveys\", \"confidence\": 0.6940576434135437, \"start\": 1633, \"end\": 1650, \"clean_text\": \"household surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1633, \"end\": 1650, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1633, \"end\": 1650, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1633, \"end\": 1650}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1633, \"end\": 1650, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9824365377426147, \"start\": 848, \"end\": 865}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6940576434135437, \"start\": 848, \"end\": 865}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.49479761719703674, \"start\": 848, \"end\": 865}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.49479761719703674, \"start\": 848, \"end\": 865}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"to Islamic inheritance law after their death - but the gender discrimination in the final wealth allocation is attenuated. Collecting novel survey data, I first provide correlational and descriptive evidence to assess to what extent gifting represents a second-best alternative to legal reform. Second, within a survey experiment I test whether providing information on public support for inheritance law reform and/or the possibility to gift one’s daughter has a causal impact on attitudes regarding women’s inheritance.\\n\\nThe primary data necessary for answering these research questions was collected in Tunisia between February and April 2023 via the phone. The total sample size in the first survey round was N=1,505. Respondents were randomly allocated to control (C), or to two different treatment arms where they received information on gifting (T1) or information on public reform support and gifting (T2). The second treatment T2 included a belief correction: respondents were told whether they underestimated, overestimated or correctly guessed public support for the introduction of gender equal inheritance law. I test the effect of the provided information on a set of inheritance-related gender attitudes, which are measured in a first endline directly following the treatment and in a second endline about four weeks later. I hypothesized that both treatments would increase respondents’ awareness of gender discrimination in inheritance and should thus lead to more gender equal attitudes. Yet, I conjectured that reform support may decrease in the treatment groups if gifting is viewed as a substitute for legal reform. Moreover, assuming that the perceived social norm functions as a constraint on individuals’ gender attitudes, the belief correction in T2 was expected to render the views of respondents who underestimate public reform support more gender progressive.\\n\\nThe observed treatment effects are mixed: T1 significantly increases respondents’ likelihood to favor a gender equal bequest allocation. Treatment effects on the other main outcomes - reform support and willingness to use gifting to achieve a more gender equitable bequest allocation - are positive but shy of statistical significance. Furthermore, these effects largely disappear when the gifting information is paired with reform information in T2. Firstand second-order beliefs are strongly correlated: underestimating the public’s support for legal reform is the best predictor of an individual’s own gender attitudes, including support for inheritance law reform.<sup>1</sup> Almost 60 percent of the respondents in the survey underestimate\\n\\n> 1Here and in the following I use the term “first-order belief” to refer to an individual’s own beliefs. i.e. “I am favorable to introducing gender equal inheritance law.” I use the term “second-order belief” to refer to an individual’s perception of what others believe. i.e. “I think that the majority of Tunisians are opposed to introducing gender equal inheritance law.”\\n\\n3\", \"datasets\": [{\"mention_name\": {\"text\": \"survey data\", \"confidence\": 0.8426319360733032, \"start\": 140, \"end\": 151, \"clean_text\": \"survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 140, \"end\": 151, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 140, \"end\": 151, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 140, \"end\": 151}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 140, \"end\": 151, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9944146871566772, \"start\": 140, \"end\": 151}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8426319360733032, \"start\": 140, \"end\": 151}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.733850359916687, \"start\": 140, \"end\": 151}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.733850359916687, \"start\": 140, \"end\": 151}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"public reform support and I find that individuals who do so are 29 percentage points less likely to support legal reform than individuals who hold correct second-order beliefs. Yet, the belief correction in T2 does not yield the results one would expect if individuals felt constrained by the belief that others disapprove of reforming the legal status quo: respondents who underestimate (overestimate) reform support show positive (negative) treatment effects on all three outcomes, but they are not statistically significant. By contrast, treated respondents who hold _correct_ beliefs are significantly less likely to choose a gender equal bequest and to support reform. I argue that, in contrast to what has been found by the previous literature with respect to beliefs about women’s labor market participation (Bursztyn et al., 2020) and political participation (Gulzar et al., 2020), the perceived social norm is unlikely to be the binding constraint in this context.\\n\\nI propose two complementary ways of reconciling these findings theoretically. First, I show that respondents who underestimate (overestimate) public reform support are generally less (more) likely to endorse gender progressive attitudes themselves. Bursztyn et al. (2020) and several other studies find that second-order beliefs causally affect first-order beliefs; the perceived social norm shapes individual preferences. However, I argue that the direction of the causality is not clear. It is also possible that first-order beliefs shape second-order beliefs - individual preferences determine what is perceived to be the social norm. In the context of this study, respondents may believe that the public hesitates to support gender equality in inheritance _because_ they themselves do not support it. If this is the case, correcting secondorder beliefs would not necessarily lead to a revision of first-order beliefs. Second, based on previous literature in related fields, I argue that this unexpected treatment effect pattern may be explained by the underlying characteristics of respondents, i.e. respondents who underestimate (overestimate) public reform support may differ from respondents who hold correct perceptions on substantive grounds.\\n\\nIn the control group, respondents who underestimate (overestimate) public reform support are the least (most) likely to hold progressive gender attitudes. Respondents who underestimate public reform support appear to be part of the educated and wealthy elite of the country. By contrast, respondents who guess correctly appear to situate themselves somewhere closer to the middle of the opinion spectrum. They make extensive use of social media and are well informed about the Tunisian feminist movement. I argue that they are likely to be the political moderates whose attitudes are malleable. This finding relates to a growing literature which assesses the effectiveness of information to shape political opinions. Alesina et al. (2018b), Alesina et al. (2018c) and Haaland and Roth (2023) for instance 4\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"have found that individuals who hold strong political beliefs are less likely to respond to informational treatments. The finding that respondents who hold correct beliefs react more strongly to treatment may thus be due to their relatively more moderate political opinions and their interest in social issues. From a policy perspective, this is a highly relevant finding as these are the respondents who should be targeted for policy interventions. However, it also implies that belief correction is unlikely to be a policy solution in this context. The group of “political moderates” is small. Hence, even though a majority of respondents underestimates public reform support, correcting this misperception does not appear to be an efficient treatment. Future research should investigate alternative strategies.\\n\\nBeyond the experimental results, this study presents new descriptive and correlational evidence on demand for gender equality in inheritance in a context where Islamic inheritance law continues to be applied. Overall, support for the introduction of a secular gender equal inheritance law is low: only about one in three respondents favors legal reform. This stands in a striking contrast with approval of gifting: two in three respondents report that they are willing to gift their daughter to attenuate gender discrimination in inheritance. The descriptive data also confirms that gifting is fairly wide-spread in Tunisia: in the surveyed sample, 27 percent of respondents have first-hand experience with gifting, meaning that they have received a gift themselves, are married to someone who has received a gift, or they have gifted a daughter or female relative (see table 3).\\n\\nHowever, this does not mean that gifting is a readily available alternative to legal reform for the whole population the population. I show that older age cohorts are more likely to favor gifting over reform, as are the most educated, wealthy and religious individuals. Conversely, demand for legal reform is significantly higher among women and individuals with low educational attainment. Taken together, these findings suggest that gifting indeed represents a socially acceptable alternative to legal reform of gender discriminatory inheritance law in Tunisia - but mostly for a relatively well-off subsample of the population.\\n\\nMy research relates to a larger literature on gender discrimination and inheritance, which has shown that well-intended legal reform may fail to empower women in the presence of regressive gender norms (Rosenblum, 2015; Roy, 2015 and Bhalotra et al., 2020). Changing gender discriminatory social norms is therefore a prerequiste for successful legal reform and sustainable development (Sunstein, 1996; Bertrand, 2020). This is likely to apply to the context of gender discrimination in wealth transmission: encouraging the practice of gifting one’s daughter could theoretically affect social norms around women’s access to property, 5\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"decreasing the cost of holding progressive gender attitudes and preparing the grounds for successful legal change in the long run (Banerjee, 1992; Mehmood et al., 2022). The descriptive evidence this study provides clearly shows that women in practice rarely make gifts. This implies that potential interventions aiming to encourage the use of gifting should focus on male family heads. However, men tend to hold more conservative gender norms than women, and the data shows that both sons and daughters receive gifts. A policy intervention must therefore be carefully designed and take into account that women until now only often do not dispose of the necessary asset ownership or bargaining power to make gifts. Considering these caveats, gifting is certainly not a panacea to gender discriminatory inheritance law and practices. It can be a viable second-best solution for a wealthy subset of the population with progressive gender views, though.\\n\\nTo the best of my knowledge, this is the first study to quantify the demand for gender equal inheritance arrangements in a context where Islamic inheritance law continues being applied. Furthermore, it is the first to assess the potential tradeoff between the introduction of secular gender equal inheritance law and gifting one’s daughter. The descriptive and correlational evidence this study provides suggests that gifting is fairly widespread and, acknowledging the caveats mentioned above, can allow to strengthen the economic empowerment of women from comparably wealthy and gender progressive families. Section 2 provides more background information on the related literature and the institutional setting. Section 3 offers a detailed description of the data collection and experimental design, as well as my conceptual framework and empirical strategy. Section 4 discusses descriptive statistics, regression results, robustness checks and limitations of the study. Section 5 concludes.\\n\\n# **2 Background Information**\\n\\n## **2.1 Related Literature**\\n\\nThis study tests the effectiveness of information on gifting and reform support for changing gender norms. While this is to the best of my knowledge the first study to investigate this particular question, there is a growing body of literature using informational experiments. These experiments create exogenous variation in how people perceive their environment, allowing researchers to gain a better understanding of individual choices and preferences. Information provision experiments have for instance been used to analyze individual demand 6\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"for redistribution (e.g. Alesina et al., 2018c or Fehr et al., 2020, Settele, 2022), xenophobic and discriminatory behavior (e.g. Alesina et al., 2018a or Grigorieff et al., 2020) or health choices (e.g. Dupas, 2011, Barari et al., 2020). A growing number of studies use informational experiments to provide a better understanding of gender norms: Bursztyn et al. (2020) show that correcting misperceived gender norms can increase female labor market participation in Saudi Arabia. Grewenig et al. (2020a) show that prescriptive gender norms have a causal impact on adolescents’ expected labor market participation in Germany. Settele (2022) uses divergent information on the gender wage gap in the US to analyze how beliefs drive the individual demand for affirmative action.\\n\\nA large literature in economics theorizes and analyzes individuals’ bequest behavior in societies where gender equality between siblings is the legal norm. A smaller literature has considered the trade-off between making gifts and bequests. Lundholm and Ohlsson (2000)’s model shows that it is optimal for parents to make inter-vivo gifts to compensate their less well-off children in the US context where equal bequest allocation is the social norm. According to their assumption, bequests are public information while gifts are private information. By making gifts, parents can deviate from the social norm of making equal transfers without negative reputational effects.\\n\\nWithin a vignette experiment run online in Germany, Tisch and Gutfleisch (2022) assess the principles which guide individuals’ allocation of monetary gifts among siblings of different sexes. The dominating principle is gender equality. However, they also find evidence for the “equity” and “need” principles: sons _and_ daughters are assigned larger monetary amounts when they help their parents or when they are unemployed. There is some evidence of gender discriminatory preferences, though: the monetary reward for helping their parents appears to be higher for sons. Conversely, unemployed daughters receive more extra money than unemployed sons.\\n\\nThis paper also relates to a larger literature on the importance of law for development. The economics literature provides evidence that equal inheritance spurs long-term growth and prosperity (Bartels et al., 2020). While a lot of the recent literature focuses on social norms, Htun et al. (2019) argue that it is _legal_ restrictions to women’s agency which best predict women’s economic empowerment, measured as the share of women who have their own bank account, own firms and participate in the labor force. Gonzales et al. (2015) argue that legal gender discrimination is positively related to large gender gaps in labor market participation.\\n\\n7\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"In contexts where legal reform has been enacted, observational studies can measure the causal impact of introducing gender equal inheritance law. Women in Western countries progressively gained property and inheritance rights in the 19<sup>_th_</sup> and 20<sup>_th_</sup> century. Shammas (1994) argues that the Married Women Property Acts which were progressively adopted by US states from 1839 improved the legal situation of women. According to the historian, the percentage of wealth owned by women in the US increased more between 1860 and 1890 than in the two centuries preceding the acts. The quantiative evidence on the effectiveness of legal reform is at best mixed, though: the experience of India illustrates that policy makers’ attempt to increase women’s access to inheritance may backfire when restrictive gender norms prevail. Using a difference-in-differences approach Rosenblum (2015) shows that granting daughters the right to inherit ancestral lands in five Indian states increased the mortality of girls - the law seems to have incentivized parents to invest less in the health of their daughters. Bhalotra et al. (2020) confirm that the introduction of gender equal inheritance rights led to higher female infant mortality and an increase in the use of sonbiased fertility stopping rules in India. Roy (2015) argues that the reform at best remained ineffective - in reality, daughters’ access to ancestral land did not increase. By contrast, families compensated them with higher dowries or increased investment in their daughters’ education.\\n\\nGender discrimination in wealth transmission may persist even where the law is gender equal: Bessi`ere and Gollac (2023) show that gender discrimination in wealth transmission is commonplace in the US and Europe and has even increased in France over the past two decades. Inter-vivo gifts can be one way for families to favor sons. Moreover, land, real estate property or family businesses are most commonly passed on to (first-born) sons with daughters being paid out “their” share. According to Bessi`ere and Gollac (2023), notaries often deliberately undervalue these assets so they can stay in the family - leading to a discrimination of the children, most often daughters and younger sons, who are paid out. Social norms and the practices of notaries and lawyers can thus contribute to the persistence of gender discrimination in wealth transmission despite gender equal inheritance law.\\n\\n# **2.2 Gender Issues in Tunisia**\\n\\nThe legal and social context of the MENA region and Tunisia in particular differs from the South Asian context, which has been researched much more extensively: there are no reports of gender-biased abortions or the use of son-based stopping rules in the region. Historically, 8\", \"datasets\": [{\"mention_name\": {\"text\": \"observational studies\", \"confidence\": 0.31712400913238525, \"start\": 49, \"end\": 70, \"clean_text\": \"observational studies\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 49, \"end\": 70, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 49, \"end\": 70, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 49, \"end\": 70}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 49, \"end\": 70, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.6566693782806396, \"start\": 49, \"end\": 70}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.31712400913238525, \"start\": 49, \"end\": 70}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8096276521682739, \"start\": 49, \"end\": 70}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8096276521682739, \"start\": 49, \"end\": 70}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"Tunisia is known for being the regional front runner with respect to women’s rights: the country abolished repudiation and polygamy upon the passage of its 1956 civil code (LatteAbdallah, 2012). Abortion was legalized in 1973, earlier than in France and Italy. Women have outnumbered men in higher education since the late 1990s (Assaad et al., 2017 and World Bank, 2022a) and have reached top positions in the public and private sector. In 2021, Nejla Bouden became the first female prime minister of the Arab world (Terriennes, 2021). Yet, gender discrimination remains problematic, especially when it comes to women’s labor market participation and property rights. Policy makers and human rights activists in Tunisia and the larger MENA region have argued that legal dispossession increases women’s vulnerability and hinders sustainable development. Women often cede the share of the bequest which they are legally entitled to male family members (see Conseil National des Droits de l’Homme, 2015) for the case of Morocco). According to a mixed method study conducted by Khodary (2018) in Egypt, this is especially common when the inheritance concerns agricultural land. The author argues that improving women’s access to inheritance is likely to have a very positive impact on development outcomes, but her results are mostly correlational.\\n\\nIn 2000, the Tunisian women’s rights advocate Yosra Frawes started a first unsuccessful petition in favor of gender equality in inheritance (Dahmani, 2018). The Tunisian Revolution ended the regime of Ben Ali and started a process of democratization in 2011. Article 21 of the 2014 constitution confirmed gender equality _before the law_ : “All citizens, male and female, have equal rights and duties, and are equal before the law without any discrimination” (Constitute Project, Org., 2014).<sup>2</sup> For progressive observers, President B ́eji Ca ̈ıd Essebsi’s creation of the COLIBE Committee in 2017 therefore seemed the logical continuation of the country’s reform efforts. Among other things, the Committee proposed an attenuated inheritance law reform according to which the family head would be allowed to bequeath a larger share to his male descendants if he expressed the wish to do so in a written testament. In the absence of explicit action, gender equality would be applied. Conservative voices, however, claimed that this proposal contradicted the first article of the constitution, which confers the status of state religion to Islam. A heated political discussion emerged and spilled over to other MENA countries where Tunisia has been perceived as a political role model (Constitute Project, Org., 2014, Blaise, 2019 & Zaireg, 2018). By contrast, the proposal for\\n\\n> 2Some Tunisian legal scholars contest that gender equality _in terms of citizenship_ as defined in the constitution of 2014 establishes the right to equal inheritance rights, one reason being that inheritance falls into the _private sphere_ , not the public sphere, to which the constitution applies (see for example Verdier„ 2020).\\n\\n9\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"a gender equal inheritance law was never voted upon in the Tunisian parliament. With the Covid-19 pandemic, Essebsi’s death and the enduring political crisis in Tunisia, the question disappeared from the political agenda. The current Tunisian president Ka ̈ıs Sa ̈ıed is fiercely opposed to inheritance law reform (Verdier, 2020) and the 2022 constitution maintains the gender discriminatory dispositions of Islamic inheritance law. Since the reforms of Atat ̈urk in Turkey in the 1920s, no MENA country has succeeded to reform inheritance law (Baderin, 2021) making Tunisia an interesting and particularly relevant case study.\\n\\n# **2.3 Institutional Setting**\\n\\nThis research focuses on gender discrimination and restrictive social norms in the context of Islamic inheritance law, a so far understudied setting. Muslim-majority countries have seen an extensive secularization in penal and in commercial law, but much less so in family law (Sait and Lim, 2008). Inheritance law is considered as one of the last strongholds of Islamic jurisprudence in North Africa, the Middle East and East Asia. Examples of other Muslimmajority countries applying Islamic inheritance law include Morocco (Zaireg, 2018), Egypt (Berger and Sonneveld, 2010), Sudan (K ̈ondgen, 2010), Jordan (Baderin, 2021), Saudia Arabia (van Eijk, 2010), the United Arab Emirates (Baderin, 2021), Iran (Mir-Hosseini, 2010), Pakistan (Lau, 2010), Bangladesh (Shoma, 2019), the Maldives (OECD Development Centre, 2019) and Indonesia (Otto, 2010). Out of the 171 countries analyzed by Htun et al. (2019) in their cross-country analysis of legal gender discrimination, almost 20 percent maintained gender discriminatory inheritance law between sons and daughters. This makes gender discrimination in inheritance the second most common form of legal discrimination that women face in the world, second only to the inability to confer their nationality to their spouse.\\n\\nIn a historical perspective, Islamic inheritance law was gender progressive, putting women in a more privileged position than in Western Europe where they often did not have a legal identity separate from that of their husband until the end of the 19<sup>_th_</sup> century. The fact that women receive a smaller inheritance share than men is generally justified with the argument that Islamic law obliges a husband to provide shelter, clothing, food and medical assistance to his wife (Sait and Lim, 2008).<sup>3</sup> Yet, the changing social reality in many Muslim-majority\\n\\n> 3There is no dowry or brideprice: Islamic law prescribes “mahr” which is a payment from the groom to the bride, not to her family. It is meant to allow the wife to survive if her husband were to die and it may have to be paid back in the case of divorce. However, over the years the practice of “mahr” has nearly\\n\\n10\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"countries increasingly questions the validity of this argument.\\n\\nInheritance is strictly regulated in Islam: the Qur’an and the Sunnah lay down precise rules for the division of bequests among the “designated heirs” of a deceased individual, that is, their surviving parents, spouse and children. These rules encompass the deceased’s monetary wealth, real estate property and land. All four legal schools in Islam uphold the “two-shares-one-share” rule for bequests from parents to their sons and daughters which this study focuses on.<sup>4</sup> Citing from the Qur’an, “Allah commands you regarding your children: the share of the male will be twice that of the female [...]” (Tanner; 2020). In a nutshell, the bequest allocated to a son must be double the size of that allocated to a daughter.\\n\\nThe precise bequest allocation will depend on the family composition. As an illustration, consider the case of a man who is survived only by his two children, one son and one daughter. According to Islamic inheritance law, the son will receive two-thirds and the daughter onethird of the total bequest. If the man were survived by three children, one son and two daughters, the son would receive one-half and each of the daughters one-fourth of the total bequest. The existence of other family members may affect the total size of the bequest allocated to the man’s children, but not the gender discriminatory split between siblings. Assume the man is survived by his wife and his two children, one son and one daughter. In this case, the wife will inherit one-eighth of the total bequest. The remaining seveneighths of the bequest will then be split between the son and the daughter according to the “two-shares-one-share” rule: the son will receive seven twelfths and the daughter will receive seven twenty-fourths of the bequest, half as much as the son. Thus, the composition of the surviving family may impact how much the children inherit, but it does not change the fact that a daughter receives half as much as a son.<sup>5</sup> Inheritance tax does not affect the gender disappeared in Tunisia. President Bourguiba led a secularization campaign since 1956. In the occasion of his own wedding in 1962, he only paid one “symbolic dinar” as mahr to comply with the Islamic law. Amendments to the Code of Personal Status have specified that the “mahr” being paid does not have to be of a sizable amount (Fournier, 2013).\\n\\n> 4Note that here and in the following, I focus on bequests from a parent to their children, which is one - if not the most - common form of inheritance, and the one that was targeted by the Tunisian reform proposal. Islamic law with respect to bequests _from_ children _to_ parents, for example, does not discriminate by gender, meaning that the mother and the father will receive equal shares of the bequest. I disregard Islamic inheritance law regarding other family relationships in this study in order to focus on the inheritance mechanism which is economically and sociologically the most relevant: the one from parents to children.\\n\\n> 5Maybe not very surprisingly, the sophistication of Islamic inheritance law stimulated some important contributions of medieval Islamic mathematicians to algebra (Gandz; 1938). Most notably, Al-Hassar, a mathematician specialized in inheritance law from nowadays Morocco, developed the notation for fractions which is still commonly used today (Høyrup; 2009). With the arrival of the Internet, online calculators have been developed to address the issue (Sait and Lim; 2008).\\n\\n11\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"# discriminatory bequest allocation.<sup>6</sup>\\n\\nTestaments can generally not be used to change the inheritance allocation among one’s children.<sup>7</sup> By contrast, if done while the parent is alive, gifting offers a way of modifying the final bequest allocation. Art. 200 XII of Tunisian Code on Personal Status establishes the procedure of “hiba” (English “gift” or “donation”). Individuals may make gifts while alive; i.e. they can favor their daughter(s) by transferring a property title, land or other assets to them. Upon the parent’s death, the remaining wealth will be split according to Islamic inheritance law. If done _pre-mortem_ , this is thus a procedure which can allow parents to arrange a more gender-equal bequest split among their children under the legal status quo. By significantly reducing property registration fees, the Tunisian Law 69 of October 28, 2006, even creates a fiscal benefit for gifts to one’s children.<sup>8</sup> According to Kuran (2011), Islamic societies have used gifting for centuries to avoid excessive wealth fragmentation. Anecdotal evidence suggested that most Tunisians are aware of gifting but may not view it as a means of tackling the gender discriminatory inheritance allocation imposed by the law. This survey experiment leverages gifting, an existing alternative to inheritance law reform, as part of an informational treatment. The next section provides a detailed description of the experimental design.\\n\\n> 6In Tunisia, inheritance taxation increases with the kinship _degree_ . i.e. sons and daughters will pay a 2.5 percent registration fee on the value of inherited goods. This fee increases to 5 percent for bequests received from siblings, 25 percent for bequests from uncles, aunts, nephews, nieces or cousins and 35 percent for bequests from individuals of an even higher kinship degree. Inheritance taxation does not differ by the gender of the heir.\\n\\n> 7The use of testaments in Islamic law is very limited. Islamic law allows the individual to freely dispose of up to one-third of their total wealth to bequeath it to another person or association. According to Art. 171 XI of the Tunisian Code on Personal Status this can be done using a written testament and if the amount to be bequeathed does not exceed one-third of the individual’s total wealth, the consent of the “designated heirs” (surviving parents, spouse and children) is not needed. However, a bequest to one of the designated heirs which will increase their share relative to that foreseen by the Qur’anic rules requires the consent of the other designated heirs. In practice, a will can therefore only be used to increase a daughter’s inheritance share if the other heirs, i.e. her brothers, agree.\\n\\n8Gifts from parents to children which are registered after this date are subject to a flat tax of 15 Tunisian dinars per act of registration. Previously, they had been subject to a proportional property registration tax just as bequests. The property registration fees for gifts and bequests coincide for transfers among individuals of a higher degree of kinship. i.e. as bequests, gifts from sisters or brothers will be subject to a registration fee of 2.5 percent of the value of the good, 5 percent for gifts from uncles, aunts, nephews, nieces or cousins and 35 percent for gifts from individuals of a higher degree of kinship.\\n\\n12\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"# **3 Data and Experimental Design**\\n\\n## **3.1 Data Collection and Sample**\\n\\nThis research is based on a phone survey, conducted between February and April 2023 and targeting a sample of 1,505 adults. The response rate was 35 percent. The survey duration was on average 37 minutes. Respondents did not receive any financial or other benefits for participating in the survey. The survey questions were translated to colloquial Tunisian Arabic and extensively piloted in January 2023. Figure 1 displays a timeline of the data collection process.\\n\\n**Figure 1:** Timeline\\n\\n<!-- Start of picture text --> First Second<br>endline endline<br>Dec. 2022 Jan. 2023 Feb. 2023 Mar. 2023 Apr. 2023<br>Data collection<br><!-- End of picture text --> The final sample was obtained by random digital dialing: phone numbers are generated randomly taking into account the country-specific area codes, which allows to reach a fairly representative sample of survey respondents (Stasny, 2001). This naturally generates a stratification by governorate and we furthermore stratified by gender to reach equal shares of female and male respondents in the first survey round.\\n\\n## **3.2 Experimental Design**\\n\\n## **3.2.1 Baseline Survey**\\n\\nFigure 2 provides an illustration of the experimental design. A baseline survey collected relevant socio-economic characteristics (gender, age, governorate of residence, number of siblings, marital status, number of children if any, educational attainment, employment status). The baseline survey also contained questions on respondents’ level of religious 13\", \"datasets\": [{\"mention_name\": {\"text\": \"phone survey\", \"confidence\": 0.6340382099151611, \"start\": 105, \"end\": 117, \"clean_text\": \"phone survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 105, \"end\": 117, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 105, \"end\": 117, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 105, \"end\": 117}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 105, \"end\": 117, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9995081424713135, \"start\": 105, \"end\": 117}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6340382099151611, \"start\": 105, \"end\": 117}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9673020839691162, \"start\": 105, \"end\": 117}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9673020839691162, \"start\": 105, \"end\": 117}}, {\"mention_name\": {\"text\": \"Baseline Survey\", \"confidence\": 0.5242633819580078, \"start\": 1195, \"end\": 1210, \"clean_text\": \"Baseline Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1195, \"end\": 1210, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1195, \"end\": 1210, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1195, \"end\": 1210}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1195, \"end\": 1210, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9972992539405823, \"start\": 511, \"end\": 526}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5242633819580078, \"start\": 511, \"end\": 526}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7001912593841553, \"start\": 511, \"end\": 526}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7001912593841553, \"start\": 511, \"end\": 526}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"conservatism,<sup>9</sup> use of social media, exposure to the local feminist movement,<sup>10</sup> as well as a subsample of questions from the Crowne-Marlowe social desirability scale (for robustness checks as in Dhar et al., 2022; Mehmood et al., 2022). The Online Appendix contains a full list of these items.\\n\\n# **3.2.2 Treatments**\\n\\nThe experiment had two treatment arms: T1 only provides information on gifting. T2 first corrects respondents’ beliefs about public reform support and then provides information on gifting. After the baseline survey, all respondents were asked about their beliefs regarding public support for inheritance law reform:\\n\\n**[Second-Order Beliefs - all respondents:]** Now let’s talk about the opinions of _other_ Tunisians:\\n\\n1. Out of 10 _women_ chosen at random from Tunisia, how many do you think believe that daughters should inherit as much as sons?\\n\\n2. Similarly, out of 10 _men_ [...]?\\n\\n3. In your opinion, how many of these _(answer)_ women would support a legal reform making gender equality in inheritance the legal norm in Tunisia?\\n\\n4. Similarly, how many of these _(answer)_ men [...]?\\n\\nAfter these questions, respondents were randomly allocated to T1, T2 and C in roughly equal shares. The interviews of respondents who were allocated to C directly continued with the outcome measures. Only individuals in T2 were informed about the true level of support for legal reform:\\n\\n> 9Respondents were asked whether the sale of pork and alcohol should be legal in Tunisia, whether banks should be allowed to charge interest, whether religious leaders such as sheiks and imams should have a say in Tunisian politics and whether Islamic inheritance law forms part of Tunisian national identity. I construct a measure of religious conservatism based on these items using principal component analysis. The individual items as well as the index measure are highly correlated with self-reported religiosity.\\n\\n> 10The baseline survey included questions asking respondents whether they knew Halima Jouini, Yosra Frawes and Bochra Belhaj Hmida, three prominent Tunisian feminists, #EnaZeda, the local offspring of #MeToo, and whether they had a family member or friend who identified as a feminist. All survey items are highly correlated among each other and I construct a measure of exposure to feminism with principal component analysis based on these items.\\n\\n14\", \"datasets\": [{\"mention_name\": {\"text\": \"Crowne-Marlowe social desirability scale\", \"confidence\": 0.8041977882385254, \"start\": 146, \"end\": 186, \"clean_text\": \"Crowne-Marlowe social desirability scale\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 146, \"end\": 186, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 146, \"end\": 186, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 146, \"end\": 186}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 146, \"end\": 186, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.7086930274963379, \"start\": 146, \"end\": 186}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8041977882385254, \"start\": 146, \"end\": 186}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.7567440271377563, \"start\": 146, \"end\": 186}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7567440271377563, \"start\": 146, \"end\": 186}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"**[Belief Correction - Only T2:]** According to your previous answers, overall you expect about -<sup>_a_</sup> out of 10 Tunisian men and women to support a legal reform making gender equality in inheritance the legal standard in Tunisia. Is this correct? _[Record_\\n\\n_respondent’s answer, allow for correction]_ _Answer:_ You indeed overestimated / underestimated this percentage / guessed correctly. A recent survey found that about 35% of Tunisian men and women support legal reform.<sup>_b_</sup>\\n\\n> _a_ Average of items 3. and 4, rounded to the next integer.\\n\\n> _b_ The stated percentage is based on a nationally representative study conducted by the researcher in Tunisia in March 2022. The question is specific and the respondents were not aware of the correct answer as the study results had not been published.\\n\\nDepending on their priors, this information may lead treated respondents to update their perceptions of public support for inheritance law reform upwards or downwards. The perceptions of respondents in T1 were not corrected.\\n\\nRespondents in both treatment groups received the following information on gifting.\\n\\n**[Gifting Info - T1 & T2:]** Did you however know that there are alternative ways of arranging a more gender equitable bequest without legal reform?\\n\\nGifting, as defined in art. 200 XII of the Code of Personal Status is a way of transferring assets to the person of one’s choice. By “gifting” their daughter, parents can arrange a more gender equitable inheritance split, and Law 69 of October 28, 2006, even creates a fiscal benefit for donations to one’s children by significantly reducing property registration fees.\\n\\n# **3.2.3 Endline Surveys and Outcomes**\\n\\nTwo endline surveys evaluated the causal effect of these treatments. The first directly followed the information treatment. The first main outcome asks respondents to choose their ideal bequest allocation in a hypothetical scenario. They were asked to allocate a bequest between two children, a son and a daughter. The most common answer was “two shares-oneshare”, which is the allocation prescribed by Islamic law. In this case, the “ideal bequest” 15\", \"datasets\": [{\"mention_name\": {\"text\": \"nationally representative study\", \"confidence\": 0.3270415961742401, \"start\": 607, \"end\": 638, \"clean_text\": \"nationally representative study\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 607, \"end\": 638, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 607, \"end\": 638, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 607, \"end\": 638}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 607, \"end\": 638, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.38622063398361206, \"start\": 607, \"end\": 638}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3270415961742401, \"start\": 607, \"end\": 638}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.568955659866333, \"start\": 607, \"end\": 638}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.568955659866333, \"start\": 607, \"end\": 638}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"variable is coded as “0”. Whenever respondents deviated to a “one-share-one-share” or 50-50 split, this variable was coded as “1”. This outcome tests the hypothesis that the treatment leads subjects to favor a more gender equitable split.\\n\\nThe two other main outcomes directly assess respondents’ support for legal reform and the use of gifting: respondents were asked whether they were personally in favor of inheritance law reform. Next, they were asked whether they would be willing to gift their daughter in order to achieve a more gender equal bequest allocation. These two outcomes were rated on a 4-point Likert scale. I code them as “1” whenever a respondent strongly or somewhat favors the gender equal option and “0” otherwise. My pre-analysis plan specified further outcomes which are dropped in the following analysis as they yield very similar results.\\n\\n# **Main Outcomes:**\\n\\n1. **Ideal Bequest:** Assuming that you can freely allocate a bequest between your two children, a son and a daughter, how would you split the bequest among them?\\n\\n2. **Reform Support:** “I am personally in favor of making gender equality in inheritance the legal norm in Tunisia.” _(4-point Likert-scale)_ 3. **Gifting:** “I would want my sons and daughters to inherit the same and if necessary, would use gifting to ensure this.” _(4-point Likert-scale)_ I collected a number of secondary outcomes to assess the prevalence of gifting in Tunisia: respondents were asked whether they or their spouse had received a gift and whether they had gifted their daughter or another female relative, if they did not have a daughter. Finally, respondents were asked how credible they find the information provided and whether they consent to participate in a follow-up survey.\\n\\nFour to eight weeks later, respondents who had given consent were recontacted.<sup>11</sup> The follow-up did not remind respondents of the initial survey or the information provided: the introduction was changed and several unrelated questions were included to limit associations with the initial survey. All main outcomes were measured again and I also included more general measures of gender norms. The enumerators moreover inquired whether respondents\\n\\n> 11Grewenig et al. (2020a), Grewenig et al. (2020b) and Settele (2022) conducted the second endline after two weeks. Two to four weeks distance to treatment are common choices in the experimental literature.\\n\\n16\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"<!-- Start of picture text --> Baseline and 2nd Order Beliefs<br>Belief a<br>Correction<br>Gifting Info ||Gifting Info<br>1st Endline<br>after 4 weeks: 2nd Endline<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"I would conclude that the effect of increasing the _salience_ of the topic dominates (Grewenig et al., 2020a): the gifting information nudges respondents to think more about women’s rights to inheritance and induces them to adopt more gender progressive views.\\n\\nAs stated in the pre-analysis plan, analysis of to what extent T1 and T2 differ in their effects was of exploratory nature. On the one hand, by revealing true reform support, T2 makes clear that legal reform is unlikely and private action is needed. This may reinforce the information about gifting and increase potential treatment effects. On the other hand, anecdotal evidence suggested that many Tunisians believe that there are more urgent issues to tackle than gender discrimination in inheritance. Many seemed tired of the debate on gender equal inheritance. I thus did not exclude that the belief correction in T2 could create a backlash reaction in some respondents.\\n\\n# **3.4 Heterogeneous Treatment Effects and Mechanisms**\\n\\nT2 exploits differences in respondents’ prior perceptions of reform support, allowing for extensive heterogeneity analysis. I use this treatment arm to assess to what extent the perceived social norm represents a constraint on individual attitudes. Do people become more favorable towards gender equal inheritance arrangements upon finding out that their peers are more gender progressive than they thought? Assuming that the perceived social norm causally influences individual attitudes, I expected the effects of T2 to increase in the degree of misperception: treated respondents who underestimate public support for legal reform should update their beliefs, realizing that the social norm is more progressive than they thought. They should thus be more likely to endorse a gender equitable inheritance arrangement and report a higher probability of using gifting to reach such an allocation than respondents in the control group who hold similar misperceptions. Such a finding would be consistent with information-based updating as documented in the previous literature (for example Bleemer and Zafar, 2018; Grewenig et al., 2020a and Bursztyn et al., 2020). Equivalently, I hypothesized that T2 would decrease demand of legal reform in respondents who initially overestimate public support because they learn that the public opinion is less progressive than they thought.\\n\\nHow the belief correction would affect the trade-off between legal reform and gifting is theoretically ambiguous. Let us take the perspective of a respondent who initially overestimates reform support. On the one hand, learning that the public is less supportive of legal 18\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"reform than one had thought may increase willingness to take private measures given that a legal reform is perceived as more unlikely. This channel would lead to an increase in willingness to gift among treated respondents who overestimated reform support. On the other, the belief correction may decrease willingness to take private measures if the individual is afraid of violating the social norm (as in Paler et al., 2018). The sign of the estimated coefficient on the interaction between treatment and overestimating reform support may reveal which of the two channels dominates. The treatment effect on individuals who misperceive the support for legal reform on the ideal bequest and gifting outcomes was thus unclear in my theoretical framework. I hypothesized that it would likely depend on factors such as religious conservatism and ability to resist social pressure.\\n\\nThe second endline was meant to evaluate whether potential treatment effects still persist several weeks later. It is less likely that treatment effects are driven by experimenter demand effects or numerical anchoring if they are still observed several weeks later (see Stantcheva, 2022). Additional more general measures of gender attitudes intended to assess whether there are “spill-over” effects of the treatment, making respondents revise their gender attitudes even with respect to other gender issues which are not directly related to inheritance.\\n\\n# **3.5 Empirical Strategy**\\n\\nI estimate the causal effect of the two treatments _Tj_ with _j ∈{_ 1 _,_ 2 _}_ using the following linear regression model: where _yi_ is the outcome variable measured for individual _i_ , _Xi_ a vector of individual-specific controls, and _α_ 1 the coefficient of interest. All outcomes are coded such that higher values imply more gender equal preferences.\\n\\nTo investigate heterogeneity by prior perceptions of reform support, I implement a procedure akin to Grewenig et al. (2020a). Recall that if allocated to T2, respondents will be told that they underestimated or overestimated public support or guessed correctly. I estimate _′ yi_ = _β_ 0+ _β_ 1 _T_ 2 _i_ + _β_ 2 _T_ 2 _i∗correcti_ + _β_ 3 _T_ 2 _i∗overesti_ + _β_ 4 _∗correcti_ + _β_ 5 _∗overesti_ + _δ Xi_ + _εi_ (2) where _correcti_ and _overesti_ are binary variables that switch on when respondents correctly 19\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"assess or overestimate the level of public support for inheritance law reform (which is 35 percent). The basegroup is respondents who underestimated reform support and _β_ 1 thus describes the average treatment effect on this group. _β_ 2 describes the additional treatment effect for respondents who held correct perceptions of public reform support. I expected that _β_ 1 _>_ 0, _β_ 2 _<_ 0 and _β_ 3 _<_ 0 - the treatment effect is larger for respondents who initially underestimated reform support than for the other groups. Consistent with social norms influencing private opinion formation, I expected that all else equal _β_ 4 _>_ 0, _β_ 5 _>_ 0 and _β_ 5 _> β_ 4: respondents who held correct perceptions or overestimated the public support for inheritance law reform are more likely to support gender equality in inheritance than those respondents who underestimated reform support.\\n\\nI furthermore investigate the correlates of support for legal reform and interest in gifting as well as having gifted a daughter using a logit model: where _yi_ is one of the three mentioned outcomes, _Xi_ a vector of individual-level controls and _εi_ an individual-specific error term. Going beyond what was specified in my pre-analysis plan, I estimate the same model using “favoring gifting over reform” as an outcome. This is a binary variable which is coded as “1” when a respondent is more favorable to gifting than to reform and “0” otherwise.\\n\\nThe treatment effect estimated based on the outcomes of the second endline would be biased if assignment to treatment and control predicts participation in the second survey. I therefore check for covariate balance across treatment and control group among those who participated in the second endline and then regress an indicator variable for participation in the follow-up on treatment indicators and covariates. I finally estimate the following model: where _SEi_ is an indicator variable for the observation being collected in the second endline. To test whether treatment effects persist over time, I will test the null hypothesis that _φ_ 1 + _φ_ 2 + _φ_ 3 = 0.\\n\\n20\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"# **4 Results**\\n\\n## **4.1 Descriptive Results**\\n\\n## **4.1.1 Baseline Balance**\\n\\nTable 1 provides summary statistics of covariate balance at baseline by treatment arm. The respondents in my sample are on average 42 years old. They have 1.9 children on average; 63 percent are married. About one-third of the sample lives in a rural area. Small imbalances are observed with respect to educational attainment. One in four respondents has completed higher education; the percentage is 29 percent in T1 and 23 percent in T2 and control. I control for educational attainment in all regressions and later introduce survey weights as a robustness check.\\n\\nOn average, 42 percent of my sample work full time and 21 percent are homemakers. The Online Appendix compares my sample’s geographical spread by governorate to the latest national statistics (Institut National de la Statistique; 2022a) and shows that it is highly geographically representative. It also compares the share of age cohorts (Institut National de la Statistique; 2022b): younger cohorts are overrepresented in my sample. In the robustness checks, I show that weighing down younger cohorts does not affect my conclusions.\\n\\n## **4.1.2 Second-Order Beliefs**\\n\\nThe belief that a majority of one’s peers opposes political change may affect individual opinion and behavior. For example, an individual may refrain from taking a particular action or expressing a political opinion because she thinks that society will disapprove of or sanction it. Generalized misperceptions about public opinion may therefore hinder social change and keep society captured in inefficient social equilibria (see Kuran, 1991). When studying socially contentious issues such as the reform of law of religious origin, paying attention to individuals’ perceptions of what is the public opinion - second-order beliefs - is thus crucial.<sup>12</sup> Figures 3, 4, 5 and 6 aggregate respondents’ second-order beliefs about their\\n\\n> 12I use the term “first-order belief” to refer to an individual’s own beliefs. i.e. “I am favorable to introducing gender equal inheritance law.” I use the term “second-order belief” to refer to an individual’s perception of what others believe. i.e. “I think that the majority of Tunisians are opposed to introducing gender equal inheritance law.” The terms “second-order belief” and perception are thus used interchangeably.\\n\\n21\", \"datasets\": [{\"mention_name\": {\"text\": \"national statistics\", \"confidence\": 0.4116002023220062, \"start\": 819, \"end\": 838, \"clean_text\": \"national statistics\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 819, \"end\": 838, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 819, \"end\": 838, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 819, \"end\": 838}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 819, \"end\": 838, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"geospatial\", \"confidence\": 0.7017004489898682, \"start\": 819, \"end\": 838}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4116002023220062, \"start\": 819, \"end\": 838}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.635034441947937, \"start\": 819, \"end\": 838}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.635034441947937, \"start\": 819, \"end\": 838}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"peers’ support for the introduction of gender equal inheritance law. Second-order beliefs were measured before the assignment of treatment which is why I use the full sample. First-order beliefs were measured in the first endline. The average first order-beliefs are thus computed based on the control group.\\n\\nFigure 3 aggregates the second-order beliefs of female and male respondents and contrasts them with the actual average first-order beliefs. Unsurprisingly, respondents expect men to be less supportive of legal reform than women: the modal belief is that 0 out of 10 men support legal reform. Perceptions about women’s reform support are much more evenly spread. On average, respondents expect 42 percent of women and 19 percent of men to support legal reform. These second-order beliefs imply that the expected gender gap in reform support amounts to 23 percentage points. This is considerably larger than the actual gender gap revealed by first-order beliefs, which I illustrate with line bars in figure 3: 42 percent of women versus 30 percent of men support a reform. The gender gap in first-order beliefs hence amounts to 12 percentage points, barely more than half the gender gap as measured by second-order beliefs. I conclude that respondents tend to overestimate the gender polarization around the introduction of gender equal inheritance law.\\n\\nWhat drives this misperception? Average second-order beliefs regarding women’s reform support are surprisingly close to the true mean. The misperception is driven by respondents underestimating male reform support. One possible hypothesis is that women avoid discussing gender equality in inheritance with men, creating information asymmetry by gender. In that case, we would expect that women have a less accurate understanding of men’s reform support than men (and vice versa for women’s reform support, see Gulzar et al., 2020). But not only women underestimate men’s reform support: figure 6 plots second-order beliefs regarding men’s reform support by gender. The distribution is strongly skewed to the left and there are hardly any differences by gender - men tend to underestimate other men’s reform support just as much as women do. Figure 5 suggests that respondents on average hold quite accurate expectations regarding female reform support. Male respondents tend to give more extreme answers. Overall, respondents tend to overestimate the gender polarization around gender equality in inheritance because they _underestimate_ men’s openness to reform.\\n\\nFigure 4 presents beliefs aggregated by gender as they were used for administering T2. Almost 60 percent of respondents underestimated public reform support and received the corresponding information if allocated to T2. This is in line with Bursztyn et al. (2023) who find that respondents in relatively less gender progressive countries tend to underestimate 22\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"support for strengthening women’s rights. Only 27 percent overestimate the public reform support. The remaining 13 percent guessed correctly.\\n\\n# **4.1.3 Main Outcomes**\\n\\nThe proposal to reform gender discriminatory inheritance law is unpopular in Tunisia. Aggregating both genders, I find that 37 percent of Tunisians are favorable to the introduction of a gender equal inheritance law regime.<sup>13</sup> The other measured outcomes provide a consistent picture of respondents’ attitudes towards gender equality in inheritance as illustrated by figure 7. Among the respondents in the control group, 35 percent of respondents prefer a gender equal inheritance split in the hypothetical bequest scenario. By contrast, reported willingness to gift one’s daughter or another female relative is high: 66 percent of all respondents in the control group declared that they would be willing to use gifting for achieving a more gender equal inheritance split among their children.\\n\\nDemand for gender equal inheritance arrangements differs by gender: women are significantly more likely to choose the ideal bequest (p-value _<_ 0.05) and significantly more likely to support legal reform (p-value _<_ 0.01). There is no statistically significant gender difference in the willingness to use gifting. The share is the same among respondents with a low score on the social desirability index, which suggests that this finding is not driven by experimenter demand effects. Table 2 shows the average marginal effects from a logit model regressing the main outcomes on respondents’ socio-economic characteristics, as well as second-order beliefs at baseline in the control group. Women are more likely to demand gender equality than men: they are almost 16 percentage points more likely to choose a gender equal ideal bequest and 19 percentage points more likely to favor legal reform, a finding which is significant at the 1 percent level.\\n\\nOther respondent characteristics matter: differences across age cohorts are not statistically significant but on average, older respondents tend to endorse less gender progressive attitudes than younger respondents. Support for gender equality in inheritance significantly decreases with educational attainment and religiosity. University graduates are 18 percentage points less likely to support legal reform; the decrease is 4 percentage points for highly religious individuals (significant at the 1 percent and 5 percent level). Importantly, respon-\\n\\n> 13This percentage mirrors the result of my initial survey which was conducted one year earlier in March 2022 and served as the basis for the misperception correction performed in T2.\\n\\n23\", \"datasets\": [{\"mention_name\": {\"text\": \"social desirability index\", \"confidence\": 0.8507440090179443, \"start\": 1355, \"end\": 1380, \"clean_text\": \"social desirability index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1355, \"end\": 1380, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1355, \"end\": 1380, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1355, \"end\": 1380}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1355, \"end\": 1380, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9489547610282898, \"start\": 539, \"end\": 564}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8507440090179443, \"start\": 539, \"end\": 564}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5900481939315796, \"start\": 539, \"end\": 564}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5900481939315796, \"start\": 539, \"end\": 564}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"dents’ first- and second-order beliefs are strongly correlated: the best predictor of favoring a gender discriminatory bequest allocation and opposing legal reform is underestimating one’s peers openness to reform. Individuals who do so are 24 percentage points less likely to choose a gender equal bequest and 29 percentage points less likely to support legal reform themselves. Both findings are significant at the 1 percent level.\\n\\n# **4.1.4 Gifting at Baseline**\\n\\nMaking gifts to one’s daughter can only be a viable policy alternative to legal reform if doing so is socially acceptable. Table 3 summarizes respondents’ previous use of and exposure to gifting as reported in the first endline: about 16 percent of respondents have received a gift themselves. The share is 18 percent among male and 14 percent among female respondents - a means difference which is statistically significant at the 10 percent level. About 13 percent report that their spouse has received a gift, and 9 percent of respondents report that they have made a gift to their daughter or another female relative. Men are more likely to make gifts to their daughter or, when they do not have a daughter, to a female relative: 13 percent of male respondents versus only 5 percent of female respondents report having done so in the past. The means difference is statistically significant at the 1 percent level.\\n\\nThe last row of table 3 aggregates these different indicator into one measure of gifting experience, which is defined as having received a gift oneself, one’s spouse having received a gift, or having made a gift to one’s daughter (or female relative when a respondent does not have a daughter). More than one in four respondents in the sample has some previous experience with gifting, confirming that gifting is a common practice in Tunisia. Among men, the share is close to 30 percent. Two stylized facts emerge from this descriptive data which are highly policy relevant: first, men gift more often than women. Second, men are more likely to receive a gift than women.\\n\\n# **4.2 The Tradeoff between Legal Reform and Gifting**\\n\\nA majority of Tunisian adults opposes a nationwide reform introducing gender equal inheritance law. Gifting daughters could be perceived as a way of circumventing the gender discriminatory inheritance split imposed by the current law. In the following subsections, I will shed light on the question whether gifting one’s daughters is an alternative to reforming gender discriminatory inheritance law. First, I assess some of the reasons which explain _why_ 24\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"an individual may be hesitant towards legal change and favor gifting instead. Second, I take a closer look at the correlational evidence to further characterize the tradeoff between legal reform and gifting: _who_ favors gifting over reform?\\n\\n# **4.2.1 Reasons for Opposing Legal Reform**\\n\\n# **Efficiency, Necessity and Priority**\\n\\nAn individual may oppose legal reform for many reasons. Legal reform may not be sufficient to promote gender equality if social norms remain biased against women. The literature provides numerous examples of legal reforms that remained inefficient or even backfired. In India, for example, the introduction of gender equal access to ancestral land has been shown to remain ineffective as families continued favoring their sons over their daughters (Roy, 2015). Beyond the main outcomes, I therefore asked respondents to assess whether legal reform would be an efficient way of promoting gender equality in inheritance, whether it was necessary, and whether they considered it a personal or political priority.<sup>14</sup> Efficiency concerns play a role: 55 percent of respondents in the control group doubt that inheritance law reform would effectively strengthen women’s rights. But more strikingly, 79 percent of respondents believe that the introduction of a gender equal law is not necessary. Only 13 percent see legal reform as a personal priority and only 14 percent think it ought to be a political priority. Neither treatment has a significant effect on respondents’ beliefs about the efficiency, necessity and urgency of legal reform. Table 4 presents the estimated average marginal effects from a logit model, regressing the respective outcome on the respondents’ socio-economic characteristics. The model presented is estimated on the control group.\\n\\nIn contrast to the previous findings regarding the demand for gender equal inheritance law and support for gifting, under or overestimating reform does not correlate in a statistically significant way with any of the outcomes. Women and men do not significantly differ in their assessment of the efficiency, necessity and importance of legal reform either: the estimated\\n\\n> 14The original wording of the items is the following: Efficient: “Establishing gender equality in inheritance in Tunisia would not strengthen women’s rights because the law is often not applied anyways.” Necessary: “Reforming inheritance law in Tunisia is not necessary because the current inheritance law is fair.” Personal Priority: “I have more urgent things to deal with in my daily life than inheritance law reform.” Political Priority: “There are more important policy issues to be taken care of right now than gender equality in inheritance.” The items were reversed and coded as binary variables, such that strong or weak agreement with the gender equal choice is indicated by “1”. Equivalently, “0” indicates that the respondent preferred the gender unequal choice.\\n\\n25\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"coefficients are close zero. Married individuals tend to hold less gender progressive views: they are almost 13 percentage points less likely to consider that gender equality in inheritance ought to be a political priority. This finding is statistically significant at the 1 percent level.\\n\\nLegal reform ought to be a particularly important topic for individuals who would be themselves materially affected by reform. Indeed, individuals whose father or mother are alive are significantly more likely to believe that gender equality in inheritance should be a political priority. Consistent with some previous research (see for example Borrell-Porta et al., 2018), respondents who have at least one daughter embrace more gender equal views on average: they are almost 5 percentage points more likely to believe that a legal reform is necessary and about 5 percentage points more likely to consider that it should be a political priority. Interestingly, individuals who have at least one son are also significantly more likely to view gender equality in inheritance as a political priority. Older individuals are significantly less likely to consider gender equality in inheritance a personal priority.\\n\\nAs in the case of the main outcome measures, education correlates negatively with all outcomes: university educated respondents are 14 percentage points less likely to consider legal reform as a personal priority than respondents with only basic education. They are also 13 percentage points less likely to believe it would be an efficient means of promoting gender equality. Material wealth is only a significant predictor of the first outcome: wealthy individuals are 5 percentage points more likely to believe in the efficiency of legal reform (significant at the 1 percent level).\\n\\n# **Need, Equity, Inter-Generational Exchange**\\n\\nThe fact that a vast majority of respondents believe that legal reform is not necessary implies that they find the status quo fair. A potential explanation would be that they consider that daughters are sufficiently taken care of by their husbands or other male family members. Indeed, 96 percent of respondents strongly or somewhat agree with the statement that men are obliged to financially provide for their wives, sisters and mothers (see table 5). The belief that women do not _need_ equal inheritance rights because they receive financial assistance may thus motivate the idea that legal reform is not necessary.\\n\\nWhat if a daughter _is_ in need? The survey included an additional outcome which triggered the respondent to re-assess their ideal bequest allocation between a son and a daughter. In this scenario, the daughter had recently divorced after being abused by her husband and the script specified that he no longer provided for her financially (see Online Appendix for 26\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [27]}}, {\"page\": 28, \"chunk\": 28, \"input_text\": \"the whole script). The respondents were asked whether they wanted to change their ideal bequest allocation. I find that about 10 percent of respondents did so, leading to 52 percent of women and 40 percent of men favoring a gender equal bequest in this scenario. The share of respondents who change their preferences when the daughter is in need is not negligible but not large enough to suggest that economic need is a guiding principle behind respondents’ preferences.\\n\\nExchange expectations and equity considerations constitute another, not necessarily competing explanation: sons may be expected to take care of their parents when in old age and it may therefore be fair to compensate them with a higher inheritance share. In a longitudinal study of US women, Light and McGarry (2004) find that exchange is a common explanation for respondents to favor unequal bequest allocations among their children. Similarly, Tisch and Gutfleisch (2022) show that German respondents in an online experiment allocate a significantly larger gift to a child who helps their parents. Inter-generational exchange and equity considerations are not an obvious explanation for respondents’ preference for gender discriminatory law in the Tunisian context, though: 98 percent of respondents consider that a son has the duty to financially support his parents when they get old. The percentage of respondents who believe that a daughter has the duty to financially support her parents when they get old is 93 percent - almost the same.\\n\\nTable 5 points out some gender differences with respect to these attitudes. Women are significantly more likely to believe that legal reform is necessary than men (24 versus 17 percent). Gender differences in family norms can motivate this finding: women are significantly less likely to adhere to the social norm that a man is responsible of his female family members. They are just as likely to expect a son to financially take in charge his elderly parents as men, but they are significantly more likely to expect daughters to do the same. Gender differences in intergenerational equity concerns may thus partially explain why women are more favorable towards reforming the law than men. Yet, even among women, a majority rejects legal reform and considers it unnecessary.\\n\\n# **4.2.2 Gifting as a Private Second-Best Alternative?**\\n\\nFrom the evidence presented so far, it is clear that the majority of Tunisians oppose a national reform introducing gender equality in inheritance. One of the most commonly held beliefs is that reform is simply not necessary. Considering this finding, I turn to the following 27\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [28]}}, {\"page\": 29, \"chunk\": 29, \"input_text\": \"# question: To what extent does gifting represent an alternative to legal reform?\\n\\nI use correlational evidence to shed further light on the potential tradeoff between legal reform and private action in the form of gifting. Table 6 shows the estimated marginal effects from a logit model: as a benchmark, column (1) shows respondents’ stated preferences regarding legal reform. The outcome of column (2) is favoring gifting over reform. This outcome is coded as “1” whenever a respondent is more favorable of gifting than of reform and “0” otherwise. Both models are estimated on the control group. In the first survey wave, 37 percent of the respondents favor gifting over reform. Column (3) uses reported gifts to daughters as an outcome. This outcome was measured in the first endline and is unlikely to have been affected by treatment, which is why I use the whole sample for estimation. As reported in table 3, about 9 percent of the total sample had already made a gift to their daughter. Figure 8 plots the estimated coefficients of interest of model (2) and (3).\\n\\nWho is most likely to favor gifting over reforming the national law? The best predictor of outcomes 1 and 2 is again underestimating reform support: respondents who do so are 28 percentage points less likely to approve of reform themselves and 25 percentage points more likely to favor gifting. Women are significantly _more_ likely to favor legal reform and significantly _less_ likely to substitute towards gifting than men. Column (3) offers a potential rationalization for this observation: mothers are 6 percentage points less likely to make gifts to their daughters than fathers. Possible reasons may include lower asset and land ownership among women, but also limited bargaining power within the household. It would be desirable to analyze whether gifting is more common among female heads of the household (i.e. widows) but the sample size does not allow a more fine-grained analysis.\\n\\nGiven the gender discriminatory bequest allocation among siblings in Islamic inheritance law, individuals who have a sibling of the opposite sex would be expected to be less willing to embrace legal reform, as it would change the inheritance allocation within their family. The effect should be stronger for men because they would themselves materially suffer from reform as their sister(s) gain equal inheritance right. Consistent with the theory, I find that respondents who have a sibling of the opposite sex are indeed less likely to favor reform (not statistically significant) and 14 percentage points more likely to favor gifting over reform (marginally statistically significant at the 10 percent level). The limited sample size does not allow for analysis by gender.\\n\\nHigher levels of education are generally associated with more gender equal preferences, as the most recent survey round of the Tunisian Arab Barometer confirms (Arab Barometer; 28\", \"datasets\": [{\"mention_name\": {\"text\": \"Tunisian Arab Barometer\", \"confidence\": 0.584068238735199, \"start\": 2871, \"end\": 2894, \"clean_text\": \"Tunisian Arab Barometer\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2871, \"end\": 2894, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2871, \"end\": 2894, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2871, \"end\": 2894}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2871, \"end\": 2894, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.999569833278656, \"start\": 431, \"end\": 454}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.584068238735199, \"start\": 431, \"end\": 454}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6230718493461609, \"start\": 431, \"end\": 454}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6230718493461609, \"start\": 431, \"end\": 454}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [29]}}, {\"page\": 30, \"chunk\": 30, \"input_text\": \"2022). Yet, attitudes towards women’s right to equal inheritance seem to be an exception: column (1) of table 6 clearly shows that reform support linearly decreases with education. University graduates are in fact the least likely to support legal reform. I control for wealth in all regressions, hence, this effect is unlikely to be driven by the correlation of education and wealth. The probability of favoring gifting and reporting an actual gift to one’s daughter increases with education, albeit not in a statistically significant way (see columns 2 and 3).\\n\\nReligious individuals would be expected to be less supportive of reforming inheritance law but it is not clear that they would favor gifting over reform: gifting daughters does not directly conflict with Islamic inheritance law but highly religious individuals may still disapprove of it if they view it as a way of circumventing the law. The regression results in table 6 shows that respondents who score above the median of the religious conservatism scale are 4 percentage points less likely to support legal reform (significant at the 1 percent level) but not significantly less likely to favor gifting. Gifting may thus to be viewed as an alternative to legal reform, even by some of those who are highly religious.\\n\\nFrom a theoretical perspective, it is not clear how wealth impacts attitudes towards women’s right to inheritance. Wealthy families would be more strongly impacted by a legal reform and may therefore be more likely to oppose it. But they also have more possibilities to use gifting. I do not find a significant impact of wealth on attitudes towards reform or having gifted one’s daughter, which is likely due to measurement issues and a lack of statistical power. However, wealthy individuals are 3 percentage points more likely to favor gifting over reform, a finding which is statistically significant at the 10 percent level. This is intuitive if one considers that gifting can only be used to alleviate gender inequality in inheritance in those families that dispose of sufficient assets and property to write over some to their daughters while alive.\\n\\n# **4.3 Treatment Effects**\\n\\n# **4.3.1 Main Effects**\\n\\nTable 7 reports the treatment effects of the pure gifting treatment (T1) and the combined treatment with belief correction and information on gifting (T2) compared to control. All regressions control for baseline second-order beliefs, respondent gender, age, marital status, education, occupation, family covariates, religious conservatism, a wealth index, and rural location. I furthermore include governorate and enumerator fixed effects. All OLS models 29\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [30]}}, {\"page\": 31, \"chunk\": 31, \"input_text\": \"are estimated using heteroskedasticity robust standard errors.\\n\\nAs expected, T1 has a marginally significant and positive effect on the ideal bequest - treated respondents are 6.9 percentage points more likely to choose a gender equal bequest allocation (significant at the 5 percent level) - but there is no statistically significant effect on the other outcomes. When pairing the information on gifting with belief correction in T2, treatment effects are zero for the first two outcomes and positive but imprecisely estimated for gifting. If anything, adding information on reform to the gifting information seems counterproductive. Yet, these aggregated treatment effects hide considerable heterogeneity by prior beliefs, as illustrated by table 8.\\n\\n# **4.3.2 By Misperceptions**\\n\\nTable 8 disaggregates the treatment effects of T2 by prior beliefs. The effects of T2 on the subgroup of respondents who underestimated reform support (almost 60 percent of the sample) are as expected positive on all three outcomes but remain shy of statistical significance. The effects of T2 on the subgroup of respondents who overestimated support are negative, but again remain shy of statistical significance. I test the null hypothesis that the estimated treatment effect on the subgroup of respondents who underestimated reform support equals the treatment effect on the subgroup of respondents who overestimated reform (T2 = T2 x Over). I fail to reject the null hypothesis for all three outcomes. The large p-values reported in the last row of table 8 indicate that there is no statistically significant treatment effect difference between the two groups. This goes against the hypothesis that the treatment would lead to belief updating and strong heterogeneity between these groups. Surprisingly, the treatment effects of T2 on the subgroup of respondents who already hold correct beliefs (Treat 2 + Treat 2 x Correct) is negative and highly statistically significant for the ideal bequest and reform support. Upon receiving T2, respondents holding correct beliefs are more than 20 percentage points less likely to favor a gender equal bequest and almost 20 percentage points less likely to support reform than respondents in control. For the ideal bequest and reform support, the treatment effects on respondents holding correct beliefs significantly differ from the effects on the groups who misperceive public reform support as can be seen from the p-values reported in table 8. For willingness to gift, the treatment effect on respondents holding correct beliefs is zero.\\n\\nInformation-based updating would imply that respondents who underestimated their 30\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [31]}}, {\"page\": 32, \"chunk\": 32, \"input_text\": \"peers’ reform support become relatively more gender progressive as they realize that their environment is more supportive of gender equality than previously thought (Bleemer and Zafar, 2018). Equivalently, respondents who overestimated reform support would be expected to become less gender progressive. Yet, this theoretical framework would expect them to react more strongly to treatment than respondents who hold correct beliefs. This is not what I observe in the data, suggesting that information-based updating cannot fully account for the observed treatment effect pattern.\\n\\nHow can the strong treatment effects on respondents holding correct beliefs be rationalized? One potential explanation is that respondents who underestimate (overestimate) reform support are inherently different from respondents who hold correct beliefs. In order to explore this hypothesis, I investigate the predictors of respondents’ second-order beliefs. Table 9 shows average marginal effects from a logit model where the outcomes are underestimating public reform support, holding correct beliefs, and overestimating public reform support. Respondents’ gender, family composition, religiosity or wealth do not predict second-order beliefs in a statistically significant way. Education does: the probability of holding pessimistic second-order beliefs significantly increases with education. University graduates are most likely to underestimate public reform support and least likely to overestimate it.\\n\\nRecall that even in the control group, education is an excellent predictor of holding conservative attitudes with respect to women’s right to inheritance as shown in table 2. This points to another potential explanation of the treatment effect patterns by misperceptions: respondents in the control group who underestimate (overestimate) public reform support are more likely to oppose (support) gender equal inheritance themselves.<sup>15</sup> It is possible that respondents oppose gender equality in inheritance _because_ they believe that their peers oppose gender equality in inheritance (analogous to Kuran (1991) and Bursztyn et al. (2020)). But a priori, the observed correlations do not allow for any conclusions regarding the direction of a potential causality. It is also possible that the respondents themselves oppose gender equality and make inferences on their peers’ gender attitudes based on their own political convictions. In this scenario, first-order beliefs shape second-order beliefs and not the other way round.\\n\\n> 15When standardizing individual reform support and beliefs regarding public reform support, I find that the correlation of respondents’ reform support and expected reform support in the control group is 0.37 and highly statistically significant. This is in line with the literature review of Bursztyn and Yang (2022) who find that attitudes and beliefs about others tend to correlate positively with individuals’ own beliefs.\\n\\n31\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [32]}}, {\"page\": 33, \"chunk\": 33, \"input_text\": \"If this mechanism is at work, then the respondents who misperceive public reform support are expected to hold stronger political opinions at baseline. The recent experimental literature suggests that the strength of political opinions at baseline crucially matters for the success of informational interventions: Settele (2022) finds that correcting beliefs about the gender wage gap suffices to shift the policy preferences of Democrats and Independents, but not those of Republicans who tend to hold strong anti-redistribution preferences at baseline. Alesina et al. (2018b) show that the respondents who hold the strongest misperceptions about migrants are the least willing to pay for corrective information, suggesting that respondents’ baseline beliefs limit their response to the informational treatment. In a similar vein, Alesina et al. (2018c) find that correcting information on social mobility increases Democrats’ demand for redistributive policies, but has little effect on Republicans. Given the strong polarization of public opinion on gender equality in inheritance in Tunisia, it is plausible that a similar mechanism is at play: respondents with strong political beliefs are less likely to respond to an informational treatment, which could explain why the observed treatment effects on these subgroups are so weak. Following this reasoning, respondents who hold correct beliefs would be expected to be relatively more moderate in their own political views. As a result, their gender norms may be more malleable and more responsive to an informational treatment.\\n\\nThe treatment effects reported in table 10 provide additional evidence for the hypothesis that the treatment effect pattern is driven by respondents’ underlying characteristics rather than by belief updating. I estimate the same model as reported in table 8 but this time using T1. Recall that T1 only provided information on gifting and did _not_ include a belief correction. Despite this, the same treatment effect pattern persists. Respondents who underestimate public reform support become significantly more likely to choose a gender equal bequest and to support reform. The interaction terms on belonging to T1 and holding correct beliefs or overestimating public reform support are negative. Importantly, respondents who held correct beliefs show very similar treatment effects as in T1. This treatment effect pattern would be roughly consistent with belief updating - but there was no belief correction. The finding that respondents in T1 and T2 react in the same way to treatment allows me to rule out that it is the belief correction in T2 which is driving the observed treatment effect pattern. Instead, it is likely that the treatment effect pattern is driven by something which is shared by respondents allocated to the different treatment arms and which conditions the way they react to the informational treatment: their underlying characteristics.\\n\\nTable 11 presents a breakdown of respondent characteristics by prior beliefs. Respondents 32\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [33]}}, {\"page\": 34, \"chunk\": 34, \"input_text\": \"who _underestimate_ public reform support are on average the most highly educated. Moreover, they score highest on the aggregated wealth index and they are the least likely to live in rural areas. Respondents who hold _correct_ beliefs score a bit lower in terms of education and wealth but the difference to the group of respondents who underestimate is not statistically significant. Respondents with correct beliefs are the least religious of the three groups. They use social media more often and have the best knowledge of Tunisian feminism - the differences in means, especially in comparison to respondents who overestimate, are highly statistically significant. They are also most likely to respond that they often or very often think about inheritance matters. Respondents who _overestimate_ public reform support are significantly more likely to favor gender equal outcomes than the other groups. However, they are significantly less educated than respondents who hold correct beliefs, less wealthy and more religious than the other groups. They make significantly less use of social media and are not well-informed about the Tunisian feminist movement. Finally, they score considerably higher on the social desirability index than the other groups. The data thus suggests that that the three groups substantially differ from each other: respondents who underestimate public reform support seem to be part of the educated and wealthy but conservative elite of the country. Respondents who hold correct beliefs have relatively more gender progressive preferences and are significantly more interested in feminism. Conversely, respondents who overestimate support do not seem to be particularly involved in gender questions but display a higher propensity to give socially desirable answers, which may partially explain their high first-order beliefs.\\n\\nThe next subsections will discuss further treatment effects heterogeneity and findings from the second endline.\\n\\n# **4.3.3 Further Heterogeneity**\\n\\nIn contrast to my pre-registered hypotheses I find little evidence of heterogeneous treatment effects by previous exposure to feminism or levels of religious conservatism. However, treatments effects are considerably larger when the models are estimated on the subsample of respondents whose father has passed away: table 4 in the Online Appendix shows that respondents in T1 in this subgroup are 11 percentage points more likely to opt for a gender equal bequest split than respondents in the control group. This is consistent with respondents’ beliefs being less malleable while their parents are alive; recent research has indeed highlighted the role of older family members in perpetuating restrictive gender norms at 33\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [34]}}, {\"page\": 35, \"chunk\": 35, \"input_text\": \"home (see for example Anukriti et al., 2020).\\n\\nConducting treatment effect analysis by gender suggests that the observed treatment effects are driven by the female respondents. In the female subsample, T1 makes choosing the gender equal bequest almost 13 percentage points more likely, a finding which is significant at the 1 percent level (see table 4 in the Online Appendix). All other treatment effects of T1 are positive and of economically important size, yet statistically not significant, possibly reflecting the smaller sample size. T2 has a statistically significant and positive effect on willingness to use gifting: among the female subsample, the likelihood to favor gifting increases by 10 percentage points (significant at the 5 percent level). This is driven by the female respondents who underestimate support as visible in table 6 of the Online Appendix: they are 11 percentage points more likely to favor gifting as a means of reaching a gender equal bequest allocation, again significant at the 5 percent level. I conclude that female respondents respond more strongly to the informational treatment and show stronger evidence of belief updating. A potential interpretation is that the perceived social norm could act as a constraint on the preferences of female respondents, but less so for male respondents.\\n\\nPersonal interest in inheritance matters is likely to play a decisive role. At baseline, a majority of respondents declared that they never think about inheritance issues. Excluding these respondents from analysis yields interesting results: table 5 in the Online Appendix illustrates that T1 is associated with a statistically significant and sizeable increase in support for legal reform in the subgroup of respondents who sometimes, often or very often think about inheritance. The other treatment effects are also positive but shy of statistical significance. Treatment 2 leads to a large increase in willingness to use gifting which is significant at the 10 percent level. Table 7 in the Online Appendix shows that this increase in willingness to use gifting is driven by respondents who hold correct beliefs, which is in line with the interpretation that their preferences are more malleable. Given the limited sample size, this finding should be taken with a grain of salt. Yet, it suggests that individuals who hold an interest in inheritance matters substitute away from gender equal inheritance towards gifting.\\n\\n# **4.3.4 Second Endline**\\n\\nMore than 90 percent of respondents who were interviewed agreed to being recontacted and 45 percent answered the call, yielding a sample size of N=672. The Online Appendix includes 34\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [35]}}, {\"page\": 36, \"chunk\": 36, \"input_text\": \"a table with covariate balance of the respondents who were successfully recontacted for the second endline. Overall, respondent characteristics are fairly balanced - the statistically significant imbalances observed with respect to educational attainment mirror those of the first endline. Table 1 in the Online Appendix shows the results from regressing an indicator variable for participation in the second endline on treatment indicators and covariates. The estimated treatment coefficients are not statistically significant, suggesting that attrition did not differ by treatment status. However, older individuals were significantly more likely to answer the call, which is why I weight down their responses in a robustness check reported in the Online Appendix.\\n\\nTable 12 reports findings from estimating model 4. The follow-up indicator are consistently statistically insignificant. T1 has a zero impact on the ideal bequest chosen by the respondents, which suggests that the treatment effect has faded over time. By contrast, respondents who answered the second call were significantly more likely to show an increase in interest in gifting in the first call. The treatment effect decreases in the follow-up but is still positive and statistically significant at the 5 percent level for T1. Is this finding driven by the fact that older individuals were more likely to answer the follow-up survey? In the Online Appendix, I estimate the same model introducing survey weights to ensure higher representativeness. The results look reassuringly similar. The positive effect of T2 on the willingness to use gifting becomes even stronger and statistically significant at the 5 percent level. The second endline thus suggests that both treatments increased respondents’ willingness to use gifting. The heterogeneity of treatment effects of T2 by prior beliefs persists, which is in line with the hypothesis that respondents who misperceive reform support differ from respondents with correct beliefs on more substantive grounds.\\n\\nA total of 23 respondents reported having gifted a daughter or another female relative since the first interview. Eight female respondents reported having received a gift since the first interview. However, consistent with the gifting pattern found at baseline (see table 3), I also observe a significant number of gifts towards men. Seven respondents reported having gifted a son or a male relative and ten male respondents having received a gift themselves. Finally, 41 percent of respondents in the second endline reported that it was likely or very likely that they would use gifting to attain a more gender equal inheritance allocation in their families in the future. These numbers do not significantly differ by treatment arm, which is likely to be driven by a lack of statistical power.\\n\\n35\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [36]}}, {\"page\": 37, \"chunk\": 37, \"input_text\": \"# **4.4 Robustness Checks**\\n\\nTo assess whether the treatment effect heterogeneity of T2 holds for smaller intervals, I estimate the model using deciles. The results are reported in the Online Appendix. The same treatment effect pattern persists: respondents holding correct beliefs are clearly the subgroup who reacts most to treatment.\\n\\nAs a more general robustness check of my findings, I re-estimate the above models without imputing missing baseline covariate values. I also drop the respondents who did not find the provided information credible (1.5 percent) or not very credible (9.1 percent). I next drop respondents who take unusually little time to respond to the survey (less than 20 minutes, 5.5 percent). My main findings also hold for these smaller subsamples. Moreover, I construct an index of respondents’ score on the Crowne and Marlowe (1960) propensity to give socially desirable answers scale and re-estimate my main models on the subsample of those who score in the bottom 6 deciles of the distribution. The positive treatment effect of T1 on the ideal bequest is smaller and no longer statistically significant but still positive. This suggests that some of the treatment effect may be driven by social desirability bias. I estimate all models without imputing missing baseline covariate values. All these results are reported in the Online Appendix.\\n\\nFinally, to address potential selection bias created by survey non-response, I follow Valliant and Dever (2018) by estimating the pseudo-inclusion probabilities into my survey sample: I treat the Tunisian Labor Market Panel Survey (TLMPS), the most recent face-to-face panel survey (ERF, 2014), as reference probability sample, which I code as “0”. I then append it to my survey data, which I code as “1”. Using a logit model, I estimate _Pr_ ( _y_ = 1 _|X, β_ ), where _X_ is a set of covariates with common support in both samples.<sup>16</sup> This allows me to estimate pseudo-weights, defined as the inverse of the predicted probabilities. This strategy addresses the imbalance with respect to educational attainment which I observe across treatment arms. Moreover, as I show in the Online Appendix, my sample of respondents is highly representative of Tunisia in terms of the geographical spread, but on average younger. Weighting allows me to address this potential concern.\\n\\nWeighting does not significantly affect my conclusions. Regression results are reported\\n\\n> 16The set of covariates _X_ includes categories of age, education and occupation, household size, assets, and rural-urban status. I do not impute missing values in ERF (2014). Weighting down highly educated respondents also addresses baseline imbalance in the first treatment arm.\\n\\n36\", \"datasets\": [{\"mention_name\": {\"text\": \"Tunisian Labor Market Panel Survey\", \"confidence\": 0.8014627695083618, \"start\": 1570, \"end\": 1604, \"clean_text\": \"Tunisian Labor Market Panel Survey\"}, \"acronym\": {\"text\": \"TLMPS\", \"confidence\": 0.5920224785804749, \"start\": 1606, \"end\": 1611, \"clean_text\": \"TLMPS\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1570, \"end\": 1604, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1570, \"end\": 1604}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1570, \"end\": 1604, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.960668683052063, \"start\": 787, \"end\": 821}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8014627695083618, \"start\": 787, \"end\": 821}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6933549642562866, \"start\": 787, \"end\": 821}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6933549642562866, \"start\": 787, \"end\": 821}}, {\"mention_name\": {\"text\": \"survey data\", \"confidence\": 0.6838034391403198, \"start\": 1746, \"end\": 1757, \"clean_text\": \"survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1746, \"end\": 1757, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1746, \"end\": 1757, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1746, \"end\": 1757}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1746, \"end\": 1757, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9893877506256104, \"start\": 97, \"end\": 108}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6838034391403198, \"start\": 97, \"end\": 108}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5717277526855469, \"start\": 97, \"end\": 108}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5717277526855469, \"start\": 97, \"end\": 108}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [37]}}, {\"page\": 38, \"chunk\": 38, \"input_text\": \"in the Online Appendix: the treatment effect of T1 on the ideal bequest chosen is smaller but still statistically significant at the 10 percent level. The treatment effect pattern by misperceptions becomes weaker but it is still respondents with correct beliefs who display the strongest treatment effects. Respondents who hold correct beliefs are 19 percentage points less likely to support reform upon receiving treatment 2 (significant at the 10 percent level). Thus, the observation that belief updating is smaller than expected still holds after introducing survey weights.\\n\\n# **4.5 Limitations**\\n\\nObserved treatment effects may be driven by (i) the information provided or by (ii) salience of the topic. My experimental design does not allow to disentangle these mechanisms. In addition to the information updating, the treatment may have additional side effects such as “uncertainty reduction, attention, and emotional responses” (Haaland et al., 2020, p.20). Ideally, a third treatment arm would test whether only providing the correct information on public reform support without adding the information on gifting already has an impact on individuals’ preferences.\\n\\nAn obvious issue is potential bias introduced by survey non-response and experimenter demand effects. The questions were framed as neutrally as possible and the enumerators were instructed to avoid influencing the respondent. They received extensive training and were followed by a supervisor during the survey. In the data, I try to minimize this source of bias further by introducing enumerator dummies in all my regressions and estimating survey weights as a robustness check.\\n\\nAnother potential drawback is that this study did not provide financial incentives for belief elicitation, which could have induced low effort on the side of the respondents. However, Bursztyn and Yang (2022) do not find that incentives correlate with the share of respondents holding misperceptions. Incorrect second-order beliefs are thus not necessarily the result of low effort from the part of the respondent. Moreover, Haaland et al. (2020) caution that financial incentives can have undesirable side-effects. Based on the previous literature, it is therefore not clear that financial incentives would have been desirable in this context - they may as well increase social desirability bias and experimenter demand effects.\\n\\nA further drawback of informational experiments is that respondents may interpret the received information differently based on their personal experiences and background. I show 37\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [38]}}, {\"page\": 39, \"chunk\": 39, \"input_text\": \"covariate balance in both survey rounds (table 1 and table 1 in the Online Appendix) and control for a large set of observable characteristics as well as baseline second-order beliefs to alleviate this concern. There are other _unobservable_ characteristics that may influence people’s reaction to the information, though: for instance, respondents may react less to the gifting information if they are already well informed. In some cases, the informational treatment would not provide any new information. I do not measure previous knowledge of gifting in baseline to avoid priming respondents who were allocated in control. I thus cannot control for it. Previous knowledge about gifing may bias treatment effect sizes downwards. A further concern is that respondents may react differently if they anticipate that a gender equal inheritance law will be passed soon. This is unlikely given the political context, though: the political consensus has shifted to the right and the incumbent president is strongly opposed to gender equality in inheritance (Verdier, 2020).\\n\\nThe survey mode imposed further restrictions on the data collection. I limited myself to quantifying the _occurrence_ of gifting and did not ask respondents to quantify the _value_ of the gifts they made or received. This is a major data limitation, yet, gaining reliable information on the value of gifts would have been difficult because of recall bias. Moreover, it would likely have increased the non-response bias as many respondents are uncomfortable disclosing information on personal wealth and financial transactions in an interview. Finally, it is beyond any doubt that adding a behavioral outcome would enhance this study. Considering the topic of research, this would require an extensive period of observation and data collection effort. Given current funding constraints, this is left to future research.\\n\\n# **5 Conclusions**\\n\\nPolicymakers and researchers worldwide agree that gender discrimination represents a major obstacle to sustainable development: women represent more than half of a country’s population and limiting their ability to participate in politics, the economy and society is likely to harm not only the women themselves but society as a whole. Gender discrimination is particularly striking when encoded in national law. Yet, reforming gender discriminatory law can be difficult, and probably even more so when it is of religious origin (Htun et al., 2019). This study assesses public opinion regarding the gender discrimination among siblings which is implied by Islamic inheritance law. Sons are systematically favored over daughters. Most Muslim-majority countries worldwide continue to apply Islamic inheritance law. This 38\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [39]}}, {\"page\": 40, \"chunk\": 40, \"input_text\": \"gender discrimination in inheritance law has been questioned by a large social movement in Tunisia since 2018 but the reform proposal has never reached a parliamentary vote.\\n\\nThis study asks whether gifting is a second-best private alternative to the reform of gender discriminatory inheritance law and whether its use can be encouraged by an informational intervention. To answer these research questions, I conducted a nationwide phone survey, reaching a sample of more than 1,500 Tunisian adults. The survey gathered extensive descriptive and correlational evidence on Tunisians’ attitudes regarding women’s right to inheritance, as well as their use of gifting. With an informational experiment which was embedded in the survey, I moreover tested whether gifting one’s daughter can be encouraged via an informational treatment.\\n\\nTunisia, probably the country with the strongest feminist tradition in North Africa, represents an ideal case study for this research. The legal status quo in Tunisia remains unchanged: a son receives twice the share of a daughter, as foreseen by the Qur’an. My study shows that the persistence of Islamic inheritance law in Tunisia is consistent with the preferences of the majority of citizens. The descriptive data I collect shows that only about one in three Tunisian adults favors the introduction of a gender equal inheritance law regime. However, legal reform does not constitute the only way of increasing women’s share in the final wealth allocation. The Tunisian civil code allows parents to make gifts to their daughter during their lifetime. By doing so, they can attenuate the gender discrimination among their children in the final wealth allocation. I find that public opinion towards gifting daughters is much more favorable than towards legal reform: two in three Tunisians state that they would be willing to use gifting to achieve a more gender equal inheritance allocation. The descriptive evidence offered by this study thus confirms that gifting is a widespread and socially accepted practice in Tunisia.\\n\\nDoes gifting one’s daughter represent a private alternative to the nationwide introduction of secular inheritance law? To some extent, yes. I conclude that gifting represents a way of attenuating the gender discrimination imposed by Islamic inheritance law but only families who are relatively well-off are fully able to exploit it. This is intuitive since gifting requires the existence of assets, such as property or land that can be written over to a child, and comes at a cost in terms of notary fees, but also in terms of effort, time and planning. The final decision regarding the within-family wealth allocation is likely to remain in the hands of the head of the household; in practice women are significantly less likely to make gifts than men. An informational intervention aiming to encourage gifting one’s daughters 39\", \"datasets\": [{\"mention_name\": {\"text\": \"nationwide phone survey\", \"confidence\": 0.8095289468765259, \"start\": 421, \"end\": 444, \"clean_text\": \"nationwide phone survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 421, \"end\": 444, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 421, \"end\": 444, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 421, \"end\": 444}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 421, \"end\": 444, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9994687438011169, \"start\": 421, \"end\": 444}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.8095289468765259, \"start\": 421, \"end\": 444}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7786902189254761, \"start\": 421, \"end\": 444}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7786902189254761, \"start\": 421, \"end\": 444}}, {\"mention_name\": {\"text\": \"descriptive data\", \"confidence\": 0.5727782845497131, \"start\": 1234, \"end\": 1250, \"clean_text\": \"descriptive data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1234, \"end\": 1250, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1234, \"end\": 1250, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1234, \"end\": 1250}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1234, \"end\": 1250, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.7270499467849731, \"start\": 330, \"end\": 346}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.5727782845497131, \"start\": 330, \"end\": 346}, \"usage_context\": {\"text\": \"background\", \"confidence\": 0.4105810523033142, \"start\": 330, \"end\": 346}, \"is_used\": {\"text\": \"False\", \"confidence\": 0.4105810523033142, \"start\": 330, \"end\": 346}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [40]}}, {\"page\": 41, \"chunk\": 41, \"input_text\": \"would therefore a priori have to target fathers to be effective. However, my descriptive data shows that first, men tend to endorse more gender conservative norms than women and second, gifts to sons and male relatives are also quite common. These are worrisome findings if one considers the growing literature that shows that well-intended reforms and policy interventions may backfire in the presence of regressive gender norms (Rosenblum, 2015, Roy, 2015, Bhalotra et al., 2020). Hence, any potential policy intervention must be carefully designed and evaluated.\\n\\nThis study evaluates the impact of two informational treatments which aim to encourage the use of gifting to increase daughters’ share in the final bequest allocation. In T1, respondents receive information on the possibility to use gifting and the fiscal benefit associated with it. In T2, respondents are first informed about the level of public support for the introduction of a gender equal inheritance regime. They are told whether they under- or overestimated public support or whether they held correct beliefs and subsequently receive the same information on gifting as respondents in T1. The evidence on the effectiveness of these informational treatments to encourage gifting women is mixed. Respondents become significantly more likely to favor a gender equal bequest when administered T1. The effect is stronger among women and among individuals whose father has passed away, which is consistent with earlier literature suggesting that preferences are less malleable in the presence of older family members (Anukriti et al., 2020). T2 significantly decreases the likelihood of favoring a gender equal bequest and supporting reform among the group of respondents who hold correct beliefs. T2 tends to increase respondents’ willingness to use gifting to achieve a more gender equal inheritance allocation. I interpret this as suggestive evidence in favor of the hypothesis that gifting is perceived as a private substitute to legal reform.\\n\\nThis study also contributes to the larger discussion on the importance of social norms for development. I show that a majority of Tunisians underestimate their peers’ willingness to reform the law. At first sight, perceived social norms appear to be closely linked to individuals’ demand for gender equality in inheritance: respondents who underestimate their peers’ support for gender equal inheritance law are significantly more likely to endorse conservative gender norms. The “gender gap” in second-order beliefs is almost twice as large as the “gender gap” in first-order beliefs: both men and women overestimate the gender polarization around the question and they do so because they underestimate men’s openness to reform. But the direction of the causality is not clear. Do second-order beliefs shape firstorder beliefs or is it maybe the other way round? The bulk of previous literature suggests that individuals adapt their preferences to what is perceived to be the social norm (Bursztyn 40\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [41]}}, {\"page\": 42, \"chunk\": 42, \"input_text\": \"et al., 2019, Bursztyn et al., 2020, Gulzar et al., 2020). However, it is plausible that the causality also runs the opposite way: individual perceptions of what the social norm is may be driven by own preferences. Indeed, my analysis of the treatment effect heterogeneity by prior perceptions shows that individuals who misperceive public reform support only weakly respond to the belief correction. I argue that this could be due to the fact that respondents who underestimate public support tend to hold relatively firm political opinions, which may make their preferences less malleable to an informational treatment. Similar results have been found in the US context where political opinions are strongly polarized with respect to issues of redistribution (Alesina et al., 2018b, Alesina et al., 2018c, Settele, 2022, Haaland and Roth, 2023). Considering the experimental evidence on treatment effect heterogeneity by prior misperceptions, I conclude that, even though Tunisians underestimate public support for inheritance law reform, this misperception is unlikely to be a binding constraint on the country’s ability to reform the law.\\n\\nAs social norms continue to evolve and younger generations reach voting age, the issues of gender discriminatory inheritance law and bequest practices will become even more pressing in Tunisia and the larger MENA region. A potential avenue for future research would be to vary the treatment: analysis of second-order beliefs shows that the misperception stems from respondents underestimating men’s reform support. Given the context, it is possible that the belief correction treatment would be stronger if it directly referred to _male_ support for introducing gender equal inheritance law, rather than aggregated support by both genders. But gifting is unlikely to be a viable alternative to legal reform in the long run given that its use is largely restricted to wealthy families. Future research should therefore investigate alternative ways of raising awareness about women’s rights and strengthening women’s access to property in the region.\\n\\n41\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099804402052432900/pdf/IDU16715ac0d1906814580188631d44bf194b848.pdf\", \"pages\": [42]}}]}"}}},{"rowIdx":98,"cells":{"doc_id":{"kind":"string","value":"34252504"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34252504\", \"last_modified_date\": \"2024-02-06T00:00:00Z\", \"authors\": {\"0\": {\"author\": \"Barnett,Carolyn Louise\"}}, \"count\": \"Morocco\", \"count_key\": \"82545\", \"docna\": {\"0\": {\"docna\": \"Gender Role Attitudes, Perceived Norms, and\\n the \\\"Double Burden'' in Morocco\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"Office of the Chief Economist (MNACE)\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34252504\"}, \"repnb\": \"WPS10694\", \"docdt\": \"2024-02-05T00:00:00Z\", \"datestored\": \"2024-02-05T06:22:20Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"To what extent do attitudes and\\n perceived norms around household roles hinder the emergence\\n of more gender-equal distributions of labor in Morocco?\\n Moroccan women undertake a disproportionate share of unpaid\\n household and care labor and participate in the labor force\\n at low rates. Yet everyday practices are shifting, and\\n normative expectations may be as well. From an online survey\\n of predominantly urban, employed Moroccans, this paper finds\\n that respondents aspire for men to be equal contributors in\\n care tasks. Yet, unpaid labor burdens remain highly unequal,\\n respondents disfavor men taking primary responsibility for\\n cooking or cleaning, and women's share of household\\n labor correlates with perceptions of what men prefer more\\n than with individuals' actual preferences. Results from\\n a conjoint survey experiment measuring preferences around\\n employment and the household division of labor confirm\\n respondents' interest in more egalitarian relations in\\n principle, but also suggest that strong preferences for a\\n male breadwinner family model will continue to drive an\\n unequal distribution of labor at home.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10694; MNACE labor and gender research programs; PEOPLE\", \"display_title\": \"Gender Role Attitudes, Perceived\\n Norms, and the \\\"Double Burden'' in Morocco\", \"disclosure_date\": \"2024-02-05T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099822202052424136/text/IDU18821daf01292214ebc1865c19a2c631d7d2f.txt\", \"listing_relative_url\": \"/research/2024/02/34252504/gender-role-attitudes-perceived-norms-double-burden-morocco\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099822202052424136/IDU18821daf01292214ebc1865c19a2c631d7d2f\", \"new_url\": \"2024/02/34252504/IDU18821daf01292214ebc1865c19a2c631d7d2f\", \"ext_pub_date\": \"2024-02-05T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Gender and Employment m1326451 618,Human Rights Norms and Mechanisms m1327321 2514,Gender and Economic Empowerment m1327622 911,Unpaid Work m1504003 2701,Employment Tools and Methods m1504070 2760\", \"docm_id\": \"IDU-8821daf0-2922-4ebc-865c-9a2c631d7d2f\", \"chronical_docm_id\": \"IDU-8821daf0-2922-4ebc-865c-9a2c631d7d2f\", \"publishtoextweb_dt\": \"2024-02-05T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10694\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"guid\": \"099822202052424136\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"64\", \"author_keys\": {\"0\": {\"author_key\": \"2206344\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099822202052424136/IDU18821daf01292214ebc1865c19a2c631d7d2f\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10694 Gender Role Attitudes, Perceived Norms, and the “Double Burden’’ in Morocco _Carolyn Barnett_ Middle East and North Africa Region Office of the Chief Economist February 2024\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10694\\n\\n# **Abstract**\\n\\nTo what extent do attitudes and perceived norms around household roles hinder the emergence of more gender-equal distributions of labor in Morocco? Moroccan women undertake a disproportionate share of unpaid household and care labor and participate in the labor force at low rates. Yet everyday practices are shifting, and normative expectations may be as well. From an online survey of predominantly urban, employed Moroccans, this paper finds that respondents aspire for men to be equal contributors in care tasks. Yet, unpaid labor burdens remain highly unequal, respondents disfavor men taking primary responsibility for cooking or cleaning, and women’s share of household labor correlates with perceptions of what men prefer more than with individuals’ actual preferences. Results from a conjoint survey experiment measuring preferences around employment and the household division of labor confirm respondents’ interest in more egalitarian relations in principle, but also suggest that strong preferences for a male breadwinner family model will continue to drive an unequal distribution of labor at home.\\n\\nThis paper is a product of the Office of the Chief Economist, Middle East and North Africa Region. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at carolynbarnett@arizona.edu.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [{\"mention_name\": {\"text\": \"online survey of predominantly urban, employed Moroccans\", \"confidence\": 0.7090256214141846, \"start\": 423, \"end\": 479, \"clean_text\": \"online survey of predominantly urban, employed Moroccans\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 423, \"end\": 479, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 423, \"end\": 479, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 423, \"end\": 479}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 423, \"end\": 479, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9988527297973633, \"start\": 423, \"end\": 479}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.7090256214141846, \"start\": 423, \"end\": 479}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.894307017326355, \"start\": 423, \"end\": 479}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.894307017326355, \"start\": 423, \"end\": 479}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"# Gender Role Attitudes, Perceived Norms, and the “Double Burden” in Morocco\\n\\nCarolyn Barnett<sup>∗</sup> Keywords: Household labor, care work, mental labor, gender norms, perceived norms, female labor force participation JEL codes: J16, J21, J22, J71\\n\\n> ∗Assistant Professor, University of Arizona. This work was supported by the MENA Chief Economist Office under the labor and gender research programs.\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# **1 Introduction**\\n\\nThe double burden many women face combining labor force participation with unpaid domestic labor and care work represents a major impediment to female labor force participation and political participation in the MENA region (ILO 2018; Nazier 2019; Women 2020). Expectations that women will take on the dominant role caring for the home and raising children lead many women to exit the labor force upon family formation (Assaad, Krafft, and Selwaness 2022). At the same time, norms around household task burdensharing are in flux, and active efforts to change norms are underway. In this domain as in others, attitudes and social practice could shift more rapidly than normative expectations of what others believe and value (or _perceived norms_ ). To what extent do attitudes and perceived norms around household roles hinder the emergence of more gender-equal distributions of labor in Morocco?\\n\\nThis study draws on an original online survey of 1,038 Moroccan adults to investigate these and related questions. A convenience sample recruited via Qualtrics in JanuaryFebruary 2023 answered questions about responsibility for household tasks within their homes, their attitudes around gender roles within the home, and their perceptions of what men, women, and their neighbors think about the same issues. The respondents also compared hypothetical ideal household configurations in a conjoint survey experiment that measured preferences over whether men and women each work outside the home and how unpaid labor and care tasks are distributed. An information experiment on the survey also tested whether encountering real data about increases in the time men spent on household tasks during the COVID-19 pandemic altered perceptions about how attitudes around household gender roles would evolve in the future.\\n\\nAmong the survey respondents as among the general Moroccan public, unpaid labor burdens remain highly unequal, yet men’s and women’s perceptions of their relative contributions diverge. I examine responses to questions about who in the respondents’ households is primarily responsible for 13 common tasks. Women’s estimates of the female share of household labor are higher than men’s estimates, with men more likely to claim that tasks are undertaken equally. Married women report near-exclusive responsibility 1\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"for cleaning, cooking, and doing the dishes, while married men report near-exclusive responsibility for paying bills and making repairs. On other tasks, women are more likely to report primary responsibility, while men are more likely to claim they and their spouse undertake these tasks equally. Unmarried respondents generally report that their mothers are responsible for a much greater share than their fathers. Women are more likely than men to report that their household responsibilities inhibit their ability to work or their career progression. Employed women do not report lower burdens of household labor than women not currently employed.\\n\\nI then examine how the distribution of household labor relates to respondents’ attitudes and perceptions of others’ attitudes around household gender roles. In this study, I use the term “perceived norms” to refer to these normative expectations, or beliefs about others’ beliefs and values, as opposed to beliefs about what others (empirically) do.<sup>1</sup> I find that the female household share of labor does not correlate with respondents’ personal attitudes or with their perceptions of what women believe, but does correlate with respondents’ perceptions of what men believe. That is, the survey findings confirm that respondents perceive the actual organization of household labor as mainly linked to what they think men value. In contrast, respondents’ personal beliefs are clearly linked to their aspirations to work outside the home (among women) or interest in having their wife work outside the home (among men). These results suggest that misperceived norms might inhibit women’s labor force participation, as other studies (Bursztyn, González, and Yanagizawa-Drott 2020) have found. Yet an information experiment included in the survey found no evidence that highlighting men’s contributions to household labor increased respondents’ expectations of more egalitarian household gender roles in the future.\\n\\nFinally, I examine a conjoint survey experiment assessing respondents’ ideal organization of labor in and outside the home. Respondents compared two hypothetical households and indicated which was closer to their ideal. I find that respondents are indifferent\\n\\n> 1. For a more extended discussion of the various dimensions of norms and perceptions of them, see Zeitoun, El Assiouty, Rahman, Osman, and Girgis (2023).\\n\\n2\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"between women taking primary responsibility for most tasks and men and women sharing those tasks equally, but disfavor men taking primary responsibility for cooking and cleaning. The intensity of these preferences, however, is dwarfed by all respondents’ strong preference that men work outside the home and men’s (weaker) preference that women not work outside the home. I argue this suggests that women’s unpaid labor burden is primarily a side effect of preferences for a male breadwinner family model which precludes men taking primary responsibility for household tasks, but does not reflect strong objections to men contributing to household tasks. Direct questions about who should work outside the home and who should be primarily responsible for household tasks confirm the strong attachment to the male breadwinner model and openness to more equal household contributions in relation to care tasks, but attachment to women’s primary responsibility for cooking and cleaning.\\n\\nIn the conclusion, I argue that taken together, these findings suggest the limited potential of efforts to alter the domestic distribution of labor as a way to “free women up” for formal labor force participation. Moroccans aspire for men to contribute more equally to household tasks—and many men already portray themselves as doing so—but strong opposition remains to men taking primary responsibility for tasks or foregoing paid employment. I argue that these findings also reflect a culture of “benevolent sexism” that valorizes women’s traditional roles while emphasizing that women deserve help at home from supportive partners. This can explain both the relative interest in “equalizing” household contributions in some spheres (and men’s interest in emphasizing the extent to which they contribute) co-existing with preferences for men working outside the home, women not working outside the home, and women taking primary responsibility for household tasks, where feasible.\\n\\nTogether, the findings suggest limited scope for expanding women’s labor force participation by altering attitudes or perceived norms around gendered responsibility for household tasks alone. Existing openness to more equal divisions of household labor in principle are constrained by the strength of attachment to a male breadwinner family model and disinterest in men taking primary responsibility for household tasks. Build3\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"ing on recent scholarship to investigate conceptions of masculinity and their influence on individuals’ preferences and behavior could be a path toward addressing the latter obstacles.\\n\\n# **2 Background**\\n\\nIn MENA as elsewhere, reconciling responsibilities at work and at home can be challenging, especially for women who have traditionally shouldered the majority of unpaid household labor. A high burden at home may pose an impediment to female labor force participation, which remains low by global standards in the MENA region despite women’s rising educational attainment (Assaad, Hendy, Lassassi, and Yassin 2020). Expectations that women will take on the dominant role caring for the home and raising children lead many women to exit the labor force upon family formation (Assaad, Krafft, and Selwaness 2022). State-funded childcare facilities are rare, and private childcare provision is usually informal, in-home, and minimally regulated. Where public childcare is available, insufficient operating hours may nonetheless inhibit or even decrease women’s labor force participation (Krafft and Lassassi 2020). Cross-national comparisons of timeuse surveys show that the ratio of time spent on unpaid labor by women versus men in the MENA is among the highest in the world (Charmes 2015, 2019). The COVID-19 pandemic also increased women’s unpaid work and care burdens in the region (Hendy and Yassin 2022; Moghadam 2021; OECD 2020).\\n\\nAt the same time, norms around household task burden-sharing are in flux, and active efforts to change norms are underway. In Morocco, for example, recent anthropological research has explored how female labor force participation contributes to the renegotiation of expected household responsibilities (Bouasria 2013, 2017; Dike 2021). Community associations working on women’s empowerment in Morocco have also undertaken awareness campaigns focused on de-stigmatizing men’s contribution to household labor (Barnett 2022). During the early phases of the pandemic, Facebook posts in the country went viral highlighting how men could relieve the burdens of lockdown by contributing 4\", \"datasets\": [{\"mention_name\": {\"text\": \"timeuse surveys\", \"confidence\": 0.9178898334503174, \"start\": 1147, \"end\": 1162, \"clean_text\": \"timeuse surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1147, \"end\": 1162, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1147, \"end\": 1162, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1147, \"end\": 1162}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1147, \"end\": 1162, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9799541234970093, \"start\": 1147, \"end\": 1162}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9178898334503174, \"start\": 1147, \"end\": 1162}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5760555863380432, \"start\": 1147, \"end\": 1162}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5760555863380432, \"start\": 1147, \"end\": 1162}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"more to household chores and featuring men proudly showing their domestic contributions. More broadly, as Krafft, Selwaness, and Sieverding (2022) argue with respect to the MENA region as a whole, “the pandemic has underscored gender inequality in care work, but also offers an opportunity to rethink care responsibilities and ultimately recognize, reduce, and redistribute care work” (3).\\n\\nIn prior work (Barnett 2022, 2023), I collected the first evidence available on the relationship between personal attitudes and perceived norms on a range of gender issues in Morocco. One striking, unexpected finding was a large gap between what Moroccans say they value and what they believe others value when it comes to men’s contributions to unpaid domestic and care work. In a nationally representative survey, 75.6% of Moroccans agreed or strongly agreed that men are just as responsible for taking care of the household and children as women are, but only 54.7% believed most of their neighbors similarly agreed. To the extent that behavior aligns more with perceived norms than personal attitudes, such misperceptions may represent an unnecessary constraint on more equal household distributions of labor and ultimately on female labor force participation.\\n\\nPrevious research in OECD contexts has highlighted gender gaps in perceptions of how others value unpaid labor and the extent to which they see them as fair (see e.g. Cerrato and Cifre (2018)). We know little, however, about individuals’ perceptions of others’ attitudes toward and time spent on unpaid labor and care work in the MENA region, with some notable recent exceptions in studies of Egypt (Zeitoun et al. 2023) and Jordan (Gauri, Rahman, and Sen 2019). At the same time, recent research has shown that people in the MENA region and South Asia, especially men, generally tend to overestimate others’ embrace of patriarchal attitudes (Bursztyn, González, and YanagizawaDrott 2020; Gulzar, Sonnet, and Khan 2020; Minardi, Akmal, Crawfurd, and Hares 2021). One policy implication of these findings is that providing more accurate information and more positive narratives about existing changes in attitudes might correct misperceived norms, in turn altering how individuals behave (Bursztyn, González, and YanagizawaDrott 2020; Gauri, Rahman, and Sen 2019), although the long-term effects of information interventions remain uncertain (Prentice and Paluck 2020; Tankard and Paluck 2017).\\n\\n5\", \"datasets\": [{\"mention_name\": {\"text\": \"nationally representative survey\", \"confidence\": 0.6390868425369263, \"start\": 773, \"end\": 805, \"clean_text\": \"nationally representative survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 773, \"end\": 805, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 773, \"end\": 805, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 773, \"end\": 805}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 773, \"end\": 805, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9967240691184998, \"start\": 773, \"end\": 805}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6390868425369263, \"start\": 773, \"end\": 805}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7555463314056396, \"start\": 773, \"end\": 805}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7555463314056396, \"start\": 773, \"end\": 805}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"This project addresses these gaps and sheds light on these questions by gathering original data on Moroccans’ attitudes, experiences, and perceptions around unpaid labor, care work, and paid work outside the home. How do Moroccans’ attitudes and perceived norms around various aspects of unpaid labor and care work within the household relate to one another and to their actual behavior? To what extent do attitudes and perceived norms around household roles hinder the emergence of more gender-equal distributions of labor?\\n\\n# **3 Data**\\n\\n## **3.1 Survey Sample**\\n\\nThe original data discussed in this paper come from a survey of 1,038 Moroccan respondents recruited through Qualtrics from January 25 to March 3, 2023. Qualtrics maintains access to a pre-recruited pool of Moroccan respondents who complete surveys in exchange for a small financial incentive.<sup>2</sup> Demographic quotas based on gender, age, and geographic location were employed to ensure a baseline level of respondent diversity. Appendix Table 1 summarizes the demographics of the sample and compares it to that of a national probability face to face survey (F2F) conducted in 2022 by the Arab Barometer. The online sample disproportionately surveys individuals who are currently employed (58% versus 38% in the F2F survey), while under-representing students, the unemployed, housewives, and retired individuals. Online respondents are also more likely to self-identify as only somewhat religious or not religious, rather than religious. Despite higher employment rates, socio-economically, the online sample respondents identify themselves as struggling a bit more, on average, than respondents in the F2F survey, in response to a question about whether their household makes enough money to meet\\n\\n> 2. In one sense, all respondents in this survey are therefore “employed.” However, completing surveys for compensation likely takes up no more than a small portion of respondents’ time and can be completed from home. While housewives and the unemployed are underrepresented in the sample, around 19% of female respondents self-identified as primarily housewives, and 12.5% of all respondents self-identified as unemployed.\\n\\n6\", \"datasets\": [{\"mention_name\": {\"text\": \"survey of 1,038 Moroccan respondents\", \"confidence\": 0.8139963150024414, \"start\": 620, \"end\": 656, \"clean_text\": \"survey of 1,038 Moroccan respondents\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 620, \"end\": 656, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 620, \"end\": 656, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 620, \"end\": 656}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 620, \"end\": 656, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9992644190788269, \"start\": 620, \"end\": 656}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.8139963150024414, \"start\": 620, \"end\": 656}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9577112197875977, \"start\": 620, \"end\": 656}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9577112197875977, \"start\": 620, \"end\": 656}}, {\"mention_name\": {\"text\": \"national probability face to face survey\", \"confidence\": 0.5711915493011475, \"start\": 1091, \"end\": 1131, \"clean_text\": \"national probability face to face survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.6402161717414856, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1091, \"end\": 1131, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1091, \"end\": 1131}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1091, \"end\": 1131, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9989330172538757, \"start\": 273, \"end\": 313}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5711915493011475, \"start\": 273, \"end\": 313}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.44234296679496765, \"start\": 273, \"end\": 313}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.44234296679496765, \"start\": 273, \"end\": 313}}, {\"mention_name\": {\"text\": \"F2F survey\", \"confidence\": 0.5781145691871643, \"start\": 1677, \"end\": 1687, \"clean_text\": \"F2F survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1677, \"end\": 1687, \"clean_text\": \"\"}, \"producer\": {\"text\": \"Arab Barometer\", \"confidence\": 0.7393770217895508, \"start\": 1163, \"end\": 1177, \"clean_text\": \"Arab Barometer\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1677, \"end\": 1687}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1677, \"end\": 1687, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9782058000564575, \"start\": 859, \"end\": 869}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5781145691871643, \"start\": 859, \"end\": 869}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7529512047767639, \"start\": 859, \"end\": 869}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7529512047767639, \"start\": 859, \"end\": 869}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"expenses and/or save. The online survey is most unrepresentative in its over-inclusion of urban residents (92%, versus 65% in the F2F sample) and those with at least some higher education (43%, versus only 13% in the F2F sample). It is thus important to interpret the results of this survey as reflecting the attitudes and beliefs of a primarily urban, relatively educated sample. Most analyses (unless otherwise indicated) nonetheless adjust the data to better approximate a national probability sample using the weighting criteria followed by the Arab Barometer to adjust their surveys conducted in 2020-2021 via telephone, based on demographics in the 2014 Moroccan Census and the (face to face) 2018 Arab Barometer survey. Details of the weighting procedure are in the Appendix.\\n\\nThe pre-registered analysis plan for the survey<sup>3</sup> and the Appendix include details about the survey questionnaire and the questions used to construct the variables used in the analysis below. The following sections briefly introduce the relevant items’ wording, scales, and modification.\\n\\n# **3.2 Time Use Survey Items**\\n\\nA battery of questions on the survey asked respondents to indicate who in their household usually did each of the following tasks (in parentheses are the short names for each of these variables used in subsequent analyses and visualizations). Following recent work that has highlighted the importance of cognitive labor or the “mental load” as both distinct from physical labor and, in other settings, more likely to be undertaken predominantly by women (Catalano Weeks 2022; Daminger 2019), I included both physical and mental or managerial tasks: _Physical tasks_\\n\\n- Minor home repairs (fixing appliances, broken doors or windows, etc.) ( _repairs_ )\\n\\n- Buying groceries and other daily necessities ( _groceries_ )\\n\\n- Cleaning the house ( _cleaning_ )\\n\\n- Cooking ( _cooking_ )\\n\\n> 3. Available at https://osf.io/b6ntz/?view_only=1ac20cd71c714ea4bc7e8f43db6a5dd2.\\n\\n7\", \"datasets\": [{\"mention_name\": {\"text\": \"2014 Moroccan Census\", \"confidence\": 0.8859774470329285, \"start\": 655, \"end\": 675, \"clean_text\": \"2014 Moroccan Census\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 655, \"end\": 675, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 655, \"end\": 675, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 655, \"end\": 675}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 655, \"end\": 675, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.870012104511261, \"start\": 655, \"end\": 675}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8859774470329285, \"start\": 655, \"end\": 675}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5294492840766907, \"start\": 655, \"end\": 675}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5294492840766907, \"start\": 655, \"end\": 675}}, {\"mention_name\": {\"text\": \"2018 Arab Barometer survey\", \"confidence\": 0.8044085502624512, \"start\": 699, \"end\": 725, \"clean_text\": \"2018 Arab Barometer survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 699, \"end\": 725, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 699, \"end\": 725, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 699, \"end\": 725}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 699, \"end\": 725, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.870012104511261, \"start\": 699, \"end\": 725}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.8044085502624512, \"start\": 699, \"end\": 725}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5294492840766907, \"start\": 699, \"end\": 725}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5294492840766907, \"start\": 699, \"end\": 725}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"- Paying bills (note: included as both a physical and mental/managerial task) ( _bills_ )\\n\\n- Helping children with schoolwork ( _homework_ )\\n\\n- Caring for sick family members ( _caring_ )\\n\\n## _Mental/managerial tasks_\\n\\n- Planning what food to buy ( _food planning_ )\\n\\n- Planning the family budget ( _budget_ )\\n\\n- Paying bills ( _bills_ )\\n\\n- Planning clothing purchases ( _clothing_ )\\n\\n- Researching options for children’s education ( _education_ )\\n\\n- Deciding when someone should see a doctor ( _doctor_ )\\n\\nRespondents chose whether they, their spouse, they and their spouse equally, a maid or servant, their mother, their father, or someone else usually did each of the above tasks. “Paying bills” is included as both a physical and a mental/managerial tasks, as in Morocco, paying bills still frequently involves physically reporting to a business or postal office in order to make a payment. Only currently married respondents were asked who is primarily responsible for making decisions about children’s education and helping them with homework. Appendix Figures 1 to 3 display the distribution of responses for each item.\\n\\n# **3.3 Attitudes and Perceived Norms Survey Items**\\n\\nThe survey also asked respondents to indicate whether they strongly agree, agree, disagree, or strongly disagree with the following items (in parentheses are the shorthand names used to refer to these questions in subsequent analyses and visualizations):\\n\\n- Taking care of the home and children is a woman’s primary responsibility. ( _primary_ )\\n\\n- It is important for fathers to be involved in daily housework and childcare tasks (for example, tasks such as cleaning, feeding, bathing, and dressing). ( _daily_ )\\n\\n- When mothers work outside the home, children suffer. ( _suffer_ )\\n\\n- It can be good for the family when mothers work outside the home. ( _work good_ )\\n\\n8\", \"datasets\": [{\"mention_name\": {\"text\": \"Attitudes and Perceived Norms Survey\", \"confidence\": 0.6482285857200623, \"start\": 1136, \"end\": 1172, \"clean_text\": \"Attitudes and Perceived Norms Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1136, \"end\": 1172, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1136, \"end\": 1172, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1136, \"end\": 1172}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1136, \"end\": 1172, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9996042847633362, \"start\": 372, \"end\": 408}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6482285857200623, \"start\": 372, \"end\": 408}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9176012873649597, \"start\": 372, \"end\": 408}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9176012873649597, \"start\": 372, \"end\": 408}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"- A man should have the last word in household decision-making about finances. ( _last word_ )\\n\\n- Housewives should be paid for their household labor. ( _pay women_ )\\n\\n- Women should have as much opportunity as men to take a break from household responsibilities to spend time with friends. ( _leisure_ )\\n\\n- Men who take on more household tasks to support women pursuing careers are not very manly. ( _manly_ )\\n\\nResponses to these items constitute the personal attitudes (PA) indicators. I also ask whether respondents believe that (1) most Moroccan women, (2) most Moroccan men, and (3) “people in your neighborhood” currently strongly agree, agree, disagree, or strongly disagree with the same set of statements. Responses to these items constitute the perceived norms indicators. To avoid systematic question order effects, whether respondents were asked about personal attitudes or perceived norms first was randomized using Qualtrics’ built-in randomization, as was the order in which respondents were asked about their perceptions of each reference group, and the order in which the individual survey items were presented within each grouping.\\n\\nFor each respondent, I generate an aggregate personal attitudes score and separate perceived norms scores with respect to each reference group (men, women, and neighbors). First, I re-code responses to the items such that 0 = strongly agree and 3 = strongly disagree when the statement as written is a conservative viewpoint, and such that 0 = strongly disagree and 3 = strongly agree when the statement as written is an egalitarian viewpoint. Higher numerical values then represent more egalitarian viewpoints for all items. The scores are the mean responses to survey items after this re-coding.<sup>4</sup> For each respondent, I also calculate a pooled perceived norm for each individual item taking the average of respondents’ perceived norms of men, women, and neighbors. From this, I calculate an overall _pooled perceived norms score_ which is the average pooled perceived norm across all eight survey items. The mean values of these scores among male and female respondents, respectively, are visualized in Appendix Figures 6 to 8.\\n\\n> 4. All questions were forced choice (respondents had to select an answer to continue the survey), so there are no missing values on the personal attitudes or perceived norms items.\\n\\n9\", \"datasets\": [{\"mention_name\": {\"text\": \"personal attitudes (PA) indicators\", \"confidence\": 0.4731481969356537, \"start\": 452, \"end\": 486, \"clean_text\": \"personal attitudes (PA) indicators\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 452, \"end\": 486, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 452, \"end\": 486, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 452, \"end\": 486}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 452, \"end\": 486, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.9791257381439209, \"start\": 452, \"end\": 486}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4731481969356537, \"start\": 452, \"end\": 486}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9441908597946167, \"start\": 452, \"end\": 486}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9441908597946167, \"start\": 452, \"end\": 486}}, {\"mention_name\": {\"text\": \"perceived norms indicators\", \"confidence\": 0.30714234709739685, \"start\": 755, \"end\": 781, \"clean_text\": \"perceived norms indicators\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 755, \"end\": 781, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 755, \"end\": 781, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 755, \"end\": 781}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 755, \"end\": 781, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.991657018661499, \"start\": 755, \"end\": 781}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.30714234709739685, \"start\": 755, \"end\": 781}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.919306755065918, \"start\": 755, \"end\": 781}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.919306755065918, \"start\": 755, \"end\": 781}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"Following a brief informational intervention (discussed in section 5), respondents reviewed the same battery of items about the gender roles within and outside the home, asking whether respondents believe that (1) most Moroccan women and (2) most Moroccan men _will_ strongly agree, agree, disagree, or strongly disagree _five years from now_ . Future perceived norms scores were calculated from these responses in the same manner as described above for current perceived norms with respect to each reference group and are visualized in Appendix Figure 9.\\n\\n# **3.4 Demographic Covariates**\\n\\nIn all regression analyses below that control for demographic covariates, I include the following predictors (henceforth these are referred to as the “standard covariates”):\\n\\n- Dummy variable for male respondents (with female respondents as the reference category).\\n\\n- Age categories: 18-25 (reference category), 26-35, 36-45, 46-55, 55+.\\n\\n- Dummy variable for urban respondents (with rural respondents as the reference category).\\n\\n- Education categories: preparatory education or less (reference category); secondary or technical education; bachelor’s degree or higher.\\n\\n- Subjective socio-economic status: a variable taking the values 1-4 based on answers to a question about whether the family’s income covers necessary expenses and they can save, where a 4 indicates the greatest socio-economic stability.\\n\\n- Dummy variable for respondents who are currently employed, with all other respondent categories (housewives, students, retired, unemployed or other) combined as the reference category.\\n\\n- Dummy variable for currently married respondents (with all other respondents in the reference category).\\n\\n- Dummy variable for respondents with at least one child (with a reference category of no children).\\n\\n- Religiosity: a variable taking values from 1 (not religious) to 3 (religious) as selfreported.\\n\\n10\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"<!-- Start of picture text --> Women<br>Bills @ 079<br>Repairs 0.78<br>Budget 0.42<br>Groceries © 0.53<br>Education 0<br>Doctor 0.14<br>Homework 0.29 ®<br>Food Planning 02<br>Clothing 0.09 0.59)<br>Care for Sick 0.03 @<br>Dishes 0.02 0.79<br>Cleaning ——0.02<br>Cooking 0 6.85)<br><!-- End of picture text --> Primarily Self (Married)\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"<!-- Start of picture text --> Women<br>Cooking @ 0.86<br>Cleaning 0.03 0.78<br>Dishes 0.03 0.7<br>Food Planning 0.5<br>Care for Sick @ 0.26<br>Doctor 0.08 0.22<br>Homework 0.08 0.22<br>Clothing 0.07 0.21<br>Education @e<br>Groceries 0.16 025<br>Budget 0.05 0.24<br>Repairs 0.04<br>Bills 0.04 @<br>Primarily Spouse (Married)<br>Women<br>Care for Sick ® 0.65<br>Clothing © 0.69<br>Homework ® 0.45<br>Dishes ® 0.26<br>Cooking 0.04) 0.11<br>Food Planning (0 22<br>Cleaning Qs<br>Doctor Ghee<br>Bills &<br>Budget ®<br>Education o.65)<br>Groceries 0.33)<br>Repairs 0.16 0.24)<br>Primarily Equal (Married)<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"by the two spouses. On the other hand, for tasks where men report being personally responsible most of the time (paying bills, making repairs, shopping for groceries), women are more likely than men to claim that these tasks are handled equally by the two parties. It is possible both that men are systematically overestimating their contributions across all tasks (as has been argued with respect to similar findings in the United States, see Barroso (2021)), and that women may be underestimating men’s contributions on some tasks where men are taking on more responsibility.\\n\\nI construct a measure that combines responses to the questions about individual household tasks while taking into account respondents’ marital status. The variable I construct, _female household share_ , takes the following values:<sup>5</sup>\\n\\n- For married women, the share of tasks they report being primarily responsible for.\\n\\n- For married men, the share of tasks they report their spouses being primarily responsible for.\\n\\n- For never-married respondents, the share of tasks they report their mothers being primarily responsible for, unless the respondent indicates that he or she personally does the majority (>50%) of the task items, in which case they are excluded from the analysis. While respondents were not asked whether they still live with their parents, applying this restriction proxies for limiting the sample of never-married respondents to those living at home for the purposes of this analysis.\\n\\nSeparated, divorced, and widowed respondents are also excluded from analysis using this variable. These represent a small minority of the sample (3.7%). Figure 4 displays density plots comparing the distribution of the _female household share_ measure for male and female respondents, respectively, among both married and never-married respondents. Overall, a total of 150 respondents are excluded from this analysis either because they are single and report responsibility for the majority of household tasks, or because they are divorced, widowed, or separated.\\n\\n5. The structure of this composite variable was pre-registered.\\n\\n13\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"<!-- Start of picture text --> 4<br>3<br>> Gender<br>e 2 Men<br>fo)<br>a) Women<br>1<br>0<br>0.00 0.25 0.50 0.75 1.00<br>FHS (Married Respondents)<br><!-- End of picture text --> <!-- Start of picture text --> 4<br>3<br>> Gender<br>e 2 Men<br>fo)<br>(a) Women<br>1<br>0<br>0.00 0.25 0.50 0.75 1.00<br>FHS (Single Respondents)<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"Respondents also estimated the number of hours they (and their spouse, if applicable) spend on household activities per week; the share of physical household labor they and their partner complete (out of 100%); and the share of mental household labor (the “mental load”) they and their partner each complete (out of 100%). Only currently married respondents were asked about the household behavior of their spouse as it is rare for unmarried couples to live together in Morocco. The patterns in these responses are similar to those found when examining the composite female household share variable.<sup>6</sup> Combined with information on women’s employment status, these patterns show the extent of the double burden female respondents face. The average unweighted value of _female household share_ did not differ significantly between employed women (0.41) and women not currently working (0.43).<sup>7</sup> Even after controlling for other demographic covariates, there is no correlation between FHS and women’s current employment status, consistent with women’s participation in the labor market failing to reduce expectations that they will take responsibility for most tasks within the home. Women are also more likely than men to report feeling as though they face a substantial “double burden” of paid and unpaid work. The survey asked respondents who are currently employed to what extent household obligations “impede your ability to succeed at work and advance your career.” Overall, 13% of those employed agreed this was true to a great extent, and another 20% agreed it was true to a moderate extent. By gender, 28% of employed men viewed household obligations as a moderate or great impediment, versus 40% of employed women. Among the latter, a majority said it was a great impediment.<sup>8</sup> Female respondents were also asked whether, in an ideal world, they would work outside the home. Currently employed women were more likely than women not currently employed to say that their ideal would personally be to _not_ work (53% vs. 42%). Of the women not currently employed who stated that they would ideally be so, 32% stated their\\n\\n> 6. Results available upon request.\\n\\n> 7. The weighted averages are identical at 0.43.\\n\\n> 8. All percentages are weighted.\\n\\n15\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"education was the primary obstacle, followed by the labor market (31%), their children (12%), or other reasons (15%), with 6% citing their husbands and another 4% citing their families as the primary obstacle. There is thus a substantial minority for whom family objections or obligations (especially childcare) are an impediment to labor market participation, along with a slight majority of working women who would prefer not to be in the labor force—potentially because of the burden of household labor they struggle to escape.<sup>9</sup>\\n\\n# **5 Female Labor Burden Correlates with Perceived Male Attitudes**\\n\\nDoes this distribution of household labor reflect respondents’ preferences, their perceptions of social norms, or both? In this section, I show that the reported _female household share_ of labor introduced above does not correlate with respondents’ personal attitudes, but does correlate with respondents’ perceptions of men’s attitudes. These results are consistent with recent findings from Jordan that women’s labor force participation is correlated with expectations of husbands’ attitudes (Gauri, Rahman, and Sen 2019).\\n\\nI hypothesized ex ante that respondents living in households with a more traditional division of labor would have more conservative personal attitudes and would perceive more conservative norms.<sup>10</sup> To test this expectation, I assess the correlation between the female household share (FHS) variable and (1) the personal attitudes score and (2) the perceived norms scores by estimating weighted OLS regressions where the outcome is either the PA score or the relevant PN score (on the scale of 0-3, where 3 indicates the most egalitarian position), and the predictors are the FHS variable and the set of standard covariates described in the Data section. The hypothesis was that the coefficient on the\\n\\n> 9. The survey also asked currently employed female respondents who would prefer _not_ to work what compelled them to do so anyway. 92% stated their financial situation obliged them to work, and 8% cited other reasons. No respondent answered that their husband required them to work.\\n\\n> 10. See pre-registered hypothesis H2(c).\\n\\n16\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"FHS term would be negative: more egalitarian attitudes and perceived norms would be associated with a lower female household share of labor. Table 1 shows the result of these regressions, presenting only the coefficient for the FHS term. Full results are available in the Appendix.\\n\\nTable 1: Female Household Share, Attitudes, and Perceived Norms\\n\\n||Attitudes|Pooled PN|PN Men|PN Women|PN Neighbors|\\n|---|---|---|---|---|---|\\n|Female Household Share|_−_0_._009|_−_0_._128*|_−_0_._212**|_−_0_._052|_−_0_._114|\\n||(0_._075)|(0_._050)|(0_._066)|(0_._073)|(0_._082)|\\n|Num.Obs.|888|888|888|888|888|\\n|Covariates Included|Yes|Yes|Yes|Yes|Yes|\\n\\n+ p _<_ 0.1, * p _<_ 0.05, ** p _<_ 0.01, *** p _<_ 0.001 _Note:_ The table shows the coefficient estimated for the variable indicating the share of household tasks undertaken primarily by women (female household share) from weighted OLS regressions using models taking as the left-hand side variable, respectively, the personal attitudes score (1), the pooled perceived norms score (2), the perceived norms of men score (3), the perceived norms of women score (4), and the perceived norms of neighbors (5). Covariates are included in all models; full results are in the Appendix.\\n\\nContrary to expectations, I find that female household share is not significantly correlated with personal attitudes. That is, respondents who report living in homes where women are responsible for a higher share of household tasks do not report more conservative personal attitudes, on average, than those living in households where women are responsible for a lower share of tasks. However, FHS is negatively correlated with the pooled perceived norms score ( _p <_ 0 _._ 05). Respondents in homes where women are responsible for a higher share of household tasks perceive other Moroccans’ attitudes as more conservative, compared to respondents where women are responsible for a lower share of tasks. This relationship is driven by respondent perceptions of men’s attitudes, which are negatively correlated with FHS ( _p <_ 0 _._ 01). FHS is not associated with respondents’ perceptions of women’s attitudes or their neighbors’ attitudes. The same relationship holds when the future perceived norms scores are used to predict current FHS (see Appendix 17\", \"datasets\": [{\"mention_name\": {\"text\": \"personal attitudes score\", \"confidence\": 0.47591957449913025, \"start\": 956, \"end\": 980, \"clean_text\": \"personal attitudes score\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 956, \"end\": 980, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 956, \"end\": 980, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 956, \"end\": 980}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 956, \"end\": 980, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.47824639081954956, \"start\": 427, \"end\": 451}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.47591957449913025, \"start\": 427, \"end\": 451}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.600067675113678, \"start\": 427, \"end\": 451}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.600067675113678, \"start\": 427, \"end\": 451}}, {\"mention_name\": {\"text\": \"pooled perceived norms score\", \"confidence\": 0.4233416020870209, \"start\": 990, \"end\": 1018, \"clean_text\": \"pooled perceived norms score\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 990, \"end\": 1018, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 990, \"end\": 1018, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 990, \"end\": 1018}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 990, \"end\": 1018, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.47824639081954956, \"start\": 461, \"end\": 489}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.4233416020870209, \"start\": 461, \"end\": 489}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.600067675113678, \"start\": 461, \"end\": 489}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.600067675113678, \"start\": 461, \"end\": 489}}, {\"mention_name\": {\"text\": \"perceived norms of men score\", \"confidence\": 0.42075568437576294, \"start\": 1028, \"end\": 1056, \"clean_text\": \"perceived norms of men score\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1028, \"end\": 1056, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1028, \"end\": 1056, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1028, \"end\": 1056}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1028, \"end\": 1056, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.47824639081954956, \"start\": 499, \"end\": 527}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.42075568437576294, \"start\": 499, \"end\": 527}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.600067675113678, \"start\": 499, \"end\": 527}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.600067675113678, \"start\": 499, \"end\": 527}}, {\"mention_name\": {\"text\": \"perceived norms of women score\", \"confidence\": 0.48640426993370056, \"start\": 1066, \"end\": 1096, \"clean_text\": \"perceived norms of women score\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1066, \"end\": 1096, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1066, \"end\": 1096, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1066, \"end\": 1096}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1066, \"end\": 1096, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.47824639081954956, \"start\": 537, \"end\": 567}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.48640426993370056, \"start\": 537, \"end\": 567}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.600067675113678, \"start\": 537, \"end\": 567}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.600067675113678, \"start\": 537, \"end\": 567}}, {\"mention_name\": {\"text\": \"future perceived norms scores\", \"confidence\": 0.37253329157829285, \"start\": 2197, \"end\": 2226, \"clean_text\": \"future perceived norms scores\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2197, \"end\": 2226, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2197, \"end\": 2226, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2197, \"end\": 2226}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2197, \"end\": 2226, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.5103315114974976, \"start\": 369, \"end\": 398}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.37253329157829285, \"start\": 369, \"end\": 398}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8411967754364014, \"start\": 369, \"end\": 398}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8411967754364014, \"start\": 369, \"end\": 398}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"Table 3).\\n\\nThis correlation likely reflects a doubly causal relationship: deference to perceived male preferences influences the household distribution of labor, and observing this distribution in turn informs perceptions of what men—but not women—believe and value, highlighting that respondents see men’s norms as the relevant factor in how the household is organized.\\n\\nI also hypothesized that women who say they would ideally hold a job outside of the home would perceive more egalitarian current and future norms than women who say they would ideally be housewives.<sup>11</sup> To assess this expectation, I conduct a similar analysis as above, but in lieu of the female household share variable, the predictor of interest is the indicator for ideally wanting to work personally. I also examine whether men’s responses about the work status of the wife in their ideal household correlate with personal attitudes and perceived norms. Table 2 presents the estimates from both sets of models for the “ideally would work” variables. Full results are available in the Appendix.\\n\\nThe data provide strong support for the second hypothesis. For women, ideally working outside the home is positively correlated with more egalitarian personal attitudes and both current and future perceived norms with _p <_ 0 _._ 001. A preference for personally working outside the home is accompanied by beliefs—and perceptions that others also believe—that such work can be good for the family, that children do not suffer, that it is not \\\"un-manly\\\" of a spouse to help his wife succeed in her career, and more generally that men should be involved in household tasks. Among men, the results are the same: believing a wife would ideally work outside the home is correlated with more egalitarian personal attitudes around household labor and more egalitarian current and future perceived norms with _p <_ 0 _._ 05.\\n\\nRespondents’ attitudes and perceived norms around household roles are thus consistent with their preferences over women’s paid work status, but not with the current distribution of labor within the home. The alignment of FHS with perceptions of men’s attitudes, but not with respondents’ current attitudes, suggests there may be room for\\n\\n> 11. See pre-registered hypothesis H2(d).\\n\\n18\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"Table 2: Women’s Ideal Work Status, Attitudes, and Perceived Norms\\n\\n||Attitudes|Pooled PN|Future PN|\\n|---|---|---|---|\\n||_Women_|||\\n|Ideal: I would work|0_._397***|0_._127***|0_._181***|\\n||(0_._034)|(0_._023)|(0_._032)|\\n|Num.Obs.|514|514|514|\\n|Covariates Included|Yes|Yes|Yes|\\n||_Men_|||\\n|Ideal: Wife would work|0_._084*|0_._095**|0_._160***|\\n||(0_._040)|(0_._029)|(0_._037)|\\n|Num.Obs.|524|524|524|\\n|Covariates included|Yes|Yes|Yes|\\n\\n+ p _<_ 0.1, * p _<_ 0.05, ** p _<_ 0.01, *** p _<_ 0.001 _Note:_ The top section of the table shows the coefficient estimated for the variable indicating, for female respondents, that they would ideally work outside of the home, from weighted OLS regressions using models taking as the left-hand side variable, respectively, the personal attitudes score (1), the pooled perceived norms score (2), and the pooled future perceived norms score (3). The bottom section of the table presents the same, but for the variable indicating, for male respondents, that in their hypothetical ideal household, the wife would work outside the home. Covariates are included in all models; full results are in the Appendix.\\n\\n19\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"better aligning individual preferences and behavior by altering perceived norms. One potential reason for a gap between what Moroccans say they value with respect to the household division of labor and what they believe others value _currently_ is that the division of labor in other households is hard to observe. Providing information about how household behavior has shifted in recent years could lead respondents to update their expectations about how much men will contribute in the future and how social norms around their contributions will evolve.\\n\\nTo test this possibility, the survey included an experiment testing whether providing information about changes in men’s contributions to household labor would induce changes in respondents’ expectations about future perceived norms. After answering the questions about their personal attitudes and current perceived norms, respondents were randomly sorted with equal probability into one of three treatment groups.<sup>12</sup> Group 1 read information about how the COVID-19 pandemic increased the time Moroccan men spend on household labor. Group 2 read information about how the COVID-19 pandemic increased the time many Moroccans spend on household labor. Group 3 did not read any information. Full text of the information treatments is available in the Appendix.\\n\\nI hypothesized that providing information about the impact of COVID-19 on men’s behavior would increase Moroccans’ perceptions that norms five years from now will be more egalitarian and that these effects would be larger among men than among women.<sup>13</sup> 12. Simple randomization was applied using Qualtrics’ built-in randomization. Appendix table 6 shows results of a balance test confirming that there were no significant differences in the standard covariates, personal attitudes scores, or current perceived norms scores across treatment groups.\\n\\n13. See pre-registered hypotheses H4(a) and H4(b). The logic for the hypothesized gender difference was that men are relatively less likely to encounter discussions of gender equality in everyday life (as suggested by Bursztyn, González, and Yanagizawa-Drott (2020)). We would expect the treatment effects to be larger among men as encountering information should address a knowledge gap that they are likely to exhibit more than women do. For women, the information may not seem particularly revealing, or to the extent that it is, they may be more pessimistic than men (who have an interest in taking a positive view of their own sex) that a post-COVID increase in household contributions by men is likely to translate into sustained change. In addition, the information may be more likely to strike women as an incomplete picture of reality, given that time spent by women on household labor also increased during the pandemic, a fact of which women are more likely to be aware.\\n\\n20\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"<!-- Start of picture text --> 2.0<br>1.9<br>©<br>8<br>” Gender<br>Zz<br>OLL 1.87 bee bed eeeee e Men<br>c 4 Women<br>©<br>oO<br>=<br>1.7<br>1.6<br>Men All Moroccans No Info<br>Treatment<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"# **6 The Strength of the Male Breadwinner Model**\\n\\nPrevious sections established that (1) women face a disproportionate burden of unpaid household labor, yet married men and women have divergent perceptions about their respective contributions, with men more likely to claim tasks are undertaken equally, and (2) the female household share of labor is uncorrelated with respondents’ attitudes toward gender roles within the home, yet does correlate with respondents’ perceptions of men’s attitudes: respondents living in households with a higher share of female labor perceive that men hold more conservative attitudes. However, providing information about men spending more time on household tasks did not shift expectations about the gender roles Moroccans will value in the future.\\n\\nThis section further probes to what extent Moroccans are open, in principle, to a more equal distribution of household labor, using a conjoint survey experiment to jointly assess preferences over multiple dimensions of household organization. Conjoints are a survey technique in which the researcher simultaneously randomly varies multiple components of a set of profiles about which respondents then answer questions. The design allows researchers to manipulate multiple independent variables while estimating the causal effect of changing one variable’s value on the likelihood that a respondent will favor a profile. In a typical design, respondents make a binary forced choice about which of two alternative profiles they prefer, for example about hypothetical immigrants or political candidates (Abramson, Kocak, and Magazinnik 2022; Hainmueller, Hopkins, and Yamamoto 2014). Recent work has used this approach to measure constraints on women’s employment in the MENA region (Barnett, A. Jamal, and Monroe 2021; Blaydes, Gengler, and Lari 2021) and couples’ attitudes toward the division of household labor in the United States (Hutchinson, Khan, and Matfess 2022). Conjoints have been shown to have good validity in relationship to real-world behavior (Hainmueller, Hangartner, and Yamamoto 2015), and the design may mitigate social desirability bias in survey responses (Horiuchi, Markovich, and Yamamoto 2022). On this survey, respondents saw descriptions of two 22\", \"datasets\": [{\"mention_name\": {\"text\": \"conjoint survey experiment\", \"confidence\": 0.36254602670669556, \"start\": 921, \"end\": 947, \"clean_text\": \"conjoint survey experiment\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 921, \"end\": 947, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 921, \"end\": 947, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 921, \"end\": 947}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 921, \"end\": 947, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9899757504463196, \"start\": 921, \"end\": 947}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.36254602670669556, \"start\": 921, \"end\": 947}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8502680659294128, \"start\": 921, \"end\": 947}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8502680659294128, \"start\": 921, \"end\": 947}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"households in which the items in brackets below were independently randomized<sup>16</sup> (that is, for each attribute, one of the options in the brackets was randomly presented, with equal probability, and displayed in each profile):\\n\\n- The husband [is / is not] employed outside the home.\\n\\n- The wife [is / is not] employed outside the home.\\n\\n- Childcare is the primary responsibility of [the mother / the father / both adults equally / a grandparent / a maid or domestic servant].\\n\\n- Cleaning and cooking are the primary responsibility of [the mother / the father / both adults equally / a grandparent / a maid or domestic servant].\\n\\n- Home repairs are the primary responsibility of [the mother / the father / both adults equally / a grandparent / a maid or domestic servant].\\n\\n- Caring for ill or elderly family members is the primary responsibility of [the mother / the father / both adults equally / a grandparent / a maid or domestic servant].\\n\\nAfter reading descriptions of the two households, respondents indicated which of the two was closer to their ideal household, even if it was not their ideal (this is the outcome variable _choice_ ).<sup>17</sup> Respondents completed three rounds of this choice task. After doing so, they also responded to individual questions that directly assessed their preference over each of the above attributes in isolation (for example, would they prefer that cleaning and cooking are the primary responsibility of the mother, the father, both adults equally, a grandparent, or a maid or domestic servant).\\n\\nWith 1,038 respondents, there are 3,114 conjoint experiment responses. All analyses cluster standard errors by respondent. Figure 6 visualizes the marginal mean values\\n\\n> 16. Simple randomization of each attribute was accomplished via random sampling from arrays in Javascript code embedded in the survey, following a procedure shared by Graham (2020).\\n\\n> 17. After each choice task, I also asked respondents whether they preferred the family profile they selected, liked both profiles equally, or liked neither profile. I collapse the latter variable into a binary indicator for whether the respondent actually preferred the family profile they selected (this is the outcome variable _preferred_ ), the results for which are displayed in Appendix Figure 11.\\n\\n23\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"of each profile attribute level, or the overall probability of selecting a profile given the presence of that version of the attribute, averaging over other attributes. Marginal means are calculated using the _cregg_ package in R (Leeper 2022).<sup>18</sup> I focus on discussing when marginal mean estimates are statistically distinguishable from 0.5 (50%), which is the marginal mean in expectation in the forced-choice conjoint design when a respondent has no preference over levels of an attribute.\\n\\n18. I present the marginal means, rather than the commonly used average marginal component effects (AMCEs), in light the greater suitability of the marginal mean for discussions of respondent preferences (particularly for subgroup analyses) and the greater ease of interpreting them (Leeper, Hobolt, and Tilley 2019). AMCE estimates are included in the Appendix; the results are substantively similar.\\n\\n24\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"<!-- Start of picture text --> Works —— x<br>o<br>Does not work —o— 3<br>Works —e— =<br>Does not work —e— °<br>Mother —_e—<br>Maid —e— S<br>Grandparents —e— Q<br>Father —e— =<br>Equally a °<br>Mother ——.<br>Maid —_e— 2)<br>Grandparents ——e— 2<br>Father ——_e— =)<br>Equally —+— o<br>Mother ——e—<br>Maid ——_e—. pa<br>Grandparents —— 8<br>Father ——_e— a<br>Equally —e—<br>Mother —e—<br>Maid ——_e— @)<br>Grandparents —e—. £.<br>Father —e— @<br>Equally ——<br>0.4 0.5 0.6<br>Marginal Mean<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [27]}}, {\"page\": 28, \"chunk\": 28, \"input_text\": \"across the various tasks, although not to an extent statistically distinguishable from the favorability for primary female responsibility. Respondents are also indifferent about fathers taking primary responsibility for childcare, care of other family members, or making repairs, but they do disfavor fathers alone taking primary responsibility for cleaning and cooking (profiles in which the father is primarily responsible for cleaning and cooking are selected with probability 0.45).\\n\\nIn contrast, respondents do have a negative view of outsourcing household tasks to a maid or domestic servant. Profiles in which a maid is responsible for childcare, cooking and cleaning, repairs, or caring for the sick are all selected with probability 0.46. Negativity toward hired domestic help has potentially important implications for women’s ability to outsource existing unpaid and care labor in order to participate effectively in the labor market.\\n\\nBy far respondents’ strongest preference, among the attributes included, is that men should have a job outside the home. Profiles in which the husband works outside the home are selected with probability 0.63. While respondents disfavor profiles in which the wife works outside the home, the intensity of this preference is not as strong. Profiles in which the wife works are selected with probability 0.47.\\n\\nExamining the marginal mean responses disaggregated by respondent gender shows that the largest divide between men and women is over their preferences for whether married women would ideally work outside the home. Figure 7 displays the marginal mean preferences for the conjoint attributes among male and female respondents, respectively.\\n\\n26\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [28]}}, {\"page\": 29, \"chunk\": 29, \"input_text\": \"<!-- Start of picture text --> Works . =<br>Does S<br>not work 9 2<br>Works ° Ss<br>Does not work ° °<br>Mother —-e @)<br>Maid —_.—— =<br>Grandparents —__.—* a<br>Father —e ©<br>Equally —— >» o<br>Mother +—*—— @)<br>Maid ——_*e —— O<br>Grandparents — is<br>Father ——_*e —— 5<br>Equally —-—— Ko)<br>Mother _——. *— x<br>Maid on O<br>Grandparents _——*— 3<br>Father + —¥_e—— a<br>Equally ———_e———_.<br>Mother _t——*—<br>Maid —— &<br>Grandparents te =.<br>Father ——_e— = e<br>Equally ———_e———.<br>0.4 0.5 0.6<br>Marginal Mean<br><!-- End of picture text --> Gender -* Men Women\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [29]}}, {\"page\": 30, \"chunk\": 30, \"input_text\": \"the same time, both male and female respondents hold identical and strong preferences for men ideally working outside of the home. The marginal mean for the scenario in which the man works outside the home is 0.63 among both men and women.\\n\\nThe direct questions about these attributes that followed the conjoint experiment confirm the strength and consistency of this finding. Isolated from other factors, 87% of men and 97% of women answered that in their ideal household, the husband would work outside the home. In contrast, 61% of women and 82% of men answer directly that women not working outside the home is their ideal. The difference between these proportions and the marginal means calculated in the conjoint experiment show that preferences over men’s and women’s employment status may be moderated when individuals consider them in the context of preferences over the overall distribution of labor within the home. Nonetheless, the continued existence of such strong gendered preferences over employment status suggests that it is these preferences, more than specific objections to a less traditional division of household labor, that determine the distribution of responsibility for household tasks.\\n\\nIn response to direct questions, respondents also indicated a preference for scenarios in which men and women share household tasks equally. The modest positive effect associated with equal responsibility for tasks in the conjoint experiment provides some confidence that these claims are not merely the result of social desirability bias. Figure 8 shows the weighted proportion of male and female respondents, respectively, indicating that the mother, father, or mother and father equally would be responsible for each task (for simplicity, I omit other response options). Around 68% of men and women believe that husbands and wives would ideally take equal responsibility for childcare and caring for sick family members, compared to 29% who believe women alone would ideally handle childcare and 21% who believe women alone would ideally care for sick relatives. Thus, with respect to care work, many Moroccans embrace an ideal of equal burden sharing and participation by both men and women. However, among those who do not, the expectation remains that these are women’s responsibilities.\\n\\n28\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [30]}}, {\"page\": 31, \"chunk\": 31, \"input_text\": \"<!-- Start of picture text --> 0.8<br>0.6<br>0.4<br>0.2<br>c<br>2<br>£<br>2<br>oO<br>0.8<br>0.6<br>0.4<br>0.2<br>Equal Father Mother Equal Father Mother<br><!-- End of picture text --> Item Gender wl Men _ Women\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [31]}}, {\"page\": 32, \"chunk\": 32, \"input_text\": \"distribution of physical tasks like cooking, cleaning, and making repairs. Despite openness to more equal household contributions, the intensity of respondent preferences for male employment _outside_ the home limit the potential impact of this theoretical openness to greater household burden-sharing. The results suggest that women’s unpaid labor burden is primarily a side effect of preferences for a male breadwinner family model, rather than strong preferences over who does specific household tasks.\\n\\n# **7 Conclusion**\\n\\nThis paper began by asking to what extent attitudes and perceived norms (or normative expectations) around household roles hinder the emergence of more gender-equal distributions of labor in Morocco. As across the MENA region, women are disproportionately out of the paid labor force and in charge of most unpaid household and care tasks. Drawing on results from an original online survey of predominantly urban, relatively educated Moroccans, with embedded experiments, I found that individuals’ attitudes toward household duties themselves appear a relatively unimportant determinant of individual behavior in this context. Instead, both (a) perceptions of men’s attitudes toward household roles and (b) attitudes toward men’s roles outside the home—specifically, strong preferences among both men and women for a male breadwinner—appear the most important factors contributing to women shouldering a disproportionate unpaid labor burden and, often, a heavy double burden of paid and unpaid labor.\\n\\nRespondents acknowledge that women shoulder a higher share of the household burden of labor than men do. Yet men also take a rosy view of their contributions: married male respondents are more likely to estimate that their own contributions to unpaid labor are relatively equal to their spouses, while married women report being primarily responsible for tasks more than sharing them equally. Among single respondents, men also provide a lower estimate of the share of household labor undertaken primarily by their mothers (potentially because they are more likely to live independently), but neither male nor female single respondents cite fathers as contributing much at all. Despite these 30\", \"datasets\": [{\"mention_name\": {\"text\": \"online survey\", \"confidence\": 0.4273264408111572, \"start\": 902, \"end\": 915, \"clean_text\": \"online survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 902, \"end\": 915, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 902, \"end\": 915, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 902, \"end\": 915}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 902, \"end\": 915, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9971363544464111, \"start\": 9, \"end\": 22}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4273264408111572, \"start\": 9, \"end\": 22}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8081192374229431, \"start\": 9, \"end\": 22}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8081192374229431, \"start\": 9, \"end\": 22}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [32]}}, {\"page\": 33, \"chunk\": 33, \"input_text\": \"differences, a majority of male and female respondents alike respond that care tasks especially (looking after children, caring for sick relatives) would ideally be undertaken by men and women equally. Among those who feel otherwise, however, women are seen as ideally responsible for these tasks. A conjoint experiment varying the responsibility for different tasks in a hypothetical household also revealed that respondents appear indifferent between women, men, or women and men equally taking responsibility for childcare, care of the sick, and minor repairs, although they continue to look unfavorably on men taking primary responsibility for cleaning and cooking.\\n\\nThis openness to more egalitarian household roles in principle does not seem to translate substantially into the actual reorganization of household responsibilities. The female household share of labor does not vary between employed and unemployed women. Nor is it correlated with respondents’ personal attitudes on a battery of questions about gender roles within the household. However, the female household share of labor does correlate with what respondents perceive men to believe and value. Respondents with more conservative perceptions of men’s attitudes tend to report a higher female burden of labor within the home. This likely reflects a doubly causal relationship: respondents exposed to men with more conservative attitudes align household roles with the perceived preferences of the patriarchal authorities within their households, and respondents who observe a higher female burden of labor likely conclude that this distribution exists because it is what men prefer. Perceptions of what men value may not be easy to change. An information experiment testing whether providing respondents with information about men’s increased contributions to household labor during COVID-19 did not have any effect on respondents’ expectations of future social norms.\\n\\nMost notably, however, both the conjoint experiment and direct survey questions revealed that (a) both men and women strongly prefer that in an ideal household, the husband will work outside the home, and (b) men, but not women, continue to have a strong preference for the wife not working outside the home in an ideal household. Around half of the women in the sample—including half of those currently employed— also believed that a woman would ideally not work outside the home.\\n\\n31\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [33]}}, {\"page\": 34, \"chunk\": 34, \"input_text\": \"These results may reflect the strength of _benevolent sexism_ in Morocco: that is, “a set of interrelated attitudes toward women that are sexist in terms of viewing women stereotypically and in restricted roles but that are subjectively positive in feeling (for the perceiver)” (Glick and Fiske 1996, 491). One manifestation of this is a valorization women’s traditional roles while emphasizing that women deserve help at home from supportive partners. In prior work, I found that 94% of Moroccans (equally among men and women) agreed or strongly agreed with the statement that “men should guard and cherish women,” while 76% of women and 73% of men agreed or strongly agreed that “men should sacrifice to provide for the women in their lives” (Barnett 2022).<sup>19</sup> Other recent research has found that while hostile sexist beliefs have declined substantially in the MENA region, benevolent sexist beliefs remain prevalent and are often called upon to justify expansions in women’s formal rights (Jones, Mitchell, and Martin 2021; Maitner and Henry 2018). The strength of benevolent sexism may account for both the relative interest in “equalizing” household contributions (and men’s interest in emphasizing the extent to which they currently contribute) co-existing with the continued preference men hold for women not working, where feasible. Thus, while openness to more equal divisions of household labor could bode well for women’s potential labor market participation, such expressions of interest in supporting women’s work may be conditional on women’s continued embrace of distinct gender roles. Alongside structural obstacles to women’s employment, the combination of benevolent sexism and commitment to the male breadwinner model may jointly form an equilibrium that makes shifting either women’s labor force participation or the actual distribution of household labor especially challenging.\\n\\nThe results also suggest that what may matter most for altering the distribution of labor is not abstract attitudes toward gender roles, but men’s reactions to their own life circumstances and how these intersect with conceptions of masculinity. Research from other contexts has shown that high-earning women may compensate for their violation of gender role norms by taking on a disproportionate share of household labor (Bittman,\\n\\n> 19. These figures are not reported in the dissertation but are calculated from responses to the telephone survey described therein.\\n\\n32\", \"datasets\": [{\"mention_name\": {\"text\": \"telephone survey\", \"confidence\": 0.38428035378456116, \"start\": 2443, \"end\": 2459, \"clean_text\": \"telephone survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2443, \"end\": 2459, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2443, \"end\": 2459, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2443, \"end\": 2459}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2443, \"end\": 2459, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9970353841781616, \"start\": 865, \"end\": 881}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.38428035378456116, \"start\": 865, \"end\": 881}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.492590993642807, \"start\": 865, \"end\": 881}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.492590993642807, \"start\": 865, \"end\": 881}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [34]}}, {\"page\": 35, \"chunk\": 35, \"input_text\": \"England, Sayer, Folbre, and Matheson 2003) or face higher rates of intimate partner violence (Krishnan, Rocca, Hubbard, Subbiah, Edmeades, and Padian 2010), and their relationships may also be more likely to end in divorce (Bertrand, Kamenica, and Pan 2015). Male psychological distress is elevated when men are solo breadwinners, but reaches its peak when men are economically dependent on their wives (Syrda 2020). In Morocco, researchers have argued that many women see protection of male dignity as a source of security for themselves and their communities (El Hajjami 2018) and have emphasized the pressure men place on themselves to fulfill the expected social role of family provider (El Feki, Heilman, and Barker 2017). The high value this study’s participants placed on men’s employment outside the home, relative to their preferences over who does which household tasks, suggests the maintenance of male dignity with respect to breadwinner expectations is relevant not only for whether women participate in market labor in the first place, but for how the distribution of household labor evolves—or not—when they do. More direct engagement with norms of masculinity may be essential to alleviating women’s double burden and ultimately facilitating their participation in paid labor on more equal terms.\\n\\n33\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099822202052424136/pdf/IDU18821daf01292214ebc1865c19a2c631d7d2f.pdf\", \"pages\": [35]}}]}"}}},{"rowIdx":99,"cells":{"doc_id":{"kind":"string","value":"34254016"},"doc":{"kind":"string","value":"{\"metadata\": {\"id\": \"34254016\", \"last_modified_date\": \"2024-02-07T00:00:00Z\", \"admreg\": \"Other,Other\", \"admreg_key\": \"912972,912972\", \"authors\": {\"0\": {\"author\": \"Kilic,Talip\"}, \"1\": {\"author\": \"Koolwal,Gayatri B.\"}, \"2\": {\"author\": \"Vundru,Wilbert Drazi\"}, \"3\": {\"author\": \"Daum,Thomas Lothar Georg\"}, \"4\": {\"author\": \"Buchwald,Hannes\"}, \"5\": {\"author\": \"Seymour,Greg\"}, \"6\": {\"author\": \"Mvula,Peter Mathias\"}, \"7\": {\"author\": \"Munthali,Alister Chaundumuka\"}, \"8\": {\"author\": \"Kachinjika,Monice\"}}, \"count\": \"World\", \"count_key\": \"517191\", \"docna\": {\"0\": {\"docna\": \"Recording the Time Divide : A Comparative\\n Study of Smartphone- and Recall-Based Approaches to Time Use Measurement\"}}, \"docty\": \"Policy Research Working Paper\", \"docty_key\": \"620265\", \"owner\": \"Living Standards Measurement (DECLS),Development Data Group (DECDG)\", \"projn\": \"1W-Measuring Time Use Well -- P170137\", \"trustfund\": \"TF0B0481-KCPIII - Measuring Time Use Well\", \"trustfund_key\": \"1770222\", \"prdln\": \"Research Activity\", \"prdln_key\": \"1582145\", \"seccl\": \"Public\", \"lang\": \"English\", \"lang_key\": \"120701\", \"entityids\": {\"entityid\": \"34254016\"}, \"repnb\": \"WPS10695\", \"docdt\": \"2024-02-07T00:00:00Z\", \"datestored\": \"2024-02-07T06:14:10Z\", \"volnb\": \"1\", \"majdocty\": \"Publications & Research\", \"majdocty_key\": \"658101\", \"abstracts\": {\"cdata!\": \"Based on a randomized survey experiment\\n in Malawi, this study examines how innovative techniques in\\n time use data collection could sidestep measurement concerns\\n with traditional recall-based time use measurement. The\\n experiment assigns random samples of households, and adult\\n men and women within, to one of two treatment arms on time\\n use measurement: a traditional 24-hour recall time use\\n diary, and a self-administered smartphone-based pictorial\\n time diary, known as the TimeTracker app, for real-time data\\n collection. Compared to the recall arm, participation in\\n employment and unpaid domestic and care work is shown to be\\n higher in the smartphone arm for both men and women. The\\n resulting estimates of gender gaps, while continuing to be\\n large, are narrower in the smartphone arm, except for care\\n work where the estimated gender gap increases. The recall\\n treatment leads to substantial underreporting of activities\\n after 6 pm, which otherwise accounts for nearly 30 percent\\n of daily reported time in the smartphone arm. Likewise, the\\n extent of simultaneous activities, particularly among women,\\n is markedly lower in the recall arm. The overall reported\\n time is, however, higher in the recall arm due to the\\n minimum 15-minute duration that was used for recording\\n activities the 24-hour recall diary, while over one-third of\\n activities lasted less than 15 minutes in the smartphone\\n arm. The analysis also shows that using stylized time use\\n questions with a 7-day recall, as opposed to a 24-hour\\n recall diary, results in an even greater overestimation of\\n reported time in employment and unpaid work.\"}, \"colti\": \"Policy Research working paper ; no. WPS\\n 10695; LSMS; KCP; PEOPLE\", \"display_title\": \"Recording the Time Divide : A\\n Comparative Study of Smartphone- and Recall-Based Approaches\\n to Time Use Measurement\", \"disclosure_date\": \"2024-02-07T00:00:00Z\", \"disclosure_type\": \"NA\", \"pdfurl\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"txturl\": \"http://documents.worldbank.org/curated/en/099814102072416574/text/IDU195d006a6196d314cc21a96310b5efaaff3cd.txt\", \"listing_relative_url\": \"/research/2024/02/34254016/recording-time-divide-comparative-study-smartphone--recall-based-approaches-time-use-measurement\", \"url_friendly_title\": \"http://documents.worldbank.org/curated/en/099814102072416574/IDU195d006a6196d314cc21a96310b5efaaff3cd\", \"new_url\": \"2024/02/34254016/IDU195d006a6196d314cc21a96310b5efaaff3cd\", \"ext_pub_date\": \"2024-02-07T00:00:00Z\", \"disclstat\": \"Disclosed\", \"topicv3\": \"Gender and Economic Empowerment m1327622 911,Unpaid Work m1504003 2701,Household Surveys m1504071 2761,Smart Phone m1505676 3093\", \"docm_id\": \"IDU-95d006a6-96d3-4cc2-a963-0b5efaaff3cd\", \"chronical_docm_id\": \"IDU-95d006a6-96d3-4cc2-a963-0b5efaaff3cd\", \"publishtoextweb_dt\": \"2024-02-07T00:00:00Z\", \"dois\": \"10.1596/1813-9450-10695\", \"versiontyp\": \"Final\", \"versiontyp_key\": \"1309935\", \"projectid\": \"P170137\", \"guid\": \"099814102072416574\", \"origu\": \"Off of Sr VP Dev Econ/Chief Econ (DECVP)\", \"no_of_pages\": \"47\", \"author_keys\": {\"0\": {\"author_key\": \"886702\"}, \"1\": {\"author_key\": \"788145\"}, \"2\": {\"author_key\": \"1379237\"}, \"3\": {\"author_key\": \"2123572\"}, \"4\": {\"author_key\": \"2136822\"}, \"5\": {\"author_key\": \"1208849\"}, \"6\": {\"author_key\": \"734305\"}, \"7\": {\"author_key\": \"1625965\"}, \"8\": {\"author_key\": \"NA\"}}, \"available_in\": \"English\", \"fullavailablein\": [], \"url\": \"http://documents.worldbank.org/curated/en/099814102072416574/IDU195d006a6196d314cc21a96310b5efaaff3cd\"}, \"model_extractions\": [{\"page\": 0, \"chunk\": 0, \"input_text\": \"Policy Research Working Paper 10695\\n\\n# Recording the Time Divide\\n\\nA Comparative Study of Smartphone- and Recall-Based Approaches to Time Use Measurement _Talip Kilic Gayatri Koolwal Wilbert Drazi Vundru Thomas Daum Hannes Buchwald Greg Seymour Peter Mvula Alister Munthali Monice Kachinjika_ Development Economics Development Data Group February 2024\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [0]}}, {\"page\": 1, \"chunk\": 1, \"input_text\": \"Policy Research Working Paper 10695\\n\\n# **Abstract**\\n\\nBased on a randomized survey experiment in Malawi, this study examines how innovative techniques in time use data collection could sidestep measurement concerns with traditional recall-based time use measurement. The experiment assigns random samples of households, and adult men and women within, to one of two treatment arms on time use measurement: a traditional 24-hour recall time use diary, and a self-administered smartphone-based pictorial time diary, known as the _TimeTracker_ app, for real-time data collection. Compared to the recall arm, participation in employment and unpaid domestic and care work is shown to be higher in the smartphone arm for both men and women. The resulting estimates of gender gaps, while continuing to be large, are narrower in the smartphone arm, except for care work where the estimated gender gap increases. The recall treatment leads to substantial underreporting of activities after 6 pm, which otherwise accounts for nearly 30 percent of daily reported time in the smartphone arm. Likewise, the extent of simultaneous activities, particularly among women, is markedly lower in the recall arm. The overall reported time is, however, higher in the recall arm due to the minimum 15-minute duration that was used for recording activities the 24-hour recall diary, while over one-third of activities lasted less than 15 minutes in the smartphone arm. The analysis also shows that using stylized time use questions with a 7-day recall, as opposed to a 24-hour recall diary, results in an even greater overestimation of reported time in employment and unpaid work.\\n\\nThis paper is a product of the Development Data Group, Development Economics. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at tkilic@worldbank.org and gkoolwal@worldbank.org.\\n\\n_The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent._ Produced by the Research Support Team\", \"datasets\": [{\"mention_name\": {\"text\": \"24-hour recall diary\", \"confidence\": 0.31737497448921204, \"start\": 1336, \"end\": 1356, \"clean_text\": \"24-hour recall diary\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1336, \"end\": 1356, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1336, \"end\": 1356, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"24-hour\", \"confidence\": 0.4269960820674896, \"start\": 1336, \"end\": 1343}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1336, \"end\": 1356, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.5746146440505981, \"start\": 469, \"end\": 489}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.31737497448921204, \"start\": 469, \"end\": 489}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9431571960449219, \"start\": 469, \"end\": 489}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9431571960449219, \"start\": 469, \"end\": 489}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [1]}}, {\"page\": 2, \"chunk\": 2, \"input_text\": \"Recording the Time Divide: A Comparative Study of Smartphone- and Recall-Based Approaches to Time Use Measurement Talip Kilic<sup>a</sup> , Gayatri Koolwal<sup>a</sup> , Wilbert Drazi Vundru<sup>a</sup> , Thomas Daum<sup>b</sup> , Hannes Buchwald<sup>c</sup> , Greg Seymour<sup>d</sup> , Peter Mvula<sup>e</sup> , Alister Munthali<sup>e</sup> , Monice Kachinjika<sup>e1</sup> JEL Codes: C83, J16, J22.\\n\\nKeywords: Time Use, Employment, Unpaid Work, Recall Bias, Smartphone Time Use Diary, Survey Methodology, Household Surveys, Gender, Malawi, Sub-Saharan Africa.\\n\\n> 1 Corresponding authors: Talip Kilic <u>(tkilic@worldbank.org) and Gayatri Koolwal (gkoolwal@worldbank.org).</u> a World Bank; b University of Gothenburg;<sup>c</sup> Independent Consultant;<sup>d</sup> International Food Policy Research Institute;<sup>e</sup> Palm Consulting. The Malawi Methodological Experiment on Time Use Measurement was implemented under the World Bank Living Standards Measurement Study (LSMS) program, in collaboration with the Measures for Advancing Gender Equality (MAGNET) initiative and PALM Consulting, the latter of whom implemented the fieldwork. The study benefited from funding from the International Fund for Agricultural Development, the World Bank Knowledge for Change Program, and the World Bank Umbrella Facility for Gender Equality, in partnership with the Bill and Melinda Gates Foundation. The authors are grateful to Andrey Prokhorenko (software programming) and Natalis Lorenz (illustrations) who were instrumental in designing the version of the _Time Tracker_ smartphone app that was used for this study; and Diego Zardetto for assistance with the sampling design. We thank Nancy Folbre and Hazel Malapit for valuable comments on the paper, as well as feedback from the participants at the 2023 International Food Policy Research Institute-RISE meeting, the Ninth International Conference of Agricultural Statistics (ICAS), the University of Maryland Conference on Measuring Time Use and Wellbeing, and the World Bank Development Data Group Let’s Talk Data Seminar.\", \"datasets\": [{\"mention_name\": {\"text\": \"Smartphone Time Use Diary\", \"confidence\": 0.7275416851043701, \"start\": 461, \"end\": 486, \"clean_text\": \"Smartphone Time Use Diary\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 461, \"end\": 486, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 461, \"end\": 486, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 461, \"end\": 486}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 461, \"end\": 486, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.920418918132782, \"start\": 461, \"end\": 486}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.7275416851043701, \"start\": 461, \"end\": 486}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7859025001525879, \"start\": 461, \"end\": 486}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7859025001525879, \"start\": 461, \"end\": 486}}, {\"mention_name\": {\"text\": \"Household Surveys\", \"confidence\": 0.43210339546203613, \"start\": 508, \"end\": 525, \"clean_text\": \"Household Surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 508, \"end\": 525, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 508, \"end\": 525, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 508, \"end\": 525}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 508, \"end\": 525, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.920418918132782, \"start\": 508, \"end\": 525}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.43210339546203613, \"start\": 508, \"end\": 525}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7859025001525879, \"start\": 508, \"end\": 525}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7859025001525879, \"start\": 508, \"end\": 525}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [2]}}, {\"page\": 3, \"chunk\": 3, \"input_text\": \"# 1. Background\\n\\nRepresentative surveys on individuals’ time use are important for understanding economic decision-making among men and women and highlighting broader inequalities within and across households. Effective design, targeting and evaluation of policies/projects related to workforce development, childcare provision, and improvements in transportation services, among others, also require an accurate measurement of men’s and women’s time use and time use disparities across different sub-populations.\\n\\nAvailable time use surveys reveal cross-country gender disparities in time allocations to different activities, including unpaid domestic and care work, which are often missed in labor force surveys and socioeconomic household surveys (Charmes, 2019; Hirway, 2017). While time use data collection has been implemented in varying formats across countries – either as part of stand-alone time use surveys or part of a time use module that is part of a labor force survey or a socioeconomic household survey, time use data collection in low- and middle-income contexts relies either on an interviewer-assisted recall-based time use diary – usually for the last 24 hours, or recall-stylized questions about time allocation to relatively aggregate groups of activities over a given reference period, which can be the last day, the last 7 days or the last 30 days (UN Women, 2021).<sup>2</sup> These survey design choices have raised concerns regarding potential cognitive errors in recalling time use (Seymour et al., 2020), particularly in contexts with lower literacy and numeracy and where economic activities are less predictable, including in countries where smallholder agriculture is prevalent (Arthi et. al, 2018; Field et al., 2023).<sup>3</sup> Activities conducted in tandem may also be more difficult to recall, depending on how often an activity occurs (regularity) and how distinguishable it is from other activities (saliency) (Menon, 1993; Brown et al., 2007). These errors can occur over shorter reference periods as well (Sudman and Bradburn, 1973).\\n\\nAlthough these measurement challenges might have direct implications for our understanding of gender disparities in time use (Floro and Miles, 2003; Budig and Folbre, 2004) and analytical applications that rely on time use data, including the design and evaluation of gender-sensitive policies to improve economic opportunities and well-being for men and women (World Economic Forum, 2021; King et al., 2021), there is scant evidence on how relying on recall-based survey methods may impact accuracy and reliability of time use data. Recall-based diaries can also be time-consuming to implement, and recent work (Field et al., 2023, ILO, 2023; Folbre, 2021) has discussed the role of streamlined time use diaries (by reducing or aggregating the number of activity categories, for example) as an alternative to more intensive data collection.\\n\\n> 2 Within these approaches, however, there is substantial variation across countries and years — including, for example, the choice of activity sets over which to report; the length of the reference period; the reliance on self- versus proxy reporting; the extent of interviewer assistance; and whether time use data collection is designed as a standalone survey or is part of an existing household survey (see UN Women, 2021, for a discussion). Open-ended diaries without pre-defined time intervals have also been used, but rarely (see Chatzitheochari et al.,2018, for a discussion).\\n\\n> 3 Farming follows an inherently irregular schedule that is often determined by factors outside the control of the farmer, such as climate variability, and the specific tasks involved in farming are often very repetitive (e.g., weeding, harvesting, and sorting).\\n\\n2\", \"datasets\": [{\"mention_name\": {\"text\": \"labor force surveys\", \"confidence\": 0.4892180263996124, \"start\": 694, \"end\": 713, \"clean_text\": \"labor force surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 694, \"end\": 713, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 694, \"end\": 713, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 694, \"end\": 713}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 694, \"end\": 713, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9747807383537292, \"start\": 694, \"end\": 713}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4892180263996124, \"start\": 694, \"end\": 713}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7550126314163208, \"start\": 694, \"end\": 713}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7550126314163208, \"start\": 694, \"end\": 713}}, {\"mention_name\": {\"text\": \"socioeconomic household surveys\", \"confidence\": 0.4919305443763733, \"start\": 718, \"end\": 749, \"clean_text\": \"socioeconomic household surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 718, \"end\": 749, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 718, \"end\": 749, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 718, \"end\": 749}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 718, \"end\": 749, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9747807383537292, \"start\": 718, \"end\": 749}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4919305443763733, \"start\": 718, \"end\": 749}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7550126314163208, \"start\": 718, \"end\": 749}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7550126314163208, \"start\": 718, \"end\": 749}}, {\"mention_name\": {\"text\": \"time use surveys\", \"confidence\": 0.7315433621406555, \"start\": 901, \"end\": 917, \"clean_text\": \"time use surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 901, \"end\": 917, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 901, \"end\": 917, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 901, \"end\": 917}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 901, \"end\": 917, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9744203090667725, \"start\": 901, \"end\": 917}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7315433621406555, \"start\": 901, \"end\": 917}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.683563232421875, \"start\": 901, \"end\": 917}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.683563232421875, \"start\": 901, \"end\": 917}}, {\"mention_name\": {\"text\": \"time use data\", \"confidence\": 0.9226007461547852, \"start\": 2598, \"end\": 2611, \"clean_text\": \"time use data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2598, \"end\": 2611, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2598, \"end\": 2611, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2598, \"end\": 2611}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2598, \"end\": 2611, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9727904200553894, \"start\": 860, \"end\": 873}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9226007461547852, \"start\": 860, \"end\": 873}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7520416975021362, \"start\": 860, \"end\": 873}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7520416975021362, \"start\": 860, \"end\": 873}}, {\"mention_name\": {\"text\": \"household survey\", \"confidence\": 0.624846875667572, \"start\": 3312, \"end\": 3328, \"clean_text\": \"household survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3312, \"end\": 3328, \"clean_text\": \"\"}, \"producer\": {\"text\": \"UN Women\", \"confidence\": 0.3724345862865448, \"start\": 3334, \"end\": 3342, \"clean_text\": \"UN Women\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3312, \"end\": 3328}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3312, \"end\": 3328, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.971609890460968, \"start\": 23, \"end\": 39}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.624846875667572, \"start\": 23, \"end\": 39}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.579433023929596, \"start\": 23, \"end\": 39}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.579433023929596, \"start\": 23, \"end\": 39}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [3]}}, {\"page\": 4, \"chunk\": 4, \"input_text\": \"Against this background, we present new findings from a large-scale randomized survey experiment conducted in Malawi to document the relative accuracy of recall-based approaches to time use survey data collection. <mark>The experiment assigns random samples of households, and adult men and women within, to one of two treatment arms on time use measurement: a traditional 24-hour recall time use diary; and a self-administered smartphone-based pictorial time diary, known as the</mark> _<mark>TimeTracker</mark>_ <mark>app.</mark> While real-time data collection on time use has been explored in other surveys in higher-income settings (Gershuny et al., 2020), our study is the largest one implemented to date in a low-income country, extending a nascent evidence base regarding the feasibility and utility of smartphone-based time use data collection - previously only available for Zambia, though for a significantly smaller, rural-only sample (Daum et al., 2018, 2019). Beyond the methodological contributions of our work, we reveal new, policy-relevant insights regarding intra-day variation and simultaneity in time use.\\n\\nThe analysis compares men’s and women’s self-reported time allocation across two survey treatment arms, over an 11-day reporting period, to provide a disaggregated, comparative analysis of real-time and recallbased survey approaches over shorter (24-hour) and longer (7-day) recall periods:\\n\\n- (1) In the first treatment arm, one adult man and adult woman (aged 15-64) in each household received an Android smartphone-based pictorial time use diary app called _TimeTracker_<sup>4</sup> to selfreport their time use across a set of activities on a continuous, real-time basis over the period. The app was developed specifically for time use data collection in low-literacy settings (Daum et al., 2018) where respondents do not necessarily have clock-based concepts of time and no prior experience with smartphones.\\n\\n- (2) In the second treatment arm, one adult man and adult woman in each household received a traditional, recall-based 24-hour diary interview on three different days during the reporting period, where interviewers asked respondents to recall their time allocation over the last 24 hours and coded the reported activities into 15-minute intervals based on the respondent’s narrative, with up to two simultaneous activities that can be recorded for each 15-minute interval. Reporting in the recall diary arm covered the same activity set as in the smartphone arm.\\n\\n- (3) Finally, at the end of the reporting period, respondents in both treatment arms received a stylized interview-based 7-day recall module covering a subset of activities in the smartphone/recall diary approaches. The 7-day module follows a structure commonly implemented in labor force surveys or socioeconomic household surveys (Seymour et al., 2020), and allows us to investigate how the same groups varied their responses over recall periods of different lengths.\\n\\nOur findings span three key areas, with important contributions to the literature. First, on participation and time across different activities, we show that the smartphone and recall arms provide very different perspectives, along with an interesting dichotomy in activity sets highlighted across the two arms. Recall, for example, leads to an underreporting of participation in employment and unpaid work, including a lower\\n\\n> 4 _Time Tracke_ r was originally developed by the University of Hohenheim and the Institute for Applied Science (University of Media, Stuttgart) (Daum et al., 2018, 2019) and enhanced for this study with World Bank LSMS support. While the app is not yet publicly available, there is a plan is to release it for public use.\\n\\n3\", \"datasets\": [{\"mention_name\": {\"text\": \"time use survey data\", \"confidence\": 0.45561572909355164, \"start\": 181, \"end\": 201, \"clean_text\": \"time use survey data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 181, \"end\": 201, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 181, \"end\": 201, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 181, \"end\": 201}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 181, \"end\": 201, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9986162185668945, \"start\": 181, \"end\": 201}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.45561572909355164, \"start\": 181, \"end\": 201}, \"usage_context\": {\"text\": \"background\", \"confidence\": 0.830862820148468, \"start\": 181, \"end\": 201}, \"is_used\": {\"text\": \"False\", \"confidence\": 0.830862820148468, \"start\": 181, \"end\": 201}}, {\"mention_name\": {\"text\": \"time use data\", \"confidence\": 0.4752330780029297, \"start\": 1759, \"end\": 1772, \"clean_text\": \"time use data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1759, \"end\": 1772, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1759, \"end\": 1772, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1759, \"end\": 1772}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1759, \"end\": 1772, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.6531230807304382, \"start\": 145, \"end\": 158}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4752330780029297, \"start\": 145, \"end\": 158}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.4924183785915375, \"start\": 145, \"end\": 158}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4924183785915375, \"start\": 145, \"end\": 158}}, {\"mention_name\": {\"text\": \"labor force surveys\", \"confidence\": 0.6607365012168884, \"start\": 2787, \"end\": 2806, \"clean_text\": \"labor force surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2787, \"end\": 2806, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2787, \"end\": 2806, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2787, \"end\": 2806}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2787, \"end\": 2806, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9908286333084106, \"start\": 305, \"end\": 324}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6607365012168884, \"start\": 305, \"end\": 324}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8746989369392395, \"start\": 305, \"end\": 324}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8746989369392395, \"start\": 305, \"end\": 324}}, {\"mention_name\": {\"text\": \"socioeconomic household surveys\", \"confidence\": 0.6240934133529663, \"start\": 2810, \"end\": 2841, \"clean_text\": \"socioeconomic household surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2810, \"end\": 2841, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2810, \"end\": 2841, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2810, \"end\": 2841}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2810, \"end\": 2841, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9908286333084106, \"start\": 328, \"end\": 359}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6240934133529663, \"start\": 328, \"end\": 359}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8746989369392395, \"start\": 328, \"end\": 359}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8746989369392395, \"start\": 328, \"end\": 359}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [4]}}, {\"page\": 5, \"chunk\": 5, \"input_text\": \"share of care work among rural women. On the other hand, we observe higher participation in leisure and self-care activities in the recall arm, including personal care, sleep and food consumption. Importantly, the smartphone arm also reflects greater participation among men in different areas of unpaid work — so that gender gaps in unpaid work in the smartphone sample, while remaining large, are narrowed when compared to the recall arm. Further, we show that the total reported time across most activities is, on average, higher in the recall arm, for both men and men and across urban and rural areas. The main driver of this finding is the set-up of the 24-hour time use diary to record activities in 15-minute intervals, while around a third of activities reported in the smartphone arm in fact last less than 15 minutes, with non-farm employment, unpaid domestic and care work, and self-care activities making up the greatest shares of these tasks. Given the common use of the 15-minute structure of time use diaries (UN Women, 2021; Seymour et al., 2020), our findings highlight how this structure can lead to “rounding” and overestimation of reported time.\\n\\nSecond, the smartphone data collection provides greater insights regarding the diversity and frequency of activities conducted throughout the day and a <mark>significantly greater incidence of activities (in particular nonleisure activities) conducted simultaneously with each other vis-à-vis the recall treatment arm.</mark> These nuances on time allocation can also help explain some of the drops in participation across activities that we observe in the recall arm. Recall, for example, leads to a substantial drop-off in reporting <mark>of active time in the early evening hours (after 6 pm), particularly among women. These hours account for a large share of daily time — about 26 percent of daily time for women and 29 percent of daily time for men in the smartphone arm. We discuss how underreporting of activities in the evening in the recall arm may be due to respondent fatigue - since each recall interview could last up to 30 to 40 minutes – as well as other interview-related factors. We also discuss how activity-linked pop-ups in the smartphone data collection to record the presence of other children, may highlight an important but understudied dimension of passive childcare (Folbre, 2005), over and above respondents’ own estimates of childcare.</mark> Third, we show that the stylized recall module with a 7-day reference period leads to even higher estimates of reported daily time compared to the 24-hour recall diary. Moving from real-time data collection to 24hour recall, and to 7-day recall therefore leads to increasing estimates of time allocation to both employment and unpaid domestic and care work. <mark>Recent studies on farm labor (Arthi et al., 2018; Gaddis et al., 2021) have also found that longer recall periods (e.g., over the last 30 days or the past season vis-àvis the last 7 days) overestimate labor time.</mark> Our findings are consistent over different reporting days, and treatment effects are robust to controlling for a range of individual, household, and community characteristics. We also discuss the successful experience with smartphone-based data collection, including the fact that the drop-offs among the selected participants were limited; that there was high, sustained participation in terms of logged activities and time; and that there was increasing proficiency in the use of the app throughout the study period. These findings can help accelerate the validation of scalable, improved approaches to individual-level data 4\", \"datasets\": [{\"mention_name\": {\"text\": \"24-hour time use diary\", \"confidence\": 0.5073368549346924, \"start\": 660, \"end\": 682, \"clean_text\": \"24-hour time use diary\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 660, \"end\": 682, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 660, \"end\": 682, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 660, \"end\": 682}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 660, \"end\": 682, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"report\", \"confidence\": 0.3025486469268799, \"start\": 660, \"end\": 682}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.5073368549346924, \"start\": 660, \"end\": 682}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.873602569103241, \"start\": 660, \"end\": 682}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.873602569103241, \"start\": 660, \"end\": 682}}, {\"mention_name\": {\"text\": \"smartphone data collection\", \"confidence\": 0.7750952839851379, \"start\": 1180, \"end\": 1206, \"clean_text\": \"smartphone data collection\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1180, \"end\": 1206, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1180, \"end\": 1206, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1180, \"end\": 1206}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1180, \"end\": 1206, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"microdata\", \"confidence\": 0.3135976195335388, \"start\": 373, \"end\": 399}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7750952839851379, \"start\": 373, \"end\": 399}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.46232491731643677, \"start\": 373, \"end\": 399}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.46232491731643677, \"start\": 373, \"end\": 399}}, {\"mention_name\": {\"text\": \"24-hour recall diary\", \"confidence\": 0.32579439878463745, \"start\": 2587, \"end\": 2607, \"clean_text\": \"24-hour recall diary\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2587, \"end\": 2607, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2587, \"end\": 2607, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2587, \"end\": 2607}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2587, \"end\": 2607, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9831584692001343, \"start\": 149, \"end\": 169}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.32579439878463745, \"start\": 149, \"end\": 169}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7118765115737915, \"start\": 149, \"end\": 169}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7118765115737915, \"start\": 149, \"end\": 169}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [5]}}, {\"page\": 6, \"chunk\": 6, \"input_text\": \"collection on time use, including efforts to scale up real-time data collection on time use through different approaches.<sup>5</sup> The paper is structured as follows. Section 2 discusses the study design. Section 3 provides an overview of the _Time Tracker_ and metrics on performance in the field. Section 4 compares time allocation across the smartphone and 24-hour recall diary arms, as well as a comparison with time reported in the 7-day recall module administered to both arms. Section 5 presents additional details on intraday variation in reporting and simultaneity of activities across treatment arms. Section 6 concludes with a discussion on future research needs and thoughts on scaling-up smartphone-based time use data collection in large-scale surveys in lower-income settings.\\n\\n# 2. Study design\\n\\nThe Malawi Methodological Experiment on Time Use Measurement was implemented from July 2022 to March 2023, across 72 enumeration areas (EAs) that were selected across the nine most populous districts in Southern and Central Malawi, with probability proportional to 2018 Population and Housing Census EAlevel household counts. The study period covers part of the dry season, which ends in October, and most of the rainy season, which goes from November to April, in order to capture seasonality in time use as much as possible.\\n\\nFurthermore, the EA sample was split evenly across Southern and Central Malawi, and by urban/rural status. In each EA, a full household listing was conducted to identify the households that had at least one adult man and one adult woman (working age 15-64). From this universe of households, five households were selected at random for each of the two treatments on time use data collection: (1) an Android-based smartphone diary with the pictorial _Time Tracker_ app, self-reported by respondents, and (2) a traditional time use diary with a 24-hour recall, administered by interviewers as part of a face-to-face interview. Within each household, following the administration of the household roster module of the household questionnaire, one adult man and one adult woman were randomly selected as the study participants.\\n\\nThis design yields a target sample of 720 households and 1,440 participants, split evenly by treatment arm and gender. However, due to app performance issues in the first two weeks of data collection (i.e., the first five EAs) that were sorted out swiftly but that led to a partial loss of smartphone data (i.e., for a subset of reported days for the respondents residing in these EAs), the number of sampled households for the smartphone arm was increased by 1 in approximately 25 percent of the remaining EAs to hedge against potential risk of data loss.<sup>6</sup> However, the feared data loss did not end up materializing, thanks to the\\n\\n> 5 This includes other approaches to high-frequency data collection; see McCullough et al. (2023) who implement a high-frequency phone survey on time use in Ethiopia.\\n\\n> 6 There was a total of 19 EAs in which six, as opposed to five, households were selected at random for the smartphone arm to hedge against the potential risk of data loss, following the experience with the first five EAs. Out of the 19 EAs, 11 EAs were visited immediately after the first five EAs, in August-September 2022. Based on the field experience and the incoming data from these 11 EAs in which there was ultimately no data loss, we reverted to sampling five households in each EA for the smartphone arm. Later, in January 2023, we had a single failure in data download from a smartphone on day 5, which we were able to resolve after\\n\\n5\", \"datasets\": [{\"mention_name\": {\"text\": \"2018 Population and Housing Census\", \"confidence\": 0.5732411742210388, \"start\": 1080, \"end\": 1114, \"clean_text\": \"2018 Population and Housing Census\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1080, \"end\": 1114, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1080, \"end\": 1114, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1080, \"end\": 1114}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1080, \"end\": 1114, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.9709566235542297, \"start\": 1080, \"end\": 1114}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5732411742210388, \"start\": 1080, \"end\": 1114}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5069783926010132, \"start\": 1080, \"end\": 1114}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5069783926010132, \"start\": 1080, \"end\": 1114}}, {\"mention_name\": {\"text\": \"time use diary\", \"confidence\": 0.3833526372909546, \"start\": 1864, \"end\": 1878, \"clean_text\": \"time use diary\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1864, \"end\": 1878, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1864, \"end\": 1878, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1864, \"end\": 1878}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1864, \"end\": 1878, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.8967908024787903, \"start\": 180, \"end\": 194}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3833526372909546, \"start\": 180, \"end\": 194}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8619642853736877, \"start\": 180, \"end\": 194}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8619642853736877, \"start\": 180, \"end\": 194}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [6]}}, {\"page\": 7, \"chunk\": 7, \"input_text\": \"updates to the app and the revision to the data retrieval protocol such that the data were successfully downloaded from and subsequently cleared from each smartphone at every check-in. Table 1 provides a breakdown of the final household sample by treatment arm and across the study districts.\\n\\nTable 1. Distribution of households across districts, by treatment arm\\n\\n||Smartphone<br>(%)|24-hour recall<br>(%)|\\n|---|---|---|\\n|Central Malawi|||\\n|Nkhotakota|0.03|0.03|\\n|Lilongwe|0.13|0.13|\\n|Dedza (urban)|0.07|0.07|\\n|Ntcheu|0.04|0.05|\\n|Lilongwe City (urban)<br>|0.23|0.23|\\n|Southern Malawi|||\\n|Machinga|0.06|0.06|\\n|Zomba|0.06|0.05|\\n|Blantyre|0.03|0.03|\\n|Thyolo (urban)|0.07|0.07|\\n|Mulanje|0.06|0.06|\\n|Zomba City (urban)|0.03|0.02|\\n|Blantyre City (urban)<br>|0.21|0.21|\\n|Number of<br>households|377|358|\\n\\nFigure 1 presents the structure of the study over the 11-day reporting period, including the visits made by interviewers to households in each treatment arm, and the administration of the 7-day recall module at the end of the survey period (day 11). As part of the study, and to inform a separate line of work in this program (Hocuk et al., in progress) respondents also wore a ActiGraph wGT3X-BT physical activity tracker over the reporting period and had height and weight measurements taken (informed by prior fieldwork in Malawi in 2017 - see Friedman et al., 2023), to inform ongoing research to develop machine learning models to reliably predict time use patterns from physical activity data and to more precisely identify physical activity demands of time use activities, across gender and different age groups.\\n\\nIn the smartphone treatment arm, an Android smartphone (Samsung Galaxy A12) was provided to each respondent for the duration of data collection. The smartphones were restricted to only allow respondents to use the _TimeTracker_ app on the device.<sup>7</sup> Sixty smartphones were reconditioned and redeployed throughout the fieldwork as survey teams moved from one enumeration area to another, and the teams were equipped with external batteries to recharge the respondents’ phones, if needed. As discussed in detail in Section 3, within the _TimeTracker_ app, respondents tapped on the images of activities (accompanied by captions in Chichewa) to begin recording time for those specific activities, and tapped the images again to stop recording time when they were done. On the day of deployment (Day 2), interviewers spent time guiding respondents on the use of the app. With the aid of paper illustrations, the interviewers provided an overview of all activities as depicted in _TimeTracker_ . Respondents were taught how to navigate\\n\\n> extensive troubleshooting. This incident did, however, stoke fears of data loss, as such we again increased the number of sampled households for the smartphone arm by one household in 8 EAs that were subsequently visited. Overall, 19 EAs were spread across eight out of 9 districts that were covered by the study, and among the 19 EAs, the rural/urban split was 12/7.\\n\\n> 7 Respondents also received a small pouch with a neck strap. A solar charger was also provided to each household in the smartphone treatment arm.\\n\\n6\", \"datasets\": [{\"mention_name\": {\"text\": \"physical activity data\", \"confidence\": 0.7899790406227112, \"start\": 1476, \"end\": 1498, \"clean_text\": \"physical activity data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1476, \"end\": 1498, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1476, \"end\": 1498, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1476, \"end\": 1498}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1476, \"end\": 1498, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.42155325412750244, \"start\": 787, \"end\": 809}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7899790406227112, \"start\": 787, \"end\": 809}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7524062991142273, \"start\": 787, \"end\": 809}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7524062991142273, \"start\": 787, \"end\": 809}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [7]}}, {\"page\": 8, \"chunk\": 8, \"input_text\": \"across the activities on the phone and how to press start and stop a given activity and were given time to practice. It was also emphasized that the recording needed to be continuous over the study period. After the initial orientation on day 2, the interviewer check-ins for the smartphone arm were conducted on days 3, 6, 9 and 11 to reinforce the basic tenets of the training provided on Day 2; download and review the time use data from respondents’ devices; identify and discuss potential anomalies in the downloaded data; and answer questions/concerns from the respondents.<sup>8</sup> Figure 1. Overview of field organization and visits to different treatment arms In the 24-hour recall diary arm, the interviewers visited the respondents three times during the study period (on days 6, 9 and 11 in each EA) to elicit their activities during the previous day, in between the time respondents woke up and when they went to sleep — during which the interviewer used the “time stamps” revealed by the respondent over the course of their conversation to code responses into the diary that was\\n\\n> 8 On the first check-in, the enumerator revisited in depth the entries between the time the enumerator left the household on day 2 and when they returned on day 3. This was done together with each respondent to make sure that he/she understood (a) the icons that are corresponding to each featured activity; (b) the idea of stopping each activity after starting it; and (c) the ability to record simultaneous activities. In this respect, the enumerator/supervisor was able to reinforce the training that was administered on day 2. On the subsequent check-in days, the enumerator and respondent also reviewed the entries to make sure that any anomalies (e.g., activities that were forgotten to be stopped) were addressed using the admin function on the app only accessible to the enumerator – see section 3.2 for a discussion. Any other queries that the respondents had were also addressed.\\n\\n7\", \"datasets\": [{\"mention_name\": {\"text\": \"time use data\", \"confidence\": 0.7054098844528198, \"start\": 422, \"end\": 435, \"clean_text\": \"time use data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 422, \"end\": 435, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 422, \"end\": 435, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 422, \"end\": 435}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 422, \"end\": 435, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.24687093496322632, \"start\": 422, \"end\": 435}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7054098844528198, \"start\": 422, \"end\": 435}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.4871467649936676, \"start\": 422, \"end\": 435}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.4871467649936676, \"start\": 422, \"end\": 435}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [8]}}, {\"page\": 9, \"chunk\": 9, \"input_text\": \"disaggregated into 15-minute intervals.<sup>9</sup> The recall diary (see Appendix Figure A1) is an adapted version of the time use module that had originally been developed as part of the IFPRI Women’s Empowerment in Agriculture Index. A similar module has also been introduced in several country contexts by the International Labor Organization (ILO) as an optional add-on module for labor force surveys (ILO, 2023). The same set of activity categories, including activity wording, were used across treatment arms to allow for one-to-one comparisons in the analysis — covering 56 different categories (see Appendix Table A1) that spanned agriculture, non-farm employment, unpaid domestic and care work, transport, leisure, schooling, personal care, sleeping and eating.\\n\\nFinally, the 7-day recall module that was administered to both treatment arms on day 11 (see Appendix Figure A2) spanned a subset of 20 key activities that were selected out of the aforementioned 56 activities (i.e., excluding activities pertaining to transport, personal care, eating/drinking, and sleep). The design of the module was informed by similar stylized questions that have been included in labor force surveys and socioeconomic household surveys. The respondents were first asked whether they conducted a given activity in the last 7 days. Then, regarding each activity that they were involved in, respondents were asked to report the number of hours they were engaged in that activity in the last 7 days.\\n\\nThe study also allows for the examination of how each survey approach affects reporting of simultaneity - in terms of participation, combinations of activities, and time spent conducting different activities together. The two treatment arms accounted for simultaneity of activities in different ways. The smartphone app, for example, allowed respondents to select multiple activities they were involved in at the same time, and click off each activity as it was finished. The 24-hour recall diary, on the other hand, first asked respondents what their main activity was for a given time interval, and then whether any secondary activity was conducted simultaneously with it — so that only two activities could be recorded simultaneously for a given 15-minute interval, and that the minimum time activities could be conducted together was 15 minutes. As seen in Appendix Figure A1, the 24-hour recall diary also included a preface to help encourage respondents to think of these secondary activities, including childcare.\\n\\nAs seen in Figure 1, additional survey modules were also implemented for both treatment arms on Day 1 (household and individual questionnaires eliciting a wide range of socioeconomic and demographic attributes) and on Day 9 (on respondents’ agency and preferences as well as perceived norms over asset ownership). There was gender matching of respondents and interviewers to help reduce potential unit and item non-response. Table 2 provides means of key individual and household attributes across treatment arms, across education, employment, and ownership of and rights to different assets such as land, financial accounts, and durables.<sup>10</sup> Balance tests across showed that the randomization worked successfully, except for few statistically significant, albeit small, differences across treatment arms.\\n\\n> 9 As a result, while the recall diary organizes the respondent’s reported time over the past day into 15-minute intervals, the interviewer does not ask the respondent to report on each 15-minute interval directly. Rather, the interviewer infers this information from the respondent’s narrative.\\n\\n> 10 The design of the modules on ownership and rights to land, financial accounts and durables follows recent international guidance on individual-disaggregated asset ownership and rights (United Nations, 2019).\\n\\n8\", \"datasets\": [{\"mention_name\": {\"text\": \"IFPRI Women’s Empowerment in Agriculture Index\", \"confidence\": 0.5076475143432617, \"start\": 189, \"end\": 235, \"clean_text\": \"IFPRI Women’s Empowerment in Agriculture Index\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.3324054181575775, \"start\": null, \"end\": null, \"clean_text\": \"\"}, \"producer\": {\"text\": \"International Labor Organization\", \"confidence\": 0.5782764554023743, \"start\": 314, \"end\": 346, \"clean_text\": \"International Labor Organization\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 189, \"end\": 235}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 189, \"end\": 235, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"indicator\", \"confidence\": 0.0, \"start\": 189, \"end\": 235}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5076475143432617, \"start\": 189, \"end\": 235}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.5897218585014343, \"start\": 189, \"end\": 235}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5897218585014343, \"start\": 189, \"end\": 235}}, {\"mention_name\": {\"text\": \"labor force surveys\", \"confidence\": 0.5134769082069397, \"start\": 386, \"end\": 405, \"clean_text\": \"labor force surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 386, \"end\": 405, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 386, \"end\": 405, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 386, \"end\": 405}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 386, \"end\": 405, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9768796563148499, \"start\": 386, \"end\": 405}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3540322184562683, \"start\": 386, \"end\": 405}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.9580660462379456, \"start\": 386, \"end\": 405}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9580660462379456, \"start\": 386, \"end\": 405}}, {\"mention_name\": {\"text\": \"socioeconomic household surveys\", \"confidence\": 0.49634572863578796, \"start\": 1199, \"end\": 1230, \"clean_text\": \"socioeconomic household surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1199, \"end\": 1230, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1199, \"end\": 1230, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1199, \"end\": 1230}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1199, \"end\": 1230, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9378021955490112, \"start\": 393, \"end\": 424}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.49634572863578796, \"start\": 393, \"end\": 424}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.44276177883148193, \"start\": 393, \"end\": 424}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.44276177883148193, \"start\": 393, \"end\": 424}}, {\"mention_name\": {\"text\": \"24-hour recall diary\", \"confidence\": 0.42329323291778564, \"start\": 2377, \"end\": 2397, \"clean_text\": \"24-hour recall diary\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2377, \"end\": 2397, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2377, \"end\": 2397, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2377, \"end\": 2397}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2377, \"end\": 2397, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.7261225581169128, \"start\": 752, \"end\": 772}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.42329323291778564, \"start\": 752, \"end\": 772}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.7400967478752136, \"start\": 752, \"end\": 772}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7400967478752136, \"start\": 752, \"end\": 772}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [9]}}, {\"page\": 10, \"chunk\": 10, \"input_text\": \"Table 2. Descriptive statistics, by treatment arm\\n\\n||24-hour<br>recall<br>S|martphone<br>diary||24-hour<br>recall|Smartphone<br>diary|\\n|---|---|---|---|---|---|\\n|Demographics|||<sup>Employment</sup>|||\\n|Household roster: HH head|0.41|0.41|Not employed|0.19*|0.16*|\\n|Household roster: spouse of HH head|0.38|0.37|Employed in agriculture|0.33|0.36|\\n|Age: 18-24|0.25|0.24|Employed in non-agriculture|0.48|0.49|\\n|Age: 25-34|0.26*|0.32*|Other employment characteristics:|||\\n|Age: 45-54*|0.12|0.11|Time (mins) to work|32.0**|27.4**|\\n|Age: 55+|0.06**|0.04**|Average monthly take-home profits in kwacha<br>(self-employed)|32,799.8|36,813.1|\\n|Married (monogamous), patrilineal|0.29|0.33|Average daily earnings in kwacha (wage)|2,350.8*|2,950.5*|\\n|Married (monogamous), matrilineal|0.41|0.37|Average daily earnings in kwacha (_ganyu_)|4,421.6|5,095.3|\\n|Separated/divorced|0.04|0.04||||\\n|Single|0.19|0.19|Ownership of assets|||\\n|Months resp. is away from HH in last year<br>|0.50<br>|0.49|Owns land, exclusively|0.25|0.23|\\n||||<sup>Owns land, jointly</sup>|0.80|0.78|\\n|Education|||<sup>Has an account: mobile money</sup>|0.55|0.58|\\n|Highest level of education: primary|0.61|0.59|Has an account: informal savings|0.13|0.15|\\n|Highest level of education: secondary or<br>above|0.32|0.34|Has an account: formal fin. institution|0.10|0.09|\\n||||<sup>Has a mobile phone</sup>|0.67|0.68|\\n|Household characteristics|||<sup>Owns a bicycle</sup>|0.28|0.28|\\n|Number of children <12 years|1.71|1.67|Owns a motorized vehicle|0.05|0.08|\\n|Household size|5.04|4.94||||\\n|HH has piped water|0.18|0.15|Number of respondents<br>|710|753|\\n|HH has electricity<br>Notes:|0.25*|0.29*||||\\n\\n<u>Notes:</u> (1) T-tests of equality of means conducted across treatment groups; ***p<0.01, **p<0.05, *p<0.10.\\n\\n# 3. Time Tracker Android smartphone app\\n\\n## 3.1 Overview\\n\\nThe _Time Tracker_ app allows for a continuous, real-time recording of time use that someone carries out during the day. The main screen (Figure 2a) shows respondents easy-to-understand illustrations of typical daily activities, that as mentioned earlier, also correspond one-to-one with activities in the 24-hour recall instrument.<sup>11</sup> Separate versions of the app with images of men and women conducting activities were provided to men and women respondents, respectively, to ensure that certain activities often linked with gender norms (such as household work, driving, and certain forms of employment) were not necessarily under-counted by either women or men. Scrolling activity description in Chichewa were also included at the top of each image to aid in the correct identification of each activity.<sup>12</sup>\\n\\n> 11 The illustrations were customized to different activities, and went through a process of editing and piloting before integrating in the _Time Tracker_ app. The app is structured in a way that the illustrations can be replaced easily in order to adapt to different country contexts.\\n\\n> 12 Appendix Table A1 provides the activity list for the smartphone app, along with the pop-ups — as well as the Chichewa captions that accompanied the images.\\n\\n9\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [10]}}, {\"page\": 11, \"chunk\": 11, \"input_text\": \"<!-- Start of picture text --> 72 One > CRE<br>29.03.2023 fad<br>o<br>4<br>\\\\<br>% Kuthilira ndi vava |): kuthilira ndi pompi \\\"<br>|<br>|<br>| @ |<br>———<br>Kupalira ndi manja |} ithira fetereza/man |<br>Y<br>&<br>} Kukolora ndima<br> KukoloraBS,SX ndi manja | yanika ndi kusunge<br>MW [e) <<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> 0s aaa 6% 4678<br>19.06.2022 ”<br>b it<br>fa<br>00:00:17 } 00:00:05<br>Ku sukkula Kuwelenga<br>=<br>si yakaneme, 4 kum Kupumula<br>— [J]<br>=]<br>Kuchen pakhomo 0, kukaonera zam<br>%<br>os s<br>Zamapemphero nupi Masew<br><!-- End of picture text --> <!-- Start of picture text --> 20:06 Gs Sal 67%8<br>Food (multiple selection)<br>Gulu fa tirigu Mitsitsi<br>SO)a ie ee)<br>Faerie went<br>ove 5<br>Mazira Nsomba<br>Pe<br>Fo<br>Mkaka ‘Shuga/Uchi<br>&. °<br>= 5<br><!-- End of picture text --> <!-- Start of picture text --> 1rce- aX ‘<br>gapo) Amachita ndir<br>i<br>saveat iMivene weingforio<br>fe<br>hih 3<br>Pandekha<br>ft<br>| cme |<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> sare + 8X1 26%8<br>Mukuganiza bwanji?<br>Okondwa Pakatikati<br>Osakondwa<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [11]}}, {\"page\": 12, \"chunk\": 12, \"input_text\": \"# 3.2 Field experience with smartphone-based data collection\\n\\nThe smartphone sample includes a total of 160,880 recorded activities across the respondents, over 9 full days of data collection between deployment (day 2) and pick up (day 11). In what follows, we present specific metrics related to the field experience with smartphone-based data collection. One concern was that respondents might leave activities on too long after pressing them. Ultimately, only 2 percent or fewer entries within an activity category in each day had reported time that was three standard deviations above the daily mean, with no variation over the reporting period. Of all recorded activities, 6.9 percent (or 11,053) were “administrative”, i.e., needed adjustment by the interviewers at the time of the check-ins. The comparison of reported time estimates as a result of including versus excluding these administrative entries showed that the administrative adjustments were primarily for recording sleep time.<sup>13</sup> Figure 3 presents additional metrics on smartphone performance over the reporting period. We exclude the ends of the reporting period (days 2 and 11), since day 2 was when phones were dropped off and considered more of a “training” period on using the smartphone app, and day 11 was when smartphones were picked up. As a result, Figures 3a-3f are informed by a total of 138,609 activities across days 3-10. Some decline is observed in the number of participating respondents over the study period, due mainly to respondent travels out of EAs and their phones being collected early as a result (Figure 3a). However, among reporting respondents, the number of diary entries per day (Figure 3b), as well as average total daily reported time (Figure 3c), were stable over the period (between days 4 and 9 in particular). Notably, the share of activities lasting less than 15 minutes in length was quite substantial for both men and women - ranging around 30 percent for men and 26 percent for women between days 4 and 9, which we will revisit in Section 4. Moreover, over the study period, the respondents improved the accuracy with which they recorded their time in the app - with fewer discontinuities (Figure 3e) as well as entries that required editing by the interviewers (i.e., administrative or deleted entries) (Figure 3f). While the time logged and number of recorded activities were not significantly different across men and women, we observe significant gender differences in Figures 3e and 3f, which can be an important consideration for future surveys and training of different groups of respondents on the app.\\n\\nAppendix Table A2 also presents regressions on additional correlates of outcomes in Figures 3e and 3f at the individual-day level, controlling for enumeration area and day of the week fixed effects. The findings reinforce that women were significantly more likely to have entries with time jumps or that required editing, and older participants (55+) as well as participants in non-agricultural employment were also more likely to have administrative or deleted entries. The respondents with secondary or higher education were less likely to have time jumps (as measured in Figure 3e) or needing editing in their smartphone logs. Having a bicycle and electricity in the household were less likely to have entries that needed editing later on.<sup>14</sup>\\n\\n> 13 The results are available upon request. The qualitative evidence from the field was that there were respondents who were disturbed by the phone light or were uncomfortable with the idea of leaving the phone on overnight. As such, they did not log sleep time and left the phone off while they were sleeping.\\n\\n> 14 Interestingly, mobile phone ownership did not have a significant effect, but this could also be because many phones owned in the sample may not have been smartphones.\\n\\n11\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [12]}}, {\"page\": 13, \"chunk\": 13, \"input_text\": \"<!-- Start of picture text --> 400<br>375<br>350 “Say ==<br>325 ><br>‘e<br>300<br>275<br>—e— Women --=-Men<br>250<br>3 4 5 6 7 8 9 10<br><!-- End of picture text --> <!-- Start of picture text --> 04<br>0.38 t<br>03 NS<br>0.34 \\\\<br>0.32 a a<br>03 t ws<br>0.28 We -----o-----8,<br>0.26 my<br>0.24<br>0.22 |—e—Women ----Men<br>0.2<br>3 4 5 6 7 8 9 10<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> 13<br>12<br>11 @&<br>10 SS<br>9 an<br>8 ne<br>7 wee *<br>6 a ae —_—'<br>5<br>4 —e— Women --=-Men<br>3<br>3 4 5 6 7 8 9 10<br><!-- End of picture text --> <!-- Start of picture text --> 35<br>30<br>25 utah Soe<br>20 pra<br>15<br>10<br>—e— Women --*- Men<br>5<br>3 4 5 6 7 8 9 10<br><!-- End of picture text --> <!-- Start of picture text --> 1700<br>1650<br>1600<br>1550 $7<br>1500 9°” De perry<br>om “ .<br>1450 Sp aN Leo<br>1400<br>1350 —s#— Women =-*=-Men<br>1300<br>3 4 5 6 7 8 9 10<br><!-- End of picture text --> <!-- Start of picture text --> 0.25<br>0.23<br>0.21<br>0.19<br>0.17<br>0.15 a eS<br>0.13 2 wma<br>0.11 ing cncnage en”<br>0.09<br>0.07 | —e—Women --«=Men<br>0.05<br>3 4 5 6 7 8 9 10<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [13]}}, {\"page\": 14, \"chunk\": 14, \"input_text\": \"# 4. Daily time allocation across the smartphone, 24-hour recall and 7-day recall approaches\\n\\n## 4.1 Time allocation and treatment effects across smartphone and 24-hour recall arms\\n\\nAs shown in Figure 1, the recall treatment was administered at three times during the weeklong survey period for each respondent — day 6 (with the recall diary corresponding to the prior 24 hours, i.e., day 5); day 9 (corresponding to time spent during day 8); and day 11 (corresponding to time spent during day 10). As the smartphone data were collected continually, the analysis in this section presents comparisons of daily time allocation in the smartphone arm matched by day with the recall arm, to account for any potential factors affecting day-to-day changes in time allocation. The results for the aggregated activity categories are presented across (1) agriculture; (2) nonfarm employment; (3) unpaid domestic and care work (obtaining services/resources for the household; cooking and cleaning; and care of children or elderly); (4) transport; (5) schooling/studying; (6) leisure; (7) community activities; (8) personal care (bathing, dressing, care for one’s own health issues); (9) sleep; and (10) eating/drinking. The descriptive statistics on the disaggregated activities underlying these categories are presented in the Appendix Tables A3 and A4.\\n\\nTaking the day 5 comparison as an example,<sup>15</sup> Table 3 shows that the smartphone arm reports significantly higher participation across several activity categories in agriculture, non-agriculture (women), unpaid domestic and care work for the household, schooling/studying, and community activities. This included higher unpaid care work among women in the smartphone arm, but also greater participation in different areas of unpaid work among men. The recall arm, on the other hand, yielded significantly higher participation rates in self-care, such as leisure, personal care, sleeping and eating. We turn to potential reasons for these differences in Section 5, including greater reporting across a range of activities conducted throughout the day in the smartphone arm, as well as <mark>multitasked activities. T</mark> able 4 also shows that greater participation in employed and unpaid work among men and women in the smartphone arm leads to narrower, albeit still substantial, gender gaps for participation in most of these activities, with a slightly greater gap for care work — 41 percent, compared to 37 percent in the recall arm.\\n\\nOn unconditional time, however, owing likely to the structure of the 24-hour recall module where time devoted to a given activity was recorded for a minimum of 15 minutes, and the fact that smartphone arm respondents recorded large shares of activities for less than 15 minutes (Table 5), the reported daily time in minutes is higher across activity categories in the recall arm, except for transportation. Along with lower participation among men in employed and unpaid work in the recall arm, this leads to higher time-related gender gaps for these activities (Table 4). Kernel density estimates in Figure 4, based on the pooled matched sample of days 5, 8, and 10, also show much greater increases in unconditional time in the recall sample for areas such as leisure, and more so for men. Only unconditional time in transport and some areas of men’s unpaid work remain higher in the smartphone arm, indicating that these activities are even further underestimated in the recall arm, given the minimum 15-minute floor for activities in the 24-hour diary.\\n\\n> 15 Appendix Figure A3 shows that there is very little variation when averaging time in activity categories across days 5, 8, and 10, for both men and women.\\n\\n13\", \"datasets\": [{\"mention_name\": {\"text\": \"smartphone data\", \"confidence\": 0.9339187741279602, \"start\": 506, \"end\": 521, \"clean_text\": \"smartphone data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 506, \"end\": 521, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 506, \"end\": 521, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 506, \"end\": 521}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 506, \"end\": 521, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.265309602022171, \"start\": 506, \"end\": 521}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.9339187741279602, \"start\": 506, \"end\": 521}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6868488192558289, \"start\": 506, \"end\": 521}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6868488192558289, \"start\": 506, \"end\": 521}}, {\"mention_name\": {\"text\": \"24-hour diary\", \"confidence\": 0.31759750843048096, \"start\": 3537, \"end\": 3550, \"clean_text\": \"24-hour diary\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3537, \"end\": 3550, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3537, \"end\": 3550, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3537, \"end\": 3550}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3537, \"end\": 3550, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.5190667510032654, \"start\": 422, \"end\": 435}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.31759750843048096, \"start\": 422, \"end\": 435}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7983964681625366, \"start\": 422, \"end\": 435}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7983964681625366, \"start\": 422, \"end\": 435}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [14]}}, {\"page\": 15, \"chunk\": 15, \"input_text\": \"The gender gaps in time allocation remain quite substantial in the smartphone arm for nonfarm employment, unpaid domestic and care work, transport, and leisure, but are magnified in the recall arm due in large part to the 24-hour diary reporting structure.\\n\\nFurthermore, Table 5 shows that that the 15-minute interval setup in the 24-hour recall diary is at odds with real-time recorded activity duration across several key areas. For instance, 14-17 percent of smartphone diary entries in agriculture and non-farm employment were recorded for less than 15 minutes. The comparable estimates were 26-29 percent for care work, 39-44 percent for obtaining services and resources for the household, and around 40 percent for eating and drinking. Men were also significantly more likely than women in the smartphone arm to log activities for less than 15 minutes, particularly for non-farm employment, unpaid domestic work, and self-care activities.\\n\\nTable 3. Participation and time (minutes) in activities, by treatment arm (Day 5 match)\\n\\n|||||Unpaid|work:|||||||||\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n||Agri-<br>culture|Nonfarm<br>employ-<br>ment|Obtaining<br>services/<br>resources<br>for HH|Cooking,<br>cleaning<br>in house|Care of<br>children,<br>elderly|TOTAL<br>UNPAID<br>WORK|Transport|School/<br>studying|Leisure|Comm.<br>Activities|Personal<br>care|Sleep|<br>Eating/<br>drinking|\\n||(1)|(2)|(3a)|(3b)|(3c)|(3d)|(4)|(5)|(6)|(7)|(8)|(9)|(10)|\\n|SHARE PARTICIPATING||||||||||||||\\n|Total||||||||||||||\\n|Recall (obs=696)|0.33***|0.33*|0.49***|0.65|0.27***|0.76***|0.94|0.04***|0.99**|0.04***|0.94***|1.00***|<br>0.99**|\\n|Smartphone (obs=734)|0.41***|0.37*|0.63***|0.67|0.34***|0.84***|0.94|0.09***|0.98**|0.14***|0.89***|0.98***|<br>0.96**|\\n|Women||||||||||||||\\n|Recall (obs=352)|0.35*|0.23*|0.68|0.98*|0.45***|0.98|0.91|0.03**|0.99|0.05***|0.96***|1.00**|<br>0.99**|\\n|Smartphone (obs=365)|0.42*|0.29*|0.72|0.95*|0.55***|0.97|0.91|0.06**|0.97|0.14***|0.90***|0.99**|<br>0.96**|\\n|Men||||||||||||||\\n|Recall (obs=344)|0.31***|0.42|0.30***|0.32**|0.08**|0.53***|0.97|0.05***|0.99|0.03***|0.92|1.00***|0.99***|\\n|Smartphone (obs=369)|0.41***|0.45|0.53***|0.40**|0.14**|0.71***|0.96|0.11***|0.98|0.14***|0.88|0.97***|0.96***|\\n|DAILY TIME, UNCONDITION|AL (MINS)|||||||||||||\\n|Total||||||||||||||\\n|Recall (obs=696)|80.2***|120.2***|39.5|129.7|57.3***|224.3|128.1***|13.9|507.9***|5.3|33.6|536.8|<br>54.8**|\\n|Smartphone (obs=734)|53.0***|84.7***|45.9|138.2|38.7***|222.8|162.0***|11.7|374.4***|7.8|34.9|529.3|44.3**|\\n|Women||||||||||||||\\n|Recall (obs=352)|74.6***|52.5|50.5|231.5|100.8**|381.6|104.9**|13.0|426.9***|5.5|38.1|549.3|<br>51.6|\\n|Smartphone (obs=365)|48.1***|47.8|46.2|240.5|69.6**|356.3|118.7**|6.8|320.1***|6.4|38.3|546.2|<br>48.2|\\n|Men||||||||||||||\\n|Recall (obs=344)|85.9***|189.3***|28.4**|25.6*|12.9|65.7**|151.9***|14.9|590.8***|5.0*|28.9|524.0|<br>58.1**|\\n|Smartphone (obs=369)|57.8***|121.3***|45.6**|36.6*|8.1|90.4**|204.9***|16.6|428.3***|9.2*|31.6|512.5|<br>47.8**|\\n\\n# <u>Notes:</u>\\n\\n(1) Within the total, women’s and men’s samples, T-tests of equality of means conducted across treatment groups; ***p<0.01, **p<0.05, *p<0.10. Grey shading = recall arm yields significantly lower estimates. (2) Average total time for women and men was 1,700 minutes in the recall sample (including main and secondary activities). In the smartphone sample, average total daily time was 1,521 minutes for women, and 1,490 minutes for men.\\n\\n14\", \"datasets\": [{\"mention_name\": {\"text\": \"24-hour recall diary\", \"confidence\": 0.3418814241886139, \"start\": 331, \"end\": 351, \"clean_text\": \"24-hour recall diary\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 331, \"end\": 351, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 331, \"end\": 351, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"24-hour\", \"confidence\": 0.4308662712574005, \"start\": 222, \"end\": 229}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 331, \"end\": 351, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.6002652049064636, \"start\": 331, \"end\": 351}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.3418814241886139, \"start\": 331, \"end\": 351}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.497372567653656, \"start\": 331, \"end\": 351}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.497372567653656, \"start\": 331, \"end\": 351}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [15]}}, {\"page\": 16, \"chunk\": 16, \"input_text\": \"Table 4. Gender gaps (women – men) in activities, by treatment arm (Day 5 match)\\n\\n|||||Unpaid|work:|||||||||\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n||Agri-<br>culture|Nonfarm<br>employ-<br>ment|Obtaining<br>services/<br>resources<br>for HH|Cooking,<br>cleaning<br>in house|Care of<br>children,<br>elderly|TOTAL<br>UNPAID<br>WORK|Transport|School/<br>studying|Leisure|Com-<br>munity<br>Activities|Personal<br>care|Sleep|<br>Eating/<br>drinking|\\n||(1)|(2)|(3a)|(3b)|(3c)|(3d)|(4)|(5)|(6)|(7)|(8)|(9)|(10)|\\n|SHARE PARTICIPATING||||||||||||||\\n|Gender gap (women-men)||||||||||||||\\n|Recall|0.04|-0.19|0.38|0.66|0.37|0.45|-0.06|-0.02|0|0.02|0.04|0|<br>0|\\n|Smartphone|0.01|-0.16|0.19|0.55|0.41|0.26|-0.05|-0.05|-0.01|0|0.02|0.02|<br>0|\\n|DAILY TIME, UNCONDITIONAL|(MINS)|||||||||||||\\n|Gender gap (women-men)||||||||||||||\\n|Recall|-11.3|-136.8|22.1|205.9|87.9|315.9|-47.0|-1.9|-163.9|0.5|9.2|25.3|<br>-6.5|\\n|Smartphone|-9.7|-73.5|0.6|203.9|61.5|265.9|-86.2|-9.8|-108.2|-2.8|6.7|33.7|0.4|\\n\\n<u>Notes:</u> (1) The gender gap takes the difference in means for women and men in each treatment arm.\\n\\nFigure 4. Kernel density estimates for daily minutes in key activity categories, by treatment arm and gender <!-- Start of picture text --> Minutes in leisure Minutes in unpaid work (cooking/cleaning,<br>care work, obtaining resources)<br>Transport Employment (agr/non-agr.)<br><!-- End of picture text --> <u>Notes:</u>\\n\\n- (1) For all graphs: p-values from Kolmogorov-Smirnov test of distributional differences (across treatment arms): 0.000\\n\\n(2) Based on the pooled sample of days 5, 8, and 10.\\n\\n15\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [16]}}, {\"page\": 17, \"chunk\": 17, \"input_text\": \"Table 5. Shares of smartphone diary entries lasting for less than 15 minutes in each aggregate activity category, by gender\\n\\n||Women|Men|\\n|---|---|---|\\n|Agriculture|0.17|0.16|\\n|Nonfarm employment|0.16*|0.14*|\\n|Unpaid work:|||\\n|Obtaining services/resources for HH|0.39***|0.44***|\\n|Cooking, cleaning in house|0.12***|0.20***|\\n|Care of children, elderly|0.26|0.29|\\n|Transport|0.41|0.42|\\n|School/studying|0.11|0.11|\\n|Leisure|0.24***|0.26***|\\n|Community activities|0.15|0.15|\\n|Personal care|0.30***|0.40***|\\n|Eating/drinking|0.38***|0.42***|\\n\\n<u>Notes:</u> (1) Based on 138,609 activities recorded across Days 3-10 of the survey period (data from deployment on Day 2 were dropped).\\n\\n- (2) T-tests of equality of means conducted across men and women. ***p<0.01, **p<0.05, *p<0.10. Sleep was excluded from the table.\\n\\nTable 6 presents individual-level, ordinary least squares (OLS) regressions of treatment effects of the recall arm on daily time allocation (unconditional minutes),<sup>16</sup> for the pooled matched sample across days 5, 8, and 10: i β β i 1 2 i i (1) ii ii ii ii yy = αα+ β ββ+ γγXX + δδ AA + δδ DD + εε In Equation (1), i represents average daily time in each of the activity categories in Table 3, for individual _i_ in household _j_ and enumeration area _k_ ; _Recall_ is the dichotomous variable that is equal to 1 if the individual ii yy was part of the recall treatment; _X_ is a vector of individual- and household-level characteristics, as included in Appendix Table A2; _A_ and _D_ represents the fixed effects for enumeration areas and days of reporting, respectively; and is the error term. The regressions were run separately for the total, men, and women samples, along with an additional disaggregation by urban/rural, in view of geographic differences in εε employment and transport.\\n\\nThe results in Table 6 reveal similar patterns observed earlier; participation is lower in the recall sample for economic activities in agriculture, unpaid work (especially care work), studying, community activities – but higher in the recall arm for leisure, personal care, sleep, and eating/drinking, as well as some unpaid domestic activities for women, focused mainly on cooking and cleaning. The reported time was lower in the recall arm concerning (a) agricultural activities of both men and women in urban areas and (b) nonfarm employment of urban women and rural men. The negative recall treatment effects on participation in unpaid domestic and care work stemmed mainly from rural areas, with some significant negative effects for men in urban areas on care work and obtaining services/resources for the household. Despite this variation in participation, however, treatment effects on time spent in different activities are generally higher in the recall arm, except for time in transport and a subset of unpaid domestic activities for men\\n\\n> 16 Descriptive statistics on conditional minutes are in Appendix Table A5.\\n\\n16\", \"datasets\": [{\"mention_name\": {\"text\": \"smartphone diary entries\", \"confidence\": 0.42207464575767517, \"start\": 19, \"end\": 43, \"clean_text\": \"smartphone diary entries\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 19, \"end\": 43, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 19, \"end\": 43, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 19, \"end\": 43}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 19, \"end\": 43, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"microdata\", \"confidence\": 0.39097392559051514, \"start\": 19, \"end\": 43}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.42207464575767517, \"start\": 19, \"end\": 43}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.8283804059028625, \"start\": 19, \"end\": 43}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.8283804059028625, \"start\": 19, \"end\": 43}}, {\"mention_name\": {\"text\": \"Descriptive statistics on conditional minutes\", \"confidence\": 0.7121233344078064, \"start\": 2873, \"end\": 2918, \"clean_text\": \"Descriptive statistics on conditional minutes\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2873, \"end\": 2918, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2873, \"end\": 2918, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2873, \"end\": 2918}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2873, \"end\": 2918, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.5818593502044678, \"start\": 495, \"end\": 540}, \"specificity_tag\": {\"text\": \"descriptive\", \"confidence\": 0.7121233344078064, \"start\": 495, \"end\": 540}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.36975371837615967, \"start\": 495, \"end\": 540}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.36975371837615967, \"start\": 495, \"end\": 540}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [17]}}, {\"page\": 18, \"chunk\": 18, \"input_text\": \"(particularly for cooking and cleaning in rural areas). One takeaway is that where there is interest in areas around employment, unpaid work, transport and other non-leisure activities, the recall arm provides greater estimates of reported time, but among a smaller, less representative subset of participating individuals.\\n\\nTable 6. Treatment effects of 24-hour recall diary on participation and daily time allocation\\n\\n|||||Unpaid|work:|||||||||\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n||Agri-<br>culture|Nonfarm<br>employ-<br>ment|Obtaining<br>services/<br>resources<br>for HH|Cooking,<br>cleaning<br>in house|Care of<br>children,<br>elderly|TOTAL<br>UNPAID<br>WORK|Transport|School/<br>studying|Leisure|Com-<br>munity<br>Activities|Personal<br>care|Sleep|<br>Eating/<br>drinking|\\n||(1)|(2)|(3a)|(3b)|(3c)|(3d)|(4)|(5)|(6)|(7)|(8)|(9)|(10)|\\n|PARTICIPATION (Y=1 N=0)||||||||||||||\\n|All areas||||||||||||||\\n|Total (obs=4,108)|-0.06***|-0.01|-0.08***|-0.03*|-0.04**|-0.02|0.01|-0.03***|0.02***|-0.07***|0.05***|0.04***|0.04***|\\n|Women (obs=2,049)|-0.06***|-0.03|0.00|0.03***|-0.04*|0.06***|0.01|-0.03***|0.02***|-0.05***|0.06***|0.02***|0.03***|\\n|Men (obs=2,059)|-0.06***|0.00|-0.16***|-0.09***|-0.04**|-0.11***|0.01|-0.02|0.02***|-0.09***|0.05***|0.06***|0.05***|\\n|Rural||||||||||||||\\n|Total (obs=2,081)|-0.04|-0.03|-0.13***|-0.08***|-0.07***|-0.06***|0.00|-0.03**|0.01*|-0.09***|0.05***|0.04***|0.04***|\\n|Women (obs=1,042)|-0.04|-0.00|-0.06*|0.03*|-0.09**|0.05***|0.01|-0.04***|0.01|-0.06***|0.08***|0.03***|0.03***|\\n|Men (obs=1,039)|-0.03|-0.06**|-0.21***|-0.19***|-0.04**|-0.17***|-0.01|-0.02|0.01|-0.11***|0.03|0.05***|0.04***|\\n|Urban||||||||||||||\\n|Total (obs=2,027)|-0.08***|-0.00|-0.03|0.04*|-0.02|0.01|0.01|-0.03*|0.03***|-0.05***|0.06***|0.04***|0.05***|\\n|Women (obs=1,007)|-0.07***|-0.07**|0.06*|0.03**|-0.01|0.07***|0.01|-0.03|0.03***|-0.04**|0.04**|0.02***|0.04***|\\n|Men (obs=1,020)|-0.10***|0.05|-0.12***|0.01|-0.04*|-0.06*|0.01|-0.02|0.03**|-0.06***|0.07***|0.06***|0.06***|\\n|DAILY TIME, UNCONDITIONA|L (MINS)|||||||||||||\\n|All areas||||||||||||||\\n|Total (obs=4,108)|26.2***|28.8***|0.1|-5.3|18.4***|23.8***|-30.8***|6.1**|141.2***|-2.3*|-2.6|11.3**|<br>6.6***|\\n|Women (obs=2,049)|18.7***|8.8|6.3**|-0.1|32.1***|53.5***|-11.4**|3.8|108.2***|-0.3|-3.0|10.8*|<br>3.4*|\\n|Men (obs=2,059)|34.8***|49.4***|-5.6|-13.1***|4.5*|-9.6|-50.9***|8.9*|170.1***|-3.9*|-2.7|14.0*|10.1***|\\n|Rural||||||||||||||\\n|Total (obs=2,081)|55.5***|14.4*|1.3|-8.7|23.9***|29.5***|-34.5***|3.1|106.6***|-2.7|3.0|19.7***|11.0***|\\n|Women (obs=1,042)|38.7***|4.7|1.9|5.0|43.3***|66.3***|-10.8|-1.2|73.3***|-2.7|4.5|13.5|<br>9.6***|\\n|Men (obs=1,039)|76.7***|22.9*|2.4|-17.5***|6.7*|-2.0|-59.1***|6.5|130.4***|-2.2|2.7|28.1***|12.1***|\\n|Urban||||||||||||||\\n|Total (obs=2,027)|-0.7|39.7***|-4.1|-2.1|8.7|11.2|-26.9***|10.1*|179.0***|-1.4|-8.0*|2.6|<br>2.4|\\n|Women (obs=1,007)|0.3|8.7|9.8**|-7.8|15.1|29.9|-11.3*|8.2|151.4***|2.2|-9.2|9.7|<br>-2.8|\\n|Men (obs=1,020)|-3.4|75.4***|-16.7*|-7.7|1.8|-18.5*|-41.7***|11.3|207.5***|-4.9|-6.5**|-1.7|<br>6.7*|\\n\\n<u>Notes:</u> (1) OLS regressions controlling for EA and day of the week Fes; the standard errors are clustered at the household-level. ***p<0.01, **p<0.05, *p<0.10. Grey shading = recall arm yields significantly lower estimates.\\n\\n(2) Based on a pooled individual-level sample for days 5, 8 and 10 (each individual appeared three times in the sample).\\n\\n(3) Descriptive statistics on conditional minutes are in Appendix Table A5.\\n\\n# 4.2 Comparisons with time use data based on stylized questions with a 7-day recall\\n\\nIn this section, we compare average daily minutes based on stylized questions on time use with a 7-day recall (obtained by dividing the total reported weekly hours by 7) with the average daily minutes under the 24-hour recall and smartphone treatments (computed across days 5, 8 and 10). The comparisons are provided for the activity categories that were common across the three survey approaches. Table 7 shows that the reliance on stylized time use questions with a 7-day recall leads to higher average daily time across 17\", \"datasets\": [{\"mention_name\": {\"text\": \"time use data\", \"confidence\": 0.6378820538520813, \"start\": 3502, \"end\": 3515, \"clean_text\": \"time use data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3502, \"end\": 3515, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3502, \"end\": 3515, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3502, \"end\": 3515}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 3502, \"end\": 3515, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.40942278504371643, \"start\": 511, \"end\": 524}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.6378820538520813, \"start\": 511, \"end\": 524}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5966583490371704, \"start\": 511, \"end\": 524}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5966583490371704, \"start\": 511, \"end\": 524}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [18]}}, {\"page\": 19, \"chunk\": 19, \"input_text\": \"= @ + BRecall + 7X = @+ BSmartphon\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"non_english\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [19]}}, {\"page\": 20, \"chunk\": 20, \"input_text\": \"time is computed across all 9 days of reporting.<sup>17</sup> Table 8 reports the 7-day recall treatment effects.<sup>18</sup> The results, in conjunction with the 24-hour recall and smartphone comparison in Section 4.1, suggest that at higher aggregations of recall time – moving from the last 24 hours to 7 days – the estimated time for key activity areas around employment and unpaid work go up. This is extremely important for understanding individuals’ productivity, for example, as well as opportunity costs for re-allocating their time across other areas. The substantial underestimation of leisure time in the 7-day recall module vis-à-vis the smartphone and 24-hour recall approaches, also indicates that the 7-day recall module may be insufficient for understanding the true distribution of daily time across activities.\\n\\nTable 8. Overall treatment effects of smartphone and 24-hour recall approaches compared to 7-day recall\\n\\n||Agri-<br>culture<br>(1)|E<br>Nonfarm<br>employ-<br>ment<br>(2)|ffect of smartphone<br>Unpaid work:<br>obtaining services/<br>resources for HH<br>(3)|and recall arms o<br>Unpaid work:<br>cooking<br>and cleaning in<br>house<br>(4)|n average dail<br>Unpaid<br>work: Care<br>of children,<br>elderly<br>(5)|y minutes spen<br>Community<br>activities<br>(6)|t in:<br>School/<br>studying<br>(7)|Leisure<br>(8)|\\n|---|---|---|---|---|---|---|---|---|\\n|Total|||||||||\\n|Smartphone arm|-16.1***|-81.9***|-8.2**|-26.5***|-49.5***|1.6|-3.6***|255.6***|\\n|24-hour recall arm|6.1|-49.1***|-3.4|-13.8***|-30.1***|-0.1|-2.0|404.4***|\\n|Women|||||||||\\n|Smartphone arm|-7.8*|-55.9***|-33.9***|-42.0***|-70.1***|0.8|-2.6*|196.6***|\\n|24-hour recall arm|11.8**|-40.8***|-19.4***|-16.4**|-32.7***|-0.8|-1.3|331.5***|\\n|Men|||||||||\\n|Smartphone arm|-24.5***|-107.0***|17.0***|-11.4**|-29.4***|2.6|-4.6**|313.3***|\\n|24-hour recall arm|0.3|-57.8***|12.6***|-11.6***|-27.4***|0.5|-2.3|477.7***|\\n\\n## <u>Notes:</u>\\n\\n(1) OLS regressions controlling for EA and day of the week FEs; the standard errors are clustered at the household-level. Regressions control for the same individual and household-level characteristics as in eq. (1). Asterisks are indicated on the smartphone diary and 24hour recall estimates if they were significantly different, respectively, from the 7-day module. ***p<0.01, **p<0.05, *p<0.10\\n\\n# 5. Factors underlying reporting differences across recall and smartphone arms: Intraday variation in time allocation and simultaneity\\n\\nApart from average treatment effects, the data collected across both treatment arms in the experimental study allow for a more disaggregated view of time allocation that has been less well-examined in time use surveys. This includes how men and women manage their time within the day — as well as how intensively they use their time, as measured by different activities being conducted simultaneously. As we discuss in this section, some of these nuances can help explain some of the findings we observed in Table 6. For example, despite the 24-hour recall arm’s 15-minute floor for time recording, we find that the smartphone arm reveals greater time across leisure and non-leisure activities during potentially productive times of the day. We also find a much higher incidence of activities conducted simultaneously in the smartphone arm, and across a wider range of activities than the recall arm, including non-leisure activities. In so doing, we\\n\\n> 17 The results are available upon request.\\n\\n> 18 For treatment effects, all days between days 3-10 were used in the smartphone sample, and days 5, 8 and 10 were used in the recall arm.\\n\\n19\", \"datasets\": [{\"mention_name\": {\"text\": \"smartphone diary\", \"confidence\": 0.5022105574607849, \"start\": 2169, \"end\": 2185, \"clean_text\": \"smartphone diary\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2169, \"end\": 2185, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2169, \"end\": 2185, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2169, \"end\": 2185}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2169, \"end\": 2185, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9874386191368103, \"start\": 324, \"end\": 340}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.5022105574607849, \"start\": 324, \"end\": 340}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5967628359794617, \"start\": 324, \"end\": 340}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5967628359794617, \"start\": 324, \"end\": 340}}, {\"mention_name\": {\"text\": \"24hour recall estimates\", \"confidence\": 0.31451570987701416, \"start\": 2190, \"end\": 2213, \"clean_text\": \"24hour recall estimates\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2190, \"end\": 2213, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2190, \"end\": 2213, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2190, \"end\": 2213}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2190, \"end\": 2213, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.9874386191368103, \"start\": 345, \"end\": 368}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.31451570987701416, \"start\": 345, \"end\": 368}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5967628359794617, \"start\": 345, \"end\": 368}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5967628359794617, \"start\": 345, \"end\": 368}}, {\"mention_name\": {\"text\": \"time use surveys\", \"confidence\": 0.8892910480499268, \"start\": 2657, \"end\": 2673, \"clean_text\": \"time use surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2657, \"end\": 2673, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2657, \"end\": 2673, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2657, \"end\": 2673}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2657, \"end\": 2673, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.95928955078125, \"start\": 336, \"end\": 352}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8892910480499268, \"start\": 336, \"end\": 352}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.45488405227661133, \"start\": 336, \"end\": 352}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.45488405227661133, \"start\": 336, \"end\": 352}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [20]}}, {\"page\": 21, \"chunk\": 21, \"input_text\": \"discuss how real-time, smartphone data collection can reveal greater insights on individuals’ opportunity costs of time, along with a wider view of different activities they are involved in.\\n\\n# 5.1 Intraday variation in time allocation\\n\\nAcross treatment arms, when looking at the intraday distribution of time across key activity categories (presented in Figure 5 by local polynomial smooth plots of the share of activities reported throughout the day),<sup>19</sup> we observe large differences in time allocation across such areas as agriculture, nonfarm employment, leisure, and unpaid work. While non-sleep, non-leisure activity reporting is much higher early in the day for the recall arm, there is a substantial drop-off in reporting for nearly all non-sleep activities by mid-day or early evening around 6 pm. The smartphone arm, on the other hand, reveals much more evenly distributed reporting, with drop-offs towards sleep only much later in the day (around 9 pm). The share of daily time in activities after 6pm accounts for about 26 percent for women and 29 percent for men the smartphone sample—compared to 17 and 19 percent for women and men in the recall sample, respectively—leading to a substantial loss of information when moving to recall-based methods. Furthermore, Table 9 highlights how time in activities such as unpaid work and non-farm employment begins to fall significantly in the recall arm vis-à-vis the smartphone arm after 12 pm, along with agriculture and leisure after 6 pm. Substantial amounts of daily time are spent by smartphone respondents in unpaid work and non-farm employment after 6 pm (17 and 33 minutes in nonfarm employment for women and men, respectively, and 83 and 17 minutes, respectively, in unpaid work). Although the recall arm tends to report higher time in leisure throughout the day, this trend flips after 6 pm as well. Men and women in the smartphone arm also spend nearly twice as much time in transport-related activities after 6 pm, compared to the recall arm. These trends are likely a reason for underreporting of participation across key activities in the recall vis-à-vis smartphone arm (Table 6).\\n\\n> 19 The results are based on the pooled Day 5, 8 and 10 samples. 95 percent confidence intervals are included as well.\\n\\n20\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [21]}}, {\"page\": 22, \"chunk\": 22, \"input_text\": \"<!-- Start of picture text --> 02 —_I~~<br>/f SSN<br>oe | . 4oX.\\\\ gei \\\\ 8<br>ot i : \\\\\\\\ 4<br>fie yk<br>/ j \\\\ ‘4<br>oe fii i\\\\ 3<br>=== te —~~~ Women,Women, smartphone recall \\\\ ood<br>0 6 12 18 24<br>Time ofday<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> 025<br>ot |fo,’ WSNY i\\\\OA \\\\ 1<br>on }, \\\\ 4<br>iie yo\\\\ 4<br>005 J yo<br>of Ne” — Women, recall “we<br>--~ Women, smartphone<br>() é 12 8 24<br>Time of day<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> »<br>7<br>I~ f/f TM<br>015: fis aN LEN<br>fi ee is fit<br>iri YA \\\\ 4<br>0 [iF \\\\ 4<br>HE ey3<br>ons: };fi \\\\\\\\ \\\\84<br>/ / — Women, recall \\\\NY<br>0.) === ~~ Women, smartphone =<br>0 6 Time ofday12 8 24<br><!-- End of picture text --> <!-- Start of picture text --> 02 /im\\\\N<br>/ “ .<br>015: |eaeesZ ee— f<br>fi \\\\\\\\ ‘ ‘<br>or } \\\\ 4<br>fi: \\\\\\\\ 4<br>005 }fi \\\\XA4<br>4 LFfoes —~~ Men,Men, recallsmartphone “Ns<br>6 é 2 8 24<br>Time ofday<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> a<br>a SXi et<br>a ink<br>ot fis/ 3\\\\<br>/ , NN H<br>—_— — Men, recall<br>O4 bessccese Pe - Men, smartphone<br>0 6 12 8 24<br>Time ofday<br><!-- End of picture text --> <!-- Start of picture text --> 025 AAoN<br>o2 | \\\\<br>/ eae<br>O15 { Nee ;<br>I: eS<br>o1 iff N33a<br>os rE| ji \\\\anay<br>A:/ — Men, recall \\\\\\\\<br>04 eee ~~ Men, smartphone “Ss<br>0 6 Time 12of day 18 a<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> 04 ivon= \\\\<br>i \\\\<br>03; fow\\\\i \\\\<br>02 | F \\\\<br>} \\\\<br>fi} i; \\\\ fat-<br>o }fi;: MSNY : .<br>of ———*y —~~ Women,Women, recallsmartphone \\\\aL.<br>0 6 2 18 24<br>Time ofday<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> 025<br>ate [ud. ff —~ \\\\ |<br>/ \\\\ 3<br>on / \\\\\\\\ 4 i<br>os /, \\\\ ‘<br>y — Women, recall ts<br>0 eet ~~ Women, smartphone —<br>0 6 12 8 24<br>Time of day<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> 025<br>o a on,<br>atin an<br>O15 { ; ee \\\\<br>/, \\\\ 4<br>o1 I,i Vi\\\\<br>os /fif \\\\\\\\ 3<br>fi/ — Women, recall \\\\\\\\<br>04 ——“\\\" + Women, smartphone cree<br>0 6 Time of12 day 8 a<br><!-- End of picture text --> <!-- Start of picture text --> 04<br>rs.<br>i \\\\<br>03. |} \\\\\\\\<br>02 |} ‘ ‘h<br>| ‘<br>fif iv: ¥\\\\ mY..<br>ot a ; Korey,all \\\\<br>04 SeeLe —~ Men,Men, recallsmartphone \\\\——hee<br>() é 2 8 m4<br>Timeof day<br><!-- End of picture text -->\\n\\n<!-- Start of picture text --> 025,<br>os aTae- \\\\\\\\<br>Jour \\\\<br>on /{- 43i<br>005 fi \\\\ 4<br>4 — Men, recall Ss<br>0 wt ~~ Men, smartphone<br>() é 12 8 24<br>Timeof day<br><!-- End of picture text --> <!-- Start of picture text --> 02 anlan<br>for. / Vi<br>ats | eel \\\\<br>i i Nik<br>» [: \\\\<br>|hi Vy:<br>one fii \\\\\\\\Vi,<br>fei: — Men, recall \\\\Naf<br>04 a“ -- Men, smartphone: ——<br>0 6 Time2of day 18 4<br><!-- End of picture text -->\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [22]}}, {\"page\": 23, \"chunk\": 23, \"input_text\": \"Table 9. Minutes spent within the day across key activity categories, by treatment arm\\n\\n||T|otal|W|omen||Men|\\n|---|---|---|---|---|---|---|\\n||Recall|Smartphone|Recall|Smartphone|Recall|Smartphone|\\n|Morning (12am-11:59am)|||||||\\n|Agriculture|59.6***|35.2***|53.1***|35.3***|66.2***|35.1***|\\n|Non-agriculture|54.9***|18.1***|28.4***|11.1***|81.7***|24.9***|\\n|Leisure|158.3***|75.4***|135.2***|61.7***|181.7***|89.0***|\\n|Unpaid work|109.2***|82.6***|177.4***|134.4***|40.0***|31.4***|\\n|Transport|62.8|65.7|55.2|52.7|70.4**|78.5**|\\n|Eating/drinking|8.4***|6.8***|8.3***|6.6***|8.4**|7.0**|\\n|Afternoon/evening/night (12pm-11:59pm)|||||||\\n|Agriculture|18.9|17.9|15.7|16.1|22.3|19.7|\\n|Non-agriculture|58.2***|70.1***|29.5**|41.1**|87.2|98.7|\\n|Leisure|306.2**|294.1**|267.8|256.6|345.2|331.1|\\n|Unpaid work|103.7***|133.3***|176.6***|210.3***|29.8***|57.1***|\\n|Transport|64.7***|95.1***|48.7***|64.5***|81.0***|125.3***|\\n|Eating/drinking|44.4***|39.9***|42.7|40.6|46.1***|39.2***|\\n|Evening/night (6pm-11:59pm)|||||||\\n|Agriculture|1.1***|3.2***|0.7***|2.2***|1.5**|4.1**|\\n|Non-agriculture|14.3***|25.1***|6.0***|16.9***|22.7**|33.3**|\\n|Leisure|106.8***|133.5***|86.5***|106.3***|127.4***|160.5***|\\n|Unpaid work|31.9***|49.6***|56.0***|82.6***|7.5***|17.0***|\\n|Transport|13.7***|34.5***|7.0***|20.5***|20.6***|48.4***|\\n|Eating/drinking|21.7|21.1|20.7|21.4|22.7*|20.8*|\\n\\nNotes: (1) T-tests of equality of means conducted across treatment groups; ***p<0.01, **p<0.05, *p<0.10.\\n\\n(2) Based on a pooled individual-level sample for days 5, 8 and 10 (each individual appeared three times in the sample).\\n\\n(3) Grey shading = recall arm yields significantly lower estimates.\\n\\nOne possibility for these differences is respondent fatigue in the recall arm, given the structure and relatively long duration of the interview. Respondents might be more actively reporting time in different productive activities at the start of the interview, and by the end of the narrative (i.e., for hours later in the day) they lose interest or track of their time. Figure 6 gives some indication that this may be the case, revealing a bimodal sleep distribution for recall respondents that is particularly pronounced for women. A significantly greater share of recall respondents, particularly women, report earlier sleep times that would conclude the diary interview — 64 percent of women in the recall arm reported sleep times between 5 pm8:59 pm, versus 54 percent in the smartphone arm. For men, these shares were 50 and 45 percent, respectively. This supports, to some extent, that fatigue may be a key factor. Other factors may be at play as well, however, including potential difficulties among those reporting earlier sleep times to pinpoint in the interview exactly when they went to sleep.<sup>20</sup>\\n\\n> 20 Also see Kaplan et al., 2020, for a discussion on recalling sleep times using the American Time Use Survey.\\n\\n22\", \"datasets\": [{\"mention_name\": {\"text\": \"pooled individual-level sample\", \"confidence\": 0.3691243827342987, \"start\": 1480, \"end\": 1510, \"clean_text\": \"pooled individual-level sample\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1480, \"end\": 1510, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1480, \"end\": 1510, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1480, \"end\": 1510}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1480, \"end\": 1510, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"other\", \"confidence\": 0.7982523441314697, \"start\": 194, \"end\": 224}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.3691243827342987, \"start\": 194, \"end\": 224}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7488421201705933, \"start\": 194, \"end\": 224}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7488421201705933, \"start\": 194, \"end\": 224}}, {\"mention_name\": {\"text\": \"American Time Use Survey\", \"confidence\": 0.9196761250495911, \"start\": 2864, \"end\": 2888, \"clean_text\": \"American Time Use Survey\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2864, \"end\": 2888, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2864, \"end\": 2888, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2864, \"end\": 2888}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2864, \"end\": 2888, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.9867161512374878, \"start\": 297, \"end\": 321}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.9196761250495911, \"start\": 297, \"end\": 321}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.9345705509185791, \"start\": 297, \"end\": 321}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.9345705509185791, \"start\": 297, \"end\": 321}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [23]}}, {\"page\": 24, \"chunk\": 24, \"input_text\": \"Figure 6. Kernel density graphs: time of sleep for men and women <!-- Start of picture text --> (6a) Recall versus smartphone: women (6b) Recall versus smartphone: men<br><!-- End of picture text --> <u>Notes:</u> (1) The estimates pool the Day 5, 8 and 10 samples for each treatment arm. K-S p-values for differences across treatment arms in (6a) and (6b): 0.000.\\n\\nTherefore, the smartphone arm provides greater variation in intraday reporting and could potentially provide an easier way to avoid issues around understanding time in the day – given how early people are reporting sleep time. If time use statistics have typically been analyzed on a day-level aggregation, however, a natural question arises as to why we should potentially care about intraday time allocation. In addition to our understanding of how men and women organize their time, the distinctions we observed across treatment arms of reporting during morning versus evening hours also have important implications for difficult-to-answer questions around infrastructural interventions affecting time use. This includes, for example, the role of household electrification in supporting time allocation in different productive areas during evening and nighttime hours (see, for example, Agénor and Agénor, 2023; Ribeiro et al., 2021; van de Walle, et. al., 2017). Taking electricity and access to energy sources more broadly, high-quality intraday data on time allocation can help us better understand the role of energy infrastructure on individuals’ economic and other intrahousehold decision making. As an example, Appendix Table A6 presents OLS regressions showing that for women in the smartphone arm in particular, electricity access (as opposed to no access, or household lighting through battery-operated torches) are significantly associated with time allocation in nonfarm employment and leisure during the evening/nighttime, but with no effects for women in the recall sample.<sup>21</sup> The choice of survey approach to time use data collection can therefore provide very different perspectives in understanding how time allocation is linked to energy infrastructure and transport.<sup>22</sup>\\n\\n> 21 The regressions control for the same right hand side variables as in Equation (1), and examine effects separately by treatment arm, men and women, as well as employed men and women — given the findings in Figure 5 and Table 9 regarding agriculture and nonfarm employment.\\n\\n> 22 In separate estimations, we also found fewer effects of electricity or motorized transport on time allocation in the morning hours across both treatment arms, supporting the relevance of the intraday distinction.\\n\\n23\", \"datasets\": [{\"mention_name\": {\"text\": \"time use statistics\", \"confidence\": 0.7863630652427673, \"start\": 596, \"end\": 615, \"clean_text\": \"time use statistics\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 596, \"end\": 615, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 596, \"end\": 615, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 596, \"end\": 615}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 596, \"end\": 615, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"census\", \"confidence\": 0.6024600267410278, \"start\": 596, \"end\": 615}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7863630652427673, \"start\": 596, \"end\": 615}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.7933728694915771, \"start\": 596, \"end\": 615}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.7933728694915771, \"start\": 596, \"end\": 615}}, {\"mention_name\": {\"text\": \"intraday data\", \"confidence\": 0.7721190452575684, \"start\": 1408, \"end\": 1421, \"clean_text\": \"intraday data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1408, \"end\": 1421, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1408, \"end\": 1421, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1408, \"end\": 1421}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 1408, \"end\": 1421, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"microdata\", \"confidence\": 0.5693390369415283, \"start\": 758, \"end\": 771}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7721190452575684, \"start\": 758, \"end\": 771}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6657208800315857, \"start\": 758, \"end\": 771}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6657208800315857, \"start\": 758, \"end\": 771}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [24]}}, {\"page\": 25, \"chunk\": 25, \"input_text\": \"# 5.2 Simultaneity of activities\\n\\nAccurately measuring multitasking — when two or more activities are conducted simultaneously — is critical for measuring total work burden and gender differences in time allocation (Craig and Brown, 2016). A longstanding literature has pointed to the relevance of data on multitasking for improving our estimates of women’s participation and time in unpaid work — based on traditional time use surveys showing that women regularly conduct different unpaid domestic and care activities together, as well as alongside other activities in employment and leisure (Floro, 1995; Budig and Folbre, 2004; Irani and Vemireddy, 2021). Accurate quantification of these overlapping time constraints can in turn inform the effective design of care policies and measuring progress towards the Sustainable Development Goal to recognize and value unpaid work across countries. Moreover, accurate quantification of the incidence and time spent in simultaneously conducted activities can help us answer open questions around the links between multitasking and a range of outcomes related to health, leisure, productivity, and stress (Bittman and Wajcman, 2000; Floro and Miles, 2003; Manhart, 2004; Craig and Brown, 2016, 2017; Zaiceva-Razzolini, 2022).\\n\\nIn this section, we show that the mode of data collection also matters in assessing the role of multitasking in individuals’ time allocation, and with implications for understanding of women’s unpaid work burden. As discussed earlier, the smartphone diary allowed respondents to select multiple activities they were involved in at a given time, allowing for a continuous examination of how different activities were layered with each other throughout the day. The recall diary, alternatively, recorded (when applicable) a single secondary activity that respondents were involved in alongside the main activity for a given 15-minute interval. Appendix Figure A1 shows that the recall diary preface read out to respondents suggested secondary activities that might include collecting water or cooking, alongside caring for children. This yields different approaches to capturing simultaneous activities — while the smartphone diary did not distinguish between main and secondary activities (every activity was selected on its own, without any limits on the number of simultaneous activities that can be recorded), the recall diary did distinguish between the two, and encouraged respondents to consider a specific secondary activity that might have typically been done alongside the main one. As a result, close to 10 percent of respondents in the smartphone treatment arm often reported three or more activities (although typically not more than three) within a 15-minute or shorter duration, compared to the maximum ceiling of two activities in the recall diary.\\n\\nThe analysis shows that nearly all (around 95 percent, for both treatment arms) of women’s daily multitasked non-leisure time was from conducting different unpaid domestic and care activities together. However, if we consider leisure activities (see Appendix Table A1 for the breakdown of leisure activities), the share of women’s multitasked time in unpaid work was greater in the smartphone arm (38 percent, compared to 23 percent in the recall arm). As discussed below, this stems from greater reporting in the recall arm of multitasked time in leisure; the smartphone arm, on the other hand, revealed higher participation in multitasked employment, unpaid work, and transport activity.\\n\\nTo provide a richer understanding of different types of activities that were conducted together in both treatment arms, Tables 10 and 11 present heat map charts for the share of women and men, respectively, 24\", \"datasets\": [{\"mention_name\": {\"text\": \"time use surveys\", \"confidence\": 0.8119231462478638, \"start\": 419, \"end\": 435, \"clean_text\": \"time use surveys\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 419, \"end\": 435, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 419, \"end\": 435, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 419, \"end\": 435}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 419, \"end\": 435, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.6139600276947021, \"start\": 419, \"end\": 435}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.8119231462478638, \"start\": 419, \"end\": 435}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.6970636248588562, \"start\": 419, \"end\": 435}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.6970636248588562, \"start\": 419, \"end\": 435}}, {\"mention_name\": {\"text\": \"recall diary\", \"confidence\": 0.521350622177124, \"start\": 2387, \"end\": 2399, \"clean_text\": \"recall diary\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2387, \"end\": 2399, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2387, \"end\": 2399, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"15-minute interval\", \"confidence\": 0.6135610342025757, \"start\": 1893, \"end\": 1911}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2387, \"end\": 2399, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"survey\", \"confidence\": 0.6198588609695435, \"start\": 645, \"end\": 657}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.32953450083732605, \"start\": 645, \"end\": 657}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.715058445930481, \"start\": 645, \"end\": 657}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.715058445930481, \"start\": 645, \"end\": 657}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [25]}}, {\"page\": 26, \"chunk\": 26, \"input_text\": \"conducting different activities together for the pooled matched samples of days 5, 8 and 10.<sup>23</sup> The tables are broken out by aggregate activity category, excluding sleep. Since the data in each treatment arm were collected by disaggregated activities within these larger activity categories, the tables include cases where different activities within a given aggregate category were conducted with each other. A key finding is that the smartphone arm, for both men and women, provides a much more diverse picture of different activities being conducted together than the recall arm. The latter primarily revealed leisure being conducted with different activities (including other leisure activities). In the smartphone arm, for example, 17 percent of women and 13 percent of men conducted transport activities while also obtaining services/resources for the household, compared to zero in the recall arm. In addition, greater shares of men and women in the smartphone arm conducted unpaid activities with agriculture and nonfarm employment, as well as transport with employment. While the smartphone arm also captured a large share of respondents conducting leisure with other activities — and the extent to which leisure is “shared” across other activities is also important to understand — the smartphone approach still reveals a much richer view of different combinations of activities individuals are involved in.\\n\\n> 23 Sleep was excluded from the table since it was rarely reported together with other activities.\\n\\n25\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [26]}}, {\"page\": 27, \"chunk\": 27, \"input_text\": \"Table 10. Average share of respondents conducting distinct activities together within a day (pairwise distribution), women\\n\\n|||||S|martphone|sample|||||||\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n|||Agri-<br>culture|Nonfarm<br>employ-<br>Ment|Obtaining<br>services/<br>resources<br>for HH|Cooking,<br>cleaning<br>in house|Care of<br>children,<br>elderly|Transport|School/<br>studying|Leisure|Comm.<br>Activitie<br>s|Personal<br>care|<br>Eating/<br>drinking|\\n|||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|(9)|(10)|(11)|\\n|Agriculture|(1)|0.03|||||||||||\\n|Nonfarm employment|(2)|0.01|0.01||||||||||\\n|Obtaining services/<br>resources for HH|(3)|0.01|0.03|0.04|||||||||\\n|Cooking, cleaning in house|(4)|0.04|0.05|0.14|0.02||||||||\\n|Care of children, elderly|(5)|0.03|0.03|0.07|0.17|0.01|||||||\\n|Transport|(6)|0.05|0.05|0.17|0.11|0.11|0.14||||||\\n|School/studying|(7)|0.00|0.00|0.00|0.00|0.00|0.01|0.00|||||\\n|Leisure|(8)|0.09|0.08|0.11|0.24|0.16|0.24|0.03|0.36||||\\n|Community Activities|(9)|0.00|0.00|0.00|0.01|0.01|0.01|0.00|0.02|0.00|||\\n|Personal care|(10)|0.01|0.01|0.01|0.07|0.04|0.05|0.00|0.11|0.01|0.02||\\n|Eating/drinking|(11)|0.02|0.04|0.02|0.18|0.09|0.04|0.00|0.28|0.00|0.07|<br>0.01|\\n|||||R|ecall sampl|e|||||||\\n|||Agri-<br>culture|Nonfarm<br>employ-<br>Ment|Obtaining<br>services/<br>resources<br>for HH|Cooking,<br>cleaning<br>in house|Care of<br>children,<br>elderly|Transport<br>|School/<br>studying|Leisure|Comm.<br>Activities|Personal<br>care|Eating/<br>drinking|\\n|||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|(9)|(10)|(11)|\\n|Agriculture|(1)|0.00|||||||||||\\n|Nonfarm employment|(2)|0.00|0.00||||||||||\\n|Obtaining services/<br>resources for HH|(3)|0.00|0.00|0.00|||||||||\\n|Cooking, cleaning in house|(4)|0.01|0.02|0.02|<br>0.00||||||||\\n|Care of children, elderly|(5)|0.03|0.04|0.08|<br>0.21|<br>0.00|||||||\\n|Transport|(6)|0.00|0.00|0.00|<br>0.00|<br>0.17|0.00||||||\\n|School/studying|(7)|0.00|0.00|0.00|<br>0.00|<br>0.00|0.00|0.00|||||\\n|Leisure|(8)|0.10|0.11|0.15|<br>0.47|<br>0.23|0.34|0.01|0.23||||\\n|Community Activities|(9)|0.00|0.00|0.00|<br>0.00|<br>0.01|0.00|0.00|0.01|0.00|||\\n|Personal care|(10)|0.00|0.00|0.00|<br>0.01|<br>0.04|0.00|0.00|0.06|0.00|0.00||\\n|Eating/drinking|(11)|0.00|0.01|0.00|<br>0.01|<br>0.16|0.00|0.00|0.45|0.00|0.00|0.00|\\n\\nNotes: (1) Pooled sample of individuals across days 5, 8 and 10.\\n\\n26\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [27]}}, {\"page\": 28, \"chunk\": 28, \"input_text\": \"Table 11. Average share of respondents conducting distinct activities together within a day (pairwise distribution), men\\n\\n|||||S|martphon|e sample|||||||\\n|---|---|---|---|---|---|---|---|---|---|---|---|---|\\n|||Agri-<br>Cultur<br>e|Nonfarm<br>employ-<br>ment|Obtaining<br>services/<br>resources<br>for HH|Cooking,<br>cleaning<br>in house|Care of<br>children,<br>elderly|Transport|School/<br>studying|Leisure|Comm.<br>Activitie<br>s|Personal<br>care|Eating/<br>drinking|\\n|||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|(9)|(10)|(11)|\\n|Agriculture|(1)|0.03|||||||||||\\n|Nonfarm employment|(2)|0.01|0.01||||||||||\\n|Obtaining services/<br>resources for HH|(3)|0.02|0.04|0.02|||||||||\\n|Cooking, cleaning in house|(4)|0.02|0.01|0.02|0.00||||||||\\n|Care of children, elderly|(5)|0.00|0.00|0.01|0.01|0.00|||||||\\n|Transport|(6)|0.06|0.10|0.13|0.04|0.02|0.18||||||\\n|School/studying|(7)|0.00|0.00|0.00|0.00|0.00|0.01|0.01|||||\\n|Leisure|(8)|0.10|0.12|0.11|0.10|0.04|0.41|0.03|0.50||||\\n|Community Activities|(9)|0.00|0.00|0.00|0.00|0.00|0.02|0.00|0.03|0.00|||\\n|Personal care|(10)|0.01|0.02|0.01|0.02|0.01|0.08|0.00|0.15|0.01|0.01||\\n|Eating/drinking|(11)|0.01|0.07|0.02|0.04|0.01|0.11|0.00|0.39|0.01|0.06|0.01|\\n||||||Recall s|ample|||||||\\n|||Agri-|Nonfarm|Obtaining|Cooking,|Care of|Transport|School/|Leisure|Comm.<br>Pe|rsonal care|Eating/|\\n|||culture|employ-|services/|cleaning|children,||Studying||Activities||drinking|\\n||||ment|resources<br>for HH|in house|elderly|||||||\\n|||(1)|(2)|(3)|(4)|(5)|(6)|(7)|(8)|(9)|(10)|(11)|\\n|Agriculture|(1)|0.00|||||||||||\\n|Nonfarm employment|(2)|0.00|0.00||||||||||\\n|Obtaining services/<br>resources for HH|(3)|0.00|0.01|<br>0.00|||||||||\\n|Cooking, cleaning in house|(4)|0.00|0.00|<br>0.00|0.00||||||||\\n|Care of children, elderly|(5)|0.00|0.00|<br>0.00|0.00|0.00|||||||\\n|Transport|(6)|0.01|0.03|<br>0.02|0.00|0.01|0.00||||||\\n|School/studying|(7)|0.00|0.00|<br>0.00|0.00|0.00|0.00|0.00|||||\\n|Leisure|(8)|0.11|0.22|<br>0.10|0.11|0.04|0.43|0.01|0.42||||\\n|Community Activities|(9)|0.00|0.00|<br>0.00|0.00|0.00|0.00|0.00|0.01|0.00|||\\n|Personal care|(10)|0.00|0.01|<br>0.00|0.00|0.00|0.01|0.00|0.03|0.00|0.00||\\n|Eating/drinking|(11)|0.00|0.03|<br>0.00|0.00|0.01|0.01|0.00|0.72|0.00|0.01|0.00|\\n\\nNotes: (1) Pooled sample of individuals across days 5, 8 and 10.\\n\\n27\", \"datasets\": [], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [28]}}, {\"page\": 29, \"chunk\": 29, \"input_text\": \"Table 12 also provides descriptive statistics on both the share of individuals and average daily shared time (in minutes) in conducting activities simultaneously, across different activity combinations — and, with a focus on how unpaid work, transport and employment were conducted with other activities. With a few exceptions, the share of individuals conducting activities simultaneously across men and women is significantly higher in the smartphone arm. Looking strictly at non-leisure activities, 74 percent of women and 60 percent of men in the smartphone arm conducted non-leisure activities together, compared to 37 and 13 percent of women and men in the recall sample (these shares for the recall arm were 84 and 85 percent, respectively, for all activities). This reinforces the role of leisure in simultaneous activity reporting in the recall diary approach. Within the smartphone arm, 31 percent of women and 17 percent of men also conducted unpaid work with transport. These shares were 10 and 15 percent, respectively, for transport with either agricultural or nonfarm employment. The shares in the recall arm, on the other hand, were much lower. Unpaid work and agricultural or nonfarm employment were conducted together among 15 percent of women and 7 percent of men in the smartphone sample, compared to 9 and 1 percent of women and men, respectively, in the recall arm.\\n\\nOnly for women’s care work is participation significantly higher in the recall arm, regardless of whether the other activity was in leisure or non-leisure. This may be due to the preface of the 24-hour diary, specifically suggesting care work as a potential secondary activity that the respondents could include (see Appendix Figure A1). For women, cooking/cleaning with any other activity was significantly higher in the recall arm, but much of this is due to leisure being the other activity, since the trend flips in Table 12 (B2). On reported time, however, since activities in the recall sample could be conducted together for a minimum of 15 minutes, the recall arm often reveals significantly higher time conducted together in simultaneous activities – though, the magnitude of the differences is quite small given the substantially lower participation rates, and, in the case of men, often not significant. An understudied area, however, that the smartphone data provides some initial insights on, is the prevalence of passive care, where children are present with adults but not directly engaged in their activities (Folbre et al., 2005). Through the _TimeTracker_ pop-up data on whom activities were conducted with (Table 13), we find that about 16 percent of women’s daily time, or about 144 minutes daily, is spent with young children (aged 0-9 years) in activities that are not actively reported as care or sleep (and thus over and above the direct time reported earlier in childcare). Most of these activities are in other areas of unpaid work, transport, or leisure, and while performing these activities, women spend roughly twice as much time with young children compared to men. Gender differences in the extent of time devoted to activities in the presence of older children (aged 10-17 years), however, are narrower, reflecting important differences in the way we need to consider passive care in time use data collection. Although the _TimeTracker_ pop-ups provide an imperfect view of passive care, the data shed some light on a significantly less-understood dimension of childcare that may not be wellcaptured by simply asking individuals whether they were involved in childcare provision.\\n\\nTable 14 also presents treatment effects, using the same structure as Equation (1), and looking at the recall arm’s effect on whether (1) activities were conducted together under different combinations, and (2) shared minutes across those activities. The estimates are presented for the total, women, and men samples, as well as by all areas/urban/rural. The results underscore the significantly lower incidence simultaneous activities in the recall arm, even when including leisure activities in column (A1). This holds 28\", \"datasets\": [{\"mention_name\": {\"text\": \"smartphone data\", \"confidence\": 0.7337389588356018, \"start\": 2344, \"end\": 2359, \"clean_text\": \"smartphone data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2344, \"end\": 2359, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2344, \"end\": 2359, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2344, \"end\": 2359}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2344, \"end\": 2359, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"microdata\", \"confidence\": 0.24693447351455688, \"start\": 651, \"end\": 666}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.7337389588356018, \"start\": 651, \"end\": 666}, \"usage_context\": {\"text\": \"supporting\", \"confidence\": 0.48465800285339355, \"start\": 651, \"end\": 666}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.48465800285339355, \"start\": 651, \"end\": 666}}, {\"mention_name\": {\"text\": \"_TimeTracker_ pop-up data\", \"confidence\": 0.6797800064086914, \"start\": 2549, \"end\": 2574, \"clean_text\": \"_TimeTracker_ pop-up data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2549, \"end\": 2574, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2549, \"end\": 2574, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2549, \"end\": 2574}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2549, \"end\": 2574, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"database\", \"confidence\": 0.32760369777679443, \"start\": 52, \"end\": 77}, \"specificity_tag\": {\"text\": \"named\", \"confidence\": 0.6797800064086914, \"start\": 52, \"end\": 77}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5762267708778381, \"start\": 52, \"end\": 77}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5762267708778381, \"start\": 52, \"end\": 77}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [29]}}, {\"page\": 30, \"chunk\": 30, \"input_text\": \"except for men in urban areas. The magnitude of underreporting widens across non-leisure activities, as well as carrying out unpaid work and agricultural/nonfarm employment with non-leisure activities. On shared time across non-leisure activities, apart from the large positive effects of 24-hour recall treatment on conducting unpaid work with non-leisure among women, most of the recall treatment effects on shared minutes lose significance, and the recall treatment has a significant negative effect on shared minutes among rural men’s non-leisure activities (column B2). Overall, the smartphone arm provides a much more granular view of the extent to which activities are conducted together, across leisure and non-leisure activities.\\n\\nTable 12. Descriptive statistics on simultaneity of activities by treatment arm\\n\\n||Share of|individuals c<br>simultane|onducting a<br>ously (s)|ctivities|Average<br>differen|daily share<br>t activities si|d time in con<br>multaneousl|ducting<br>y (mins)|\\n|---|---|---|---|---|---|---|---|---|\\n||Wo<br>|men|Me<br>|n|Wom<br>|en|M<br>|en|\\n||Smart-||Smart-||Smart-||Smart-||\\n||phone|Recall|phone|Recall|phone|Recall|phone|Recal|\\n|(A) Any activities together:|||||||||\\n|Across any activity categories|0.89***|0.84***|0.86***|0.85***|125.0***|267.3***|109.6***|272.3***|\\n|Across non-leisure activities|0.74***|0.37***|0.60***|0.13***|52.4**|61.7**|15.7*|11.9*|\\n|(B1 Unpaid work with any activity|0.68|0.69|0.38***|0.25***|79.3***|150.8***|13.1***|25.8***|\\n|Specific areas of unpaid work with<br>|||||||||\\n|any activity:|||||||||\\n|Obtaining services/resources for HH|0.38***|0.24***|0.25***|0.13***|10.8***|14.1***|6.6***|12.9***|\\n|Cooking and cleaning|0.56***|0.62***|0.17***|0.12***|47.3***|79.9***|4.4***|7.8***|\\n|Care of children and/or elderly|0.31***|0.36***|0.08***|0.05***|42.7***|82.1***|2.3***|5.7***|\\n|(B2) Unpaid work w/any non-leisure activity|0.61***|0.35***|0.29***|0.05***|48.0***|60.7***|4.2|3.2|\\n|Specific areas of unpaid work with any|||||||||\\n|non-leisure activity:|||||||||\\n|Obtaining services/resources for HH|0.35***|0.11***|0.20***|0.02***|8.8***|5.4***|2.1|1.7|\\n|Cooking and cleaning|0.48***|0.26***|0.12***|0.02***|30.5***|23.9***|1.7***|0.8***|\\n|Care of children and/or elderly|0.27**|0.30**|0.05***|0.02***|32.6***|56.6***|0.8|1.3|\\n|(B3) Unpaid work w/any transport activity|0.31***|0.17***|0.17***|0.02***|8.7***|14.3***|1.7|1.5|\\n|Specific areas of unpaid work with any|||||||||\\n|transport activity:|||||||||\\n|Obtaining services/resources for HH|0.17***|0.01***|0.13***|0.02***|2.1***|0.07***|1.1|0.8|\\n|Cooking and cleaning|0.11***|0.002***|0.04***|0.002***|1.4***|0.03***|0.3***|0.03***|\\n|Care of children and/or elderly|0.11***|0.17***|0.02**|0.01**|5.6***|14.2***|0.3|0.6|\\n|(C) Agricultural or non-agr. work with:|||||||||\\n|Any activity|0.34***|0.26***|0.35|0.37|20.8***|52.4***|26.3***|100.6***|\\n|Any non-leisure activity|0.26***|0.10***|0.26***|0.07***|10.2*|13.2*|7.1|8.8|\\n|Any unpaid work|0.15***|0.09***|0.07***|0.01***|7.7***|12.3***|1.2|0.9|\\n|Any transport activity|0.10***|0.002***|0.15***|0.03***|0.8***|0.04***|2.7|4.7|\\n\\nNotes: (1) T-tests of equality of means conducted across treatment arms, for men and women separately; ***p<0.01, **p<0.05, *p<0.10.\\n\\n(2) Pooled sample of individuals across days 5, 8 and 10.\\n\\n29\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [30]}}, {\"page\": 31, \"chunk\": 31, \"input_text\": \"Table 13. Smartphone data: average daily time (mins), and share of daily time, spent in passive care across non-care and non-sleep activities\\n\\n|||Po|p-up data: activity|conducted with|:||\\n|---|---|---|---|---|---|---|\\n||Any ch<br>(aged<br>|ildren<br>0-17)<br>|Children a<br>|ged 0-9<br>|Children a<br>|ged 10-17<br>|\\n||Women|Men|Women|Men|Women|Men|\\n|Average daily minutes:|||||||\\n|Agriculture|8.8|7.0|3.1|1.8|6.2|5.5|\\n|Nonfarm employment|5.5***|16.3***|3.2|5.4|3.0***|13.0***|\\n|Unpaid work:|||||||\\n|Obtaining services/resources for HH|5.9***|2.4***|3.9***|0.8***|2.3|1.9|\\n|Cooking, cleaning in house|42.1***|5.1***|29.6***|1.6***|16.9***|3.7***|\\n|Transport|21.5|19.1|14.8***|5.4***|8.1***|14.7***|\\n|School/studying|3.8|5.8|0.02|0.5|3.7|5.3|\\n|Leisure|100.9|107.3|67.1***|48.2***|57.4**|73.3**|\\n|Community activities|1.7|2.7|0.6|0.2|1.2|2.7|\\n|Personal care|6.1**|1.2**|4.8**|0.6**|3.2|0.6|\\n|Eating/drinking|22.2***|13.9***|16.4***|7.5***|10.3***|7.6***|\\n|All non-care and non-sleep activities|218.4***|180.8***|143.5***|72.0***|112.3*|128.4*|\\n|Share of daily time (within activity<br>categories; among those involved):|||||||\\n|Agriculture|0.17**|0.12**|0.07***|0.03***|0.10|0.09|\\n|Nonfarm employment|0.13|0.11|0.08***|0.03***|0.05*|0.09*|\\n|Unpaid work:|||||||\\n|Obtaining services/resources for HH|0.13**|0.09**|0.09***|0.04***|0.04|0.06|\\n|Cooking, cleaning in house|0.18**|0.14**|0.13***|0.06***|0.07**|0.10**|\\n|Transport|0.19***|0.10***|0.14***|0.03***|0.07|0.08|\\n|School/studying|0.27|0.25|0.08|0.04|0.19|0.21|\\n|Leisure|0.30***|0.25***|0.21***|0.12***|0.15|0.17|\\n|Community activities|0.20|0.17|0.09**|0.02**|0.12|0.16|\\n|Personal care|0.07***|0.04***|0.05***|0.02***|0.02|0.03|\\n|Eating/drinking|0.50***|0.31***|0.36***|0.16***|0.22***|0.18***|\\n|All non-care and non-sleep activities|0.24***|0.19***|0.16***|0.07***|0.12*|0.13*|\\n\\n# <u>Notes:</u>\\n\\n(1) Based on 1,026 and 1,035 observations for women and men respondents, respectively, for Days 5, 8 and 10 of the survey period.\\n\\n(2) T-tests of equality of means conducted across men and women. ***p<0.01, **p<0.05, *p<0.10 (3) After each recorded activity in the _Time Tracker_ app, respondents were asked whom the activity was conducted with.\\n\\n30\", \"datasets\": [{\"mention_name\": {\"text\": \"Smartphone data\", \"confidence\": 0.4179115295410156, \"start\": 10, \"end\": 25, \"clean_text\": \"Smartphone data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 10, \"end\": 25, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 10, \"end\": 25, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 10, \"end\": 25}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 10, \"end\": 25, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"estimates\", \"confidence\": 0.45307546854019165, \"start\": 10, \"end\": 25}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.4179115295410156, \"start\": 10, \"end\": 25}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.40338069200515747, \"start\": 10, \"end\": 25}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.40338069200515747, \"start\": 10, \"end\": 25}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [31]}}, {\"page\": 32, \"chunk\": 32, \"input_text\": \"Table 14. Treatment effects: effect of recall arm on average daily participation and shared time (minutes) in conducting different activities simultaneously\\n\\n||(A)||(B|Effect of<br>|recall arm:<br>(C)||(D)||\\n|---|---|---|---|---|---|---|---|---|\\n||<br>Across any<br>categories (<br>leisur|<br>activity<br>including<br>e):|<br>Across no<br>activit|<br>n-leisure<br>ies:|<br>Unpaid work<br>non-leisure|<br>with any<br>activity:|<br>Work in agricu<br>farm emp. Wit<br>leisure ac|<br>lture/non-<br>h any non-<br>tivity:|\\n||Conducted<br>activities<br>together<br>(Y=1 N=0)|Shared<br>minutes|Conducted<br>activities<br>together<br>(Y=1 N=0)|Shared<br>minutes|Conducted<br>activities<br>together<br>(Y=1 N=0)|Shared<br>minutes|Conducted<br>activities<br>together<br>(Y=1 N=0)|Shared<br>minutes|\\n||(A1)|(A2)|(B1)|(B2)|(C1)|(C2)|(D1)|(D2)|\\n|All areas|||||||||\\n|Total|-0.03*|158.17***|-0.42***|3.14|-0.25***|6.45*|-0.17***|2.74|\\n|Women|-0.05***|144.92***|-0.36***|10.62|-0.25***|13.58**|-0.15***|4.46*|\\n|Men|-0.01|168.95***|-0.47***|-4.77|-0.24***|-1.03|-0.20***|0.76|\\n|Rural areas|||||||||\\n|Total|-0.09***|136.39***|-0.46***|3.65|-0.27***|10.17**|-0.20***|1.87|\\n|Women|-0.11***|131.90***|-0.42***|19.53**|-0.29***|23.27***|-0.17***|8.10**|\\n|Men|-0.08***|141.65***|-0.51***|-9.86***|-0.25***|-1.06|-0.23***|-2.80|\\n|Urban areas|||||||||\\n|Total|0.04***|176.12***|-0.39***|0.01|-0.23***|0.21|-0.16***|3.36|\\n|Women|0.02|153.99***|-0.32***|-1.33|-0.24***|0.56|-0.14***|0.80|\\n|Men|0.05**|194.34***|-0.46***|0.73|-0.25***|-1.00|-0.19***|5.06|\\n\\n<u>Notes:</u> (1) OLS regressions controlling for EA and day of the week Fes; the standard errors are clustered at the household-level. ***p<0.01, **p<0.05, *p<0.10 (2) Based on a pooled individual-level sample for days 5, 8 and 10 (each individual appeared three times in the sample).\\n\\n(3) Regressions control for the same individual and household-level characteristics as in eq. (1).\\n\\n31\", \"datasets\": [], \"classifier_skipped\": true, \"skip_reason\": \"no_data\", \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [32]}}, {\"page\": 33, \"chunk\": 33, \"input_text\": \"# 6. Discussion\\n\\nThis study reports on a large-scale survey experiment that was implemented for the first time in a lowincome country to provide a comparative assessment of real-time, smartphone-based self-administered time use diaries versus recall-based, interviewer-assisted approaches to time use measurement. The experimental study design, repeated measurements of smartphone- and recall-based data, and the scope of the smartphone app functionality unlock several empirical insights regarding the practical field experience with smartphone-based data collection and the quality of reporting across smartphone and recall-based approaches. The analysis shows that compared to real-time time use data collection using a pictorial smartphone app, the reliance on a recall-based 24-hour diary leads to underestimation of participation in employment and unpaid work; mischaracterizes intra-day variation in time use with policyrelevant implications; fails to capture the true extent of multitasking, particularly among women; and yet, results in overestimation of time devoted to a wide range of activities. The latter is <mark>due to the minimum 15 minute-duration that was used for recording activities the 24-hour recall diary – while over one-third of activities lasted less than 15 minutes in the smartphone arm. The design of the 24-hour recall diary also results in “rounding” of estimated time, which is not a concern in the use of the open-ended smartphone time use diary. The analysis also demonstrates that compared to the smartphone time use diary, the reliance on stylized time use questions with a 7-day recall (as opposed to a 24-hour recall diary) results in an even greater overestimation of time devoted to employment and unpaid work. Ultimately, recall-based time use measurement methods provide an incomplete and misleading representation of men’s and women’s time use outcomes in this study population.</mark> Looking forward, survey practitioners can consider using the existing _TimeTracker_ app and training materials for deployment in alternative contexts. The app illustrations and translations can be updated with ease, and the fixed costs of smartphone procurement can be managed by procuring a limited pool of smartphones that can redeployed throughout the fieldwork. The app pop-ups can also be tailored to collect additional policy-relevant contextual information regarding reported activities, including passive childcare provision, as demonstrated in this study. At the same time, there are key issues that need to be addressed prior to proposing concrete recommendations for scaling up real-time time use data collection in low- and middle-income countries. First, it would be essential to replicate our study in alternative contexts, including with additional recall variants that have emerged in recent literature (Field et al., 2023; McCullough et al., 2023), to verify the external validity of our findings and strengthen the evidence base regarding the utility and requirements of improved approaches to time use measurement. Second, future research should provide guidance on whether the duration of smartphone data collection can be reduced, for example from nine to three full days of reporting, to save on costs without losing the analytical utility of the resulting data. Finally, survey practitioners could benefit from guidance on whether real-time, smartphone-based data collection can be limited to a sub-sample with the idea of relying on model-based within-survey imputation of time use measures – again as a cost saving mechanism while improving the quality of time use data collected in large-scale surveys.\\n\\n32\", \"datasets\": [{\"mention_name\": {\"text\": \"real-time time use data\", \"confidence\": 0.37689197063446045, \"start\": 2620, \"end\": 2643, \"clean_text\": \"real-time time use data\"}, \"acronym\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2620, \"end\": 2643, \"clean_text\": \"\"}, \"producer\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2620, \"end\": 2643, \"clean_text\": \"\"}, \"reference_year\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2620, \"end\": 2643}, \"geography\": {\"text\": \"\", \"confidence\": 0.0, \"start\": 2620, \"end\": 2643, \"clean_text\": \"\"}, \"typology_tag\": {\"text\": \"microdata\", \"confidence\": 0.3807242810726166, \"start\": 755, \"end\": 778}, \"specificity_tag\": {\"text\": \"vague\", \"confidence\": 0.37689197063446045, \"start\": 755, \"end\": 778}, \"usage_context\": {\"text\": \"primary\", \"confidence\": 0.5213189125061035, \"start\": 755, \"end\": 778}, \"is_used\": {\"text\": \"True\", \"confidence\": 0.5213189125061035, \"start\": 755, \"end\": 778}}], \"classifier_skipped\": false, \"skip_reason\": null, \"document\": {\"source\": \"http://documents.worldbank.org/curated/en/099814102072416574/pdf/IDU195d006a6196d314cc21a96310b5efaaff3cd.pdf\", \"pages\": [33]}}]}"}}}],"truncated":false,"partial":false},"paginationData":{"pageIndex":0,"numItemsPerPage":100,"numTotalItems":8569,"offset":0,"length":100}},"jwt":"eyJhbGciOiJFZERTQSIsImtpZCI6IjVHZDBvd0g5MTM2eDZjc1FvbE1zcktNWUZoRVFoUm5PVVVybEpjOUhaUEEifQ.eyJyZWFkIjp0cnVlLCJwZXJtaXNzaW9ucyI6eyJyZXBvLmNvbnRlbnQucmVhZCI6dHJ1ZX0sImlhdCI6MTc4ODYxNzAwOCwianRpIjoiZjUxZjA3MGQtYmEyYy00NmQzLWEwZDYtYWYyNDY2NDQzZjZlIiwic3ViIjoiL2RhdGFzZXRzL3JhZm1hY2FsYWJhL2RhdGF1c2VfcmF3X2V4dHJhY3Rpb25zIiwiZXhwIjoxNzg4NjIwNjA4LCJpc3MiOiJodHRwczovL2h1Z2dpbmdmYWNlLmNvIn0.ZsOwiRs-DIHaVsi8daqhjDeogurBrIjWXkQX9EGniEqkdWqncI48unyn62Lpmkw6lefgLFx2SISRyXCSALyiAQ","displayUrls":true,"splitSizeSummaries":[{"config":"default","split":"train","numRows":8569,"numBytesParquet":419674544}]},"discussionsStats":{"closed":0,"open":1,"total":1},"fullWidth":true,"hasGatedAccess":true,"hasFullAccess":true,"isEmbedded":false,"savedQueries":{"community":[],"user":[]}}">
doc_id stringlengths 6 8 | doc stringlengths 5.92k 14.2M |
|---|---|
34175989 | {"metadata": {"id": "34175989", "last_modified_date": "2023-11-15T00:00:00Z", "authors": {"0": {"author": "Eslava,Marcela"}, "1": {"author": "Melendez Arjona,Marcela"}, "2": {"author": "Tenjo Galvis,Laura Daniela"}, "3": {"author": "Urdaneta,Nicolás"}}, "count": "Latin America", "count_key": "517204", "docna": {"0": {"... |
34176024 | {"metadata": {"id": "34176024", "last_modified_date": "2023-10-06T00:00:00Z", "admreg": "East Asia and Pacific,East Asia and Pacific", "admreg_key": "119225,119225", "authors": {"0": {"author": "Hapsari,Indira Maulani"}, "1": {"author": "Shu Yu"}, "2": {"author": "Pape,Utz Johann"}, "3": {"author": "Mansour,Wael"}}, "c... |
34178367 | {"metadata": {"id": "34178367", "last_modified_date": "2024-12-05T00:00:00Z", "admreg": "Eastern and Southern Africa,Eastern and Southern Africa", "admreg_key": "1927319,1927319", "authors": {"0": {"author": "Keyser,John C."}, "1": {"author": "Kunaka,Charles"}, "2": {"author": "Walkenhorst,Peter"}}, "count": "Eastern a... |
34179149 | {"metadata": {"id": "34179149", "last_modified_date": "2023-11-16T00:00:00Z", "authors": {"0": {"author": "Yarrow,Noah Bunce"}, "1": {"author": "Cahu,Paul Marie Michel"}, "2": {"author": "Breeding,Mary E."}, "3": {"author": "Afkar,Rythia"}}, "count": "East Asia and Pacific", "count_key": "517199", "docna": {"0": {"docn... |
34186262 | {"metadata": {"id": "34186262", "last_modified_date": "2023-10-26T00:00:00Z", "authors": {"0": {"author": "Orecchia,Carlo"}, "1": {"author": "Cala,Valerio Ferdinando"}, "2": {"author": "de Cristofaroa,Fabiana"}, "3": {"author": "Dudu,Hasan"}}, "count": "Italy", "count_key": "82554", "docna": {"0": {"docna": "Assessing ... |
34188936 | {"metadata": {"id": "34188936", "last_modified_date": "2023-10-31T00:00:00Z", "admreg": "East Asia and Pacific,East Asia and Pacific", "admreg_key": "119225,119225", "authors": {"0": {"author": "Freund,Caroline"}, "1": {"author": "Mattoo,Aaditya"}, "2": {"author": "Mulabdic,Alen"}, "3": {"author": "Ruta,Michele"}}, "co... |
34192225 | {"metadata": {"id": "34192225", "last_modified_date": "2023-11-07T00:00:00Z", "authors": {"0": {"author": "Roche Rodriguez,Jaime Alfonso"}, "1": {"author": "Lopez-Acevedo,Gladys C."}, "2": {"author": "Robertson,Raymond"}, "3": {"author": "Zárate,Daniela Ruiz"}}, "count": "Morocco", "count_key": "82545", "docna": {"0": ... |
34192232 | {"metadata": {"id": "34192232", "last_modified_date": "2024-01-09T00:00:00Z", "admreg": "South Asia,South Asia", "admreg_key": "119231,119231", "authors": {"0": {"author": "Bedoya Arguelles,Guadalupe"}, "1": {"author": "Belyakova,Yulia"}, "2": {"author": "Coville,Aidan"}, "3": {"author": "Escande,Thomas"}, "4": {"autho... |
34192813 | {"metadata": {"id": "34192813", "last_modified_date": "2023-11-07T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Ashwin,Julian"}, "1": {"author": "Chhabra,Aditya"}, "2": {"author": "Rao,Vijayendra"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "Us... |
34192820 | {"metadata": {"id": "34192820", "last_modified_date": "2023-11-15T00:00:00Z", "authors": {"0": {"author": "Dahis,Ricardo"}, "1": {"author": "Ricca,Bernardo"}, "2": {"author": "De Gouvea Scot De Arruda,Thiago"}, "3": {"author": "Machado Sales,Nathalia"}, "4": {"author": "Nascimento Moreira,Lucas"}}, "count": "N/A,Brazil... |
34193411 | {"metadata": {"id": "34193411", "last_modified_date": "2023-11-17T00:00:00Z", "admreg": "Middle East and North Africa,Middle East and North Africa", "admreg_key": "119229,119229", "authors": {"0": {"author": "Lofgren,Hans"}, "1": {"author": "Cicowiez,Martín"}, "2": {"author": "Mele,Gianluca"}}, "count": "Yemen,Republic... |
34193594 | {"metadata": {"id": "34193594", "last_modified_date": "2023-11-08T00:00:00Z", "authors": {"0": {"author": "Robinson,James Alan"}, "1": {"author": "Vostroknutov,Alexander"}, "2": {"author": "Vostroknutova,Ekaterina"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "Endogenous Institutions and Economi... |
34194176 | {"metadata": {"id": "34194176", "last_modified_date": "2023-11-09T00:00:00Z", "authors": {"0": {"author": "Zavala,Lucas"}, "1": {"author": "Fernandes,Ana Margarida"}, "2": {"author": "Haygood,Ryan"}, "3": {"author": "Reed,Tristan"}, "4": {"author": "Reyes,Jose Daniel"}}, "count": "World", "count_key": "517191", "docna"... |
34194257 | {"metadata": {"id": "34194257", "last_modified_date": "2023-11-09T00:00:00Z", "authors": {"0": {"author": "Gill,Indermit S."}, "1": {"author": "Pinto,Brian R."}}, "count": "Ethiopia,Ghana,Zambia", "count_key": "82611,82621,82701", "docna": {"0": {"docna": "Making the Low-Income Country Debt\n Sustainability ... |
34195858 | {"metadata": {"id": "34195858", "last_modified_date": "2023-11-27T00:00:00Z", "authors": {"0": {"author": "Cull,Robert J."}, "1": {"author": "Foster,Vivien"}, "2": {"author": "Jolliffe,Dean Mitchell"}, "3": {"author": "Lederman,Daniel"}, "4": {"author": "Mare,Davide Salvatore"}, "5": {"author": "Veerappan,Malarvizhi"}}... |
34196317 | {"metadata": {"id": "34196317", "last_modified_date": "2024-04-18T00:00:00Z", "authors": {"0": {"author": "Davalos,Maria Eugenia"}, "1": {"author": "Londono Aguirre,Diana Isabel"}, "2": {"author": "Medina Gaspar,Daniel Santiago"}}, "count": "Colombia", "count_key": "82629", "docna": {"0": {"docna": "How Have Gender Gap... |
34196964 | {"metadata": {"id": "34196964", "last_modified_date": "2023-11-15T00:00:00Z", "authors": {"0": {"author": "Bloem,Jeffrey Richard"}, "1": {"author": "Damon,Amy"}, "2": {"author": "Francis,David C."}, "3": {"author": "Mitchell,Harrison"}}, "count": "Nigeria", "count_key": "82531", "docna": {"0": {"docna": "Herder-Related... |
34196967 | {"metadata": {"id": "34196967", "last_modified_date": "2023-11-15T00:00:00Z", "admreg": "East Asia and Pacific,East Asia and Pacific", "admreg_key": "119225,119225", "authors": {"0": {"author": "Ablaza,Christine Marie Jimenez"}, "1": {"author": "Alladi,Vinayak"}, "2": {"author": "Pape,Utz Johann"}}, "count": "Indonesia... |
34198137 | {"metadata": {"id": "34198137", "last_modified_date": "2023-11-16T00:00:00Z", "authors": {"0": {"author": "Fontagné,Lionel"}, "1": {"author": "Lebrand,Mathilde Sylvie Maria"}, "2": {"author": "Murray,Siobhan"}, "3": {"author": "Ruta,Michele"}, "4": {"author": "Santoni,Gianluca"}}, "count": "Africa", "count_key": "51719... |
34199941 | {"metadata": {"id": "34199941", "last_modified_date": "2023-11-20T00:00:00Z", "admreg": "East Asia and Pacific,East Asia and Pacific", "admreg_key": "119225,119225", "authors": {"0": {"author": "Chattha,Muhammad Khudadad"}, "1": {"author": "Tyas,Prabaning"}, "2": {"author": "Soko,Naranggi Pramudya"}, "3": {"author": "F... |
34200623 | {"metadata": {"id": "34200623", "last_modified_date": "2023-11-21T00:00:00Z", "authors": {"0": {"author": "Arroyo Marioli,Francisco"}, "1": {"author": "Khadan,Jeetendra"}, "2": {"author": "Ohnsorge,Franziska Lieselotte"}, "3": {"author": "Yamazaki,Takefumi"}}, "count": "World", "count_key": "517191", "docna": {"0": {"d... |
34200651 | {"metadata": {"id": "34200651", "last_modified_date": "2024-01-10T00:00:00Z", "authors": {"0": {"author": "Bossavie,Laurent Loic Yves"}, "1": {"author": "Rozo Villarraga,Sandra Viviana"}, "2": {"author": "Urbina Florez,Maria Jose"}}, "count": "Afghanistan,Tajikistan", "count_key": "82662,82573", "docna": {"0": {"docna"... |
34200675 | {"metadata": {"id": "34200675", "last_modified_date": "2023-11-21T00:00:00Z", "admreg": "Africa,Africa", "admreg_key": "119222,119222", "authors": {"0": {"author": "Carletto,Calogero"}, "1": {"author": "Letta,Marco"}, "2": {"author": "Montalbano,Pierluigi"}, "3": {"author": "Paolantonio,Adriana"}, "4": {"author": "Zezz... |
34200726 | {"metadata": {"id": "34200726", "last_modified_date": "2024-01-25T00:00:00Z", "authors": {"0": {"author": "Tillan,Pablo Antonio"}, "1": {"author": "Kokas,Deeksha"}, "2": {"author": "Bezerra De Goes,Carlos Andre"}, "3": {"author": "Lopez-Acevedo,Gladys C."}}, "count": "Egypt, Arab Republic of", "count_key": "82608", "do... |
34201505 | {"metadata": {"id": "34201505", "last_modified_date": "2024-03-22T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Hoy,Christopher Alexander"}, "1": {"author": "Kim,Yeon Soo"}, "2": {"author": "Nguyen,Minh Cong"}, "3": {"author": "Sosa,Mariano Ernesto"}, "4": {"author": "... |
34203039 | {"metadata": {"id": "34203039", "last_modified_date": "2023-12-04T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Combes,Pierre Philippe"}, "1": {"author": "Gorin,Clément"}, "2": {"author": "Nakamura,Shohei"}, "3": {"author": "Roberts,Mark"}, "4": {"author": "Stewart,Ben... |
34203046 | {"metadata": {"id": "34203046", "last_modified_date": "2023-11-27T00:00:00Z", "authors": {"0": {"author": "Barriga Cabanillas,Oscar Eduardo"}, "1": {"author": "Chawla,Parth"}, "2": {"author": "Redaelli,Silvia"}, "3": {"author": "Yoshida,Nobuo"}}, "count": "Afghanistan", "count_key": "82662", "docna": {"0": {"docna": "U... |
34203219 | {"metadata": {"id": "34203219", "last_modified_date": "2023-11-27T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Nakamura,Shohei"}, "1": {"author": "Combes,Pierre Philippe"}, "2": {"author": "Moellerherm,Robin"}, "3": {"author": "Robert,Charlotte Genevieve Julka"}, "4":... |
34203226 | {"metadata": {"id": "34203226", "last_modified_date": "2023-11-27T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "World Bank"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "Is Natural Capital a Complement to Human\n Capital ? Evidence fr... |
34203250 | {"metadata": {"id": "34203250", "last_modified_date": "2023-11-27T00:00:00Z", "admreg": "Middle East and North Africa,Middle East and North Africa", "admreg_key": "119229,119229", "authors": {"0": {"author": "Ishak,Phoebe Saad Wasfy"}, "1": {"author": "Aghajanian,Alia Jane"}, "2": {"author": "Ghorpade,Yashodhan"}}, "co... |
34204435 | {"metadata": {"id": "34204435", "last_modified_date": "2023-11-29T00:00:00Z", "authors": {"0": {"author": "Doan,Miki Khanh"}, "1": {"author": "Hill,Ruth"}, "2": {"author": "Hallegatte,Stephane"}, "3": {"author": "Corral Rodas,Paul Andres"}, "4": {"author": "Brunckhorst,Ben James"}, "5": {"author": "Nguyen,Minh"}, "6": ... |
34205862 | {"metadata": {"id": "34205862", "last_modified_date": "2024-05-13T00:00:00Z", "authors": {"0": {"author": "Cust,James Frederick"}, "1": {"author": "Collier,Paul"}, "2": {"author": "Rivera-Ballesteros, lexis"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "Are the Poorest Catching Up ?"}}, "docty":... |
34206486 | {"metadata": {"id": "34206486", "last_modified_date": "2023-12-01T00:00:00Z", "authors": {"0": {"author": "Lebrand,Mathilde Sylvie Maria"}, "1": {"author": "Vasishtha,Garima"}, "2": {"author": "Yilmazkuday,Hakan"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "Energy Price Shocks and Current Accou... |
34208238 | {"metadata": {"id": "34208238", "last_modified_date": "2024-01-29T00:00:00Z", "authors": {"0": {"author": "Mandon,Pierre Jean-Claude"}, "1": {"author": "Nossek,Vincent Remi"}, "2": {"author": "Sandjong Tomi,Diderot Guy D Estaing"}}, "count": "Central Africa", "count_key": "517196", "docna": {"0": {"docna": "Stuck in a ... |
34208252 | {"metadata": {"id": "34208252", "last_modified_date": "2023-12-04T00:00:00Z", "authors": {"0": {"author": "Miller,Amisha"}, "1": {"author": "Lall,Saurabh A."}, "2": {"author": "Goldstein,Markus P."}, "3": {"author": "Montalvao Machado,Joao H. C."}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "Ask... |
34208542 | {"metadata": {"id": "34208542", "last_modified_date": "2023-12-05T00:00:00Z", "authors": {"0": {"author": "Bertoli,Simone"}, "1": {"author": "Mckenzie,David J."}, "2": {"author": "Murard,Elie"}}, "count": "World,United States", "count_key": "517191,82584", "docna": {"0": {"docna": "Migration, families, and counterfactu... |
34208588 | {"metadata": {"id": "34208588", "last_modified_date": "2023-12-05T00:00:00Z", "authors": {"0": {"author": "Matekenya,Dunstan"}, "1": {"author": "Mulangu,Francis Muamba"}, "2": {"author": "Newhouse,David Locke"}}, "count": "Madagascar", "count_key": "82548", "docna": {"0": {"docna": "Malnourished but Not Destitute : The... |
34208602 | {"metadata": {"id": "34208602", "last_modified_date": "2023-12-05T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Mckenzie,David J."}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "Is There Still A Role for Direct Government\n Support to ... |
34209286 | {"metadata": {"id": "34209286", "last_modified_date": "2023-12-06T00:00:00Z", "admreg": "East Asia and Pacific,East Asia and Pacific", "admreg_key": "119225,119225", "authors": {"0": {"author": "Ghorpade,Yashodhan"}, "1": {"author": "Abdur Rahman,Amanina Binti"}, "2": {"author": "Binti Jasmin,Alyssa Farha"}}, "count": ... |
34209351 | {"metadata": {"id": "34209351", "last_modified_date": "2023-12-06T00:00:00Z", "admreg": "East Asia and Pacific,East Asia and Pacific", "admreg_key": "119225,119225", "authors": {"0": {"author": "Ghorpade,Yashodhan"}, "1": {"author": "Binti Jasmin,Alyssa Farha"}, "2": {"author": "Abdur Rahman,Amanina Binti"}}, "count": ... |
34209391 | {"metadata": {"id": "34209391", "last_modified_date": "2023-12-07T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Masaki,Takaaki"}, "1": {"author": "Madson,B."}}, "count": "Africa,World", "count_key": "517192,517191", "docna": {"0": {"docna": "Data Gaps in Microdata in t... |
34210169 | {"metadata": {"id": "34210169", "last_modified_date": "2024-03-28T00:00:00Z", "authors": {"0": {"author": "Beck,Thorsten"}, "1": {"author": "Cull,Robert J."}, "2": {"author": "Mare,Davide Salvatore"}, "3": {"author": "Valenzuela,Patricio"}}, "count": "Africa", "count_key": "517192", "docna": {"0": {"docna": "Banking in... |
34210179 | {"metadata": {"id": "34210179", "last_modified_date": "2024-04-18T00:00:00Z", "authors": {"0": {"author": "Vargas Da Cruz,Marcio Jose"}, "1": {"author": "Pereira Lopez,Mariana De La Paz"}, "2": {"author": "Salgado Chavez,Edgar"}}, "count": "Africa", "count_key": "517192", "docna": {"0": {"docna": "Disruptive Technologi... |
34212577 | {"metadata": {"id": "34212577", "last_modified_date": "2023-12-11T00:00:00Z", "authors": {"0": {"author": "Kee,Hiau Looi"}, "1": {"author": "Nicita,Alessandro"}}, "count": "World,European Union,United Kingdom", "count_key": "517191,82617", "docna": {"0": {"docna": "Quantifying Economic Impacts of Trade\n Agr... |
34212650 | {"metadata": {"id": "34212650", "last_modified_date": "2023-12-11T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Winkler,Deborah Elisabeth"}, "1": {"author": "Kruse,Hagen"}, "2": {"author": "Aguilar Luna,Luis Alejandro"}, "3": {"author": "Maliszewska,Maryla"}}, "count":... |
34212676 | {"metadata": {"id": "34212676", "last_modified_date": "2023-12-13T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Johansson De Silva,Sara"}, "1": {"author": "Santos,Indhira Vanessa"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "Productive Longevit... |
34212706 | {"metadata": {"id": "34212706", "last_modified_date": "2023-12-11T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Shepherd,Ben"}, "1": {"author": "Winkler,Deborah Elisabeth"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "Leveraging Trade for More a... |
34212724 | {"metadata": {"id": "34212724", "last_modified_date": "2023-12-12T00:00:00Z", "admreg": "Europe and Central Asia,Europe and Central Asia", "admreg_key": "119226,119226", "authors": {"0": {"author": "Robayo,Monica"}, "1": {"author": "Rude,Britta Laurin"}}, "count": "Romania", "count_key": "82513", "docna": {"0": {"docna... |
34212730 | {"metadata": {"id": "34212730", "last_modified_date": "2023-12-11T00:00:00Z", "authors": {"0": {"author": "Londoño-Vélez,Juliana"}, "1": {"author": "Tortarolo,Dario"}}, "count": "Argentina", "count_key": "82668", "docna": {"0": {"docna": "Revealing 21% of GDP in Hidden Assets :\n Evidence from Argentina"}}, ... |
34213264 | {"metadata": {"id": "34213264", "last_modified_date": "2023-12-12T00:00:00Z", "admreg": "East Asia and Pacific,East Asia and Pacific", "admreg_key": "119225,119225", "authors": {"0": {"author": "Bosker,Maarten"}, "1": {"author": "Roberts,Mark"}, "2": {"author": "Tiwari,Sailesh"}, "3": {"author": "Wibisana,Putu Sanjiwac... |
34213298 | {"metadata": {"id": "34213298", "last_modified_date": "2023-12-12T00:00:00Z", "admreg": "Europe and Central Asia,Europe and Central Asia", "admreg_key": "119226,119226", "authors": {"0": {"author": "Bruhn,Miriam"}, "1": {"author": "Demirguc-Kunt,Asli"}, "2": {"author": "Singer,Dorothe"}}, "count": "World,Europe and Cen... |
34213411 | {"metadata": {"id": "34213411", "last_modified_date": "2025-01-10T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Ghose,Devaki"}, "1": {"author": "Pinheiro Fraga,Eduardo"}, "2": {"author": "Fernandes,Ana Margarida"}}, "count": "Sri Lanka,World", "count_key": "82538,51719... |
34214449 | {"metadata": {"id": "34214449", "last_modified_date": "2023-12-13T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Goicoechea,Ana"}, "1": {"author": "Lang,Megan Elizabeth"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "Firms and Climate Change in Lo... |
34214489 | {"metadata": {"id": "34214489", "last_modified_date": "2024-04-05T00:00:00Z", "authors": {"0": {"author": "Nana,Ibrahim"}, "1": {"author": "Motelle,Sephooko Ignatius"}, "2": {"author": "Starnes,Susan K."}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "The Trade-Growth Nexus: Evidence of\n ... |
34215230 | {"metadata": {"id": "34215230", "last_modified_date": "2024-01-10T00:00:00Z", "authors": {"0": {"author": "El-Haddad,Amirah"}, "1": {"author": "Ishak,Phoebe Saad Wasfy"}}, "count": "Egypt, Arab Republic of", "count_key": "82608", "docna": {"0": {"docna": "The Effect of COVID-19 on the Gender\n Employment Gap... |
34216042 | {"metadata": {"id": "34216042", "last_modified_date": "2024-01-29T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "He He"}, "1": {"author": "Kim,Chaeyoung"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "Vehicle and Fuel Taxation for Transport\n ... |
34218168 | {"metadata": {"id": "34218168", "last_modified_date": "2023-12-18T00:00:00Z", "authors": {"0": {"author": "Ha,Jongrim"}, "1": {"author": "Kose,Ayhan"}, "2": {"author": "Ohnsorge,Franziska Lieselotte"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "What Explains Global Inflation"}}, "docty": "Polic... |
34218204 | {"metadata": {"id": "34218204", "last_modified_date": "2023-12-18T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Dauda,Seidu"}, "1": {"author": "Pop,Georgiana"}, "2": {"author": "Iootty De Paiva Dias,Mariana"}}, "count": "Romania,World", "count_key": "82513,517191", "do... |
34218347 | {"metadata": {"id": "34218347", "last_modified_date": "2023-12-18T00:00:00Z", "authors": {"0": {"author": "Pérez-Sebastián,Fidel"}, "1": {"author": "Serrano Quintero,Rafael"}, "2": {"author": "Steinbuks,Jevgenijs"}}, "count": "Brazil", "count_key": "82649", "docna": {"0": {"docna": "Spatial Misallocation of Complementa... |
34218381 | {"metadata": {"id": "34218381", "last_modified_date": "2023-12-18T00:00:00Z", "admreg": "Eastern and Southern Africa,Eastern and Southern Africa", "admreg_key": "1927319,1927319", "authors": {"0": {"author": "Bedi,Tara Sylvia"}, "1": {"author": "Buehren,Niklas"}, "2": {"author": "Goldstein,Markus P."}, "3": {"author": ... |
34219083 | {"metadata": {"id": "34219083", "last_modified_date": "2024-08-12T00:00:00Z", "authors": {"0": {"author": "Iregui-Bohórquez,Ana María"}, "1": {"author": "Melo-Becerra,Ligia Alba"}, "2": {"author": "Ramírez-Giraldo,María Teresa"}, "3": {"author": "Tribin Uribe,Ana Maria"}, "4": {"author": "Zárate-Solano,Héctor M."}}, "c... |
34220483 | {"metadata": {"id": "34220483", "last_modified_date": "2023-12-20T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "S Anukriti"}, "1": {"author": "Calvi,Rossella"}, "2": {"author": "Chakravarty,Abhishek"}}, "count": "India,World", "count_key": "82602,517191", "docna": {"0"... |
34220589 | {"metadata": {"id": "34220589", "last_modified_date": "2024-01-29T00:00:00Z", "authors": {"0": {"author": "Okunogbe,Oyebola Motunrayo"}, "1": {"author": "Tourek,Gabriel Zenon"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "How Can Lower-Income Countries Collect More\n Taxes ? The Role ... |
34220656 | {"metadata": {"id": "34220656", "last_modified_date": "2023-12-20T00:00:00Z", "authors": {"0": {"author": "Zhang,Kexin"}, "1": {"author": "Takamatsu,Shinya"}, "2": {"author": "Yoshida,Nobuo"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "Correcting Sampling and Nonresponse Bias in\n Ph... |
34234558 | {"metadata": {"id": "34234558", "last_modified_date": "2024-02-12T00:00:00Z", "admreg": "Europe and Central Asia,Europe and Central Asia", "admreg_key": "119226,119226", "authors": {"0": {"author": "Robayo,Monica"}, "1": {"author": "Cabrera,Maynor Vinicio"}}, "count": "Bulgaria", "count_key": "82642", "docna": {"0": {"... |
34234665 | {"metadata": {"id": "34234665", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "Fiuratti,Federico Ivan"}, "1": {"author": "Nikolova,Desislava Enikova"}, "2": {"author": "Pennings,Steven Michael"}, "3": {"author": "Schiffbauer,Marc Tobias"}}, "count": "European Union", "docna": {"0": {"docna":... |
34234742 | {"metadata": {"id": "34234742", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "Assaad,Ragui A"}, "1": {"author": "Krafft,Caroline Gould"}}, "count": "Middle East and North Africa", "count_key": "517203", "docna": {"0": {"docna": "Connecting People to Projects : A New\n Approach to... |
34234814 | {"metadata": {"id": "34234814", "last_modified_date": "2024-02-06T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Wollburg,Philip Randolph"}, "1": {"author": "Markhof,Yannick Valentin"}, "2": {"author": "Bentze,Thomas Patrick"}, "3": {"author": "Ponzini,Giulia"}}, "count... |
34234826 | {"metadata": {"id": "34234826", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "El Mekkaoui,Najat"}, "1": {"author": "Loukili,Sara"}, "2": {"author": "Fourouheshfar,Yeganeh"}, "3": {"author": "Eissa,Nada Omer"}}, "count": "Morocco,Egypt, Arab Republic of,Jordan", "count_key": "82545,82608,825... |
34234978 | {"metadata": {"id": "34234978", "last_modified_date": "2024-02-06T00:00:00Z", "admreg": "Latin America & Caribbean,Latin America & Caribbean", "admreg_key": "119228,119228", "authors": {"0": {"author": "Stokenberga,Aiga"}, "1": {"author": "Ivarsson Molina,Linda Ellin Maria"}, "2": {"author": "Fulponi,Juan Ignacio"}, "3... |
34235628 | {"metadata": {"id": "34235628", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "Mohieldin,Mahmoud"}, "1": {"author": "Ramadan,Racha"}}, "count": "Middle East and North Africa", "count_key": "517203", "docna": {"0": {"docna": "Could Digital Inclusion Close the Gender\n Economic Gap ... |
34235906 | {"metadata": {"id": "34235906", "last_modified_date": "2024-08-14T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Amankwah,Akuffo"}, "1": {"author": "Castaing,Pauline"}, "2": {"author": "Owoo,Nkechi Srodah"}, "3": {"author": "Palacios-Lopez,Amparo"}}, "count": "Ghana,Wor... |
34236962 | {"metadata": {"id": "34236962", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "Jakubowski,Maciej Jan"}, "1": {"author": "Gajderowicz,Tomasz Janusz"}, "2": {"author": "Patrinos,Harry Anthony"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "COVID-19, School Closures, and ... |
34237443 | {"metadata": {"id": "34237443", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "Kalliny,Marize"}, "1": {"author": "Zaki,Chahir"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "Are Global Value Chains Women Friendly in\n Developing Countries ? Evidence from Fir... |
34237636 | {"metadata": {"id": "34237636", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "Floreani,Vincent Arthur"}, "1": {"author": "Rama,Martin G."}}, "count": "Nepal", "count_key": "82535", "docna": {"0": {"docna": "Costing Disasters : Hedonic Pricing,\n Neighborhood Effects, and the Nepa... |
34239665 | {"metadata": {"id": "34239665", "last_modified_date": "2024-02-06T00:00:00Z", "admreg": "East Asia and Pacific,East Asia and Pacific", "admreg_key": "119225,119225", "authors": {"0": {"author": "Cho,Yoonyoung"}, "1": {"author": "Molina,Teresa Sacro"}}, "count": "Philippines", "count_key": "82506", "docna": {"0": {"docn... |
34239706 | {"metadata": {"id": "34239706", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "Wahby,Sarah Mohammad Osman"}, "1": {"author": "Assaad,Ragui A"}}, "count": "Syrian Arab Republic,Jordan", "count_key": "82691,82556", "docna": {"0": {"docna": "Job Finding and Separation among Syrian\n ... |
34239717 | {"metadata": {"id": "34239717", "last_modified_date": "2024-02-06T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Horton,Susan E."}, "1": {"author": "Fracassi,Patrizia"}, "2": {"author": "Siekmans,Kendra"}, "3": {"author": "Kato,Tomoko"}, "4": {"author": "Hanley-Cook,Gil... |
34239758 | {"metadata": {"id": "34239758", "last_modified_date": "2024-02-06T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "AlAzzawi,Shireen"}, "1": {"author": "Dang,Hai-Anh H."}, "2": {"author": "Hlasny,Vladimir"}, "3": {"author": "Abanokova,Ksenia"}, "4": {"author": "Behrman,Jer... |
34240433 | {"metadata": {"id": "34240433", "last_modified_date": "2024-02-06T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Stacy,Brian William"}, "1": {"author": "Kitzmüller,Lucas"}, "2": {"author": "Wang,Xiaoyu"}, "3": {"author": "Mahler,Daniel Gerszon"}, "4": {"author": "Seraju... |
34241066 | {"metadata": {"id": "34241066", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "Najam,Rafiuddin"}}, "count": "Afghanistan", "count_key": "82662", "docna": {"0": {"docna": "Closing the gap : Effect of a gender quota\n on women's access to education in Afghanistan"}}, "docty": "Polic... |
34241163 | {"metadata": {"id": "34241163", "last_modified_date": "2024-03-08T00:00:00Z", "admreg": "Latin America & Caribbean,Latin America & Caribbean", "admreg_key": "119228,119228", "authors": {"0": {"author": "Fietz,Katharina Maria"}, "1": {"author": "Morgandi,Matteo"}, "2": {"author": "Lyrio De Oliveira,Gabriel"}, "3": {"aut... |
34241308 | {"metadata": {"id": "34241308", "last_modified_date": "2024-02-06T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Thomas,Alastair Geoffrey Arthur"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "VAT Rate Structures in Theory and Practice"}}, "docty"... |
34243167 | {"metadata": {"id": "34243167", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "Canavire Bacarreza,Gustavo Javier"}, "1": {"author": "Gómez Aliaga,Guillermo"}, "2": {"author": "Britton,Chevanne Regina Lygia"}, "3": {"author": "Rios Avila,Fernando"}, "4": {"author": "Jimenez Pozo,Wilson Leonar... |
34243399 | {"metadata": {"id": "34243399", "last_modified_date": "2024-02-06T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Atamanov,Aziz"}, "1": {"author": "Hoogeveen,Johannes G."}, "2": {"author": "Reese,Benjamin Christopher"}}, "count": "Uganda,World", "count_key": "82583,51719... |
34243842 | {"metadata": {"id": "34243842", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "Ambler,Kate"}, "1": {"author": "Herskowitz,Sylvan Rene"}, "2": {"author": "Maredia,Mywish Karimbhai"}}, "count": "Malawi", "count_key": "82524", "docna": {"0": {"docna": "Rural Labor and Long Recall Loss"}}, "doct... |
34243850 | {"metadata": {"id": "34243850", "last_modified_date": "2024-02-06T00:00:00Z", "admreg": "East Asia and Pacific,East Asia and Pacific", "admreg_key": "119225,119225", "authors": {"0": {"author": "Cameron,Lisa"}, "1": {"author": "Contreras Suarez,Diana Stella"}, "2": {"author": "Setyonaluri,Diahhadi"}}, "count": "Indones... |
34244968 | {"metadata": {"id": "34244968", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "Feriga,Moustafa Amgad Moustafa Ahmed Moustafa"}, "1": {"author": "Lozano Gracia,Nancy"}, "2": {"author": "Serneels,Pieter Maria"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "The Impact of ... |
34247621 | {"metadata": {"id": "34247621", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "Gascoigne,Jon"}, "1": {"author": "Baquie,Sandra"}, "2": {"author": "Vinha,Katja Pauliina"}, "3": {"author": "Skoufias,Emmanuel"}, "4": {"author": "Calcutt,Evie Isabel Neall"}, "5": {"author": "Kshirsagar,Varun Sri... |
34247735 | {"metadata": {"id": "34247735", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "Islam,Asif Mohammed"}, "1": {"author": "Gatti,Roberta V."}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "Dysfunctional Family Management :\n Family-Managed Businesses and the Qual... |
34248405 | {"metadata": {"id": "34248405", "last_modified_date": "2024-02-09T00:00:00Z", "authors": {"0": {"author": "Cortina Toro,Magdalena"}, "1": {"author": "Jimenez,Juan Miguel"}, "2": {"author": "Rozo Villarraga,Sandra Viviana"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "Little Nomads : Economic and... |
34248497 | {"metadata": {"id": "34248497", "last_modified_date": "2024-02-06T00:00:00Z", "admreg": "Middle East and North Africa,Middle East and North Africa", "admreg_key": "119229,119229", "authors": {"0": {"author": "Amjad,Beenish"}, "1": {"author": "Carrasco Nunez,Haydeeliz"}, "2": {"author": "Finn,Arden Jeremy"}, "3": {"auth... |
34249073 | {"metadata": {"id": "34249073", "last_modified_date": "2024-04-14T00:00:00Z", "authors": {"0": {"author": "Bou Habib,Chadi"}}, "count": "Lebanon", "count_key": "82571", "docna": {"0": {"docna": "Globalization, Dutch Disease, and\n Vulnerability to External Shocks in a Small Open Economy :\n The Ca... |
34249090 | {"metadata": {"id": "34249090", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "Behrer,Arnold Patrick"}, "1": {"author": "Heft-Neal,Sam"}}, "count": "World", "count_key": "517191", "docna": {"0": {"docna": "In Most Low- and Middle-Income Countries\n Pollution Levels Are Higher in W... |
34249700 | {"metadata": {"id": "34249700", "last_modified_date": "2024-02-06T00:00:00Z", "admreg": "Africa,Africa", "admreg_key": "119222,119222", "authors": {"0": {"author": "Wu,Haoyu"}, "1": {"author": "Atamanov,Aziz"}, "2": {"author": "Bundervoet,Tom"}, "3": {"author": "Paci,Pierella"}}, "count": "Africa", "count_key": "517192... |
34249732 | {"metadata": {"id": "34249732", "last_modified_date": "2024-07-16T00:00:00Z", "authors": {"0": {"author": "Xuan Hoang,Trung"}}, "count": "Iran, Islamic Republic of,Iraq", "count_key": "82604,82603", "docna": {"0": {"docna": "The Long-Term and Gender-Equalizing Impacts\n of the Islamic Republic of Iran-Iraq W... |
34249852 | {"metadata": {"id": "34249852", "last_modified_date": "2024-03-08T00:00:00Z", "authors": {"0": {"author": "Goldstein,Markus P."}, "1": {"author": "Gonzalez Martinez,Paula Lorena"}, "2": {"author": "Kilic,Talip"}, "3": {"author": "Papineni,Sreelakshmi"}, "4": {"author": "Wollburg,Philip Randolph"}}, "count": "World", "c... |
34252495 | {"metadata": {"id": "34252495", "last_modified_date": "2024-04-14T00:00:00Z", "authors": {"0": {"author": "Hauser,Christina Sarah"}}, "count": "Tunisia", "count_key": "82575", "docna": {"0": {"docna": "Tackling Gender Discriminatory Inheritance\n Law Privately : Lessons from a Survey Experiment in Tunisia"}}... |
34252504 | {"metadata": {"id": "34252504", "last_modified_date": "2024-02-06T00:00:00Z", "authors": {"0": {"author": "Barnett,Carolyn Louise"}}, "count": "Morocco", "count_key": "82545", "docna": {"0": {"docna": "Gender Role Attitudes, Perceived Norms, and\n the \"Double Burden'' in Morocco"}}, "docty": "Policy Researc... |
34254016 | {"metadata": {"id": "34254016", "last_modified_date": "2024-02-07T00:00:00Z", "admreg": "Other,Other", "admreg_key": "912972,912972", "authors": {"0": {"author": "Kilic,Talip"}, "1": {"author": "Koolwal,Gayatri B."}, "2": {"author": "Vundru,Wilbert Drazi"}, "3": {"author": "Daum,Thomas Lothar Georg"}, "4": {"author": "... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.