CH AD P OV ER T Y A S S E S S ME N T Investing in rural income growth, human capital, and resilience to support sustainable poverty reduction 2 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION © 2021 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 3 ACKNOWLEDGMENTS This report has been prepared by Aboudrahyme Savadogo (Economist, EAWPV) and Aly Sanoh (Sr. Economist, EAWPV). The core team also included Clarence Tsimpo Nkengne (Sr. Economist, EAWPV), Aissatou Ouedraogo (Economist, EAWPV), Awa Sanou, Essama Nssah, Mohammed Coulibaly, Sidi Mohamed Sawadogo, Arnaud Gotoraye, Sean Lothrop, and Oscar Parlback (Consultants, EAWPV). The team would like to thank constructive and detailed comments and suggestions from peer-reviewers Nadia Belhaj Hassine Belghith (Senior Economist, EEAPV), Nga Thi Viet Nguyen (Senior Economist, EECPV), Emmanuel Skoufias (Lead Economist, EAEPV), and Arthur Alik Lagrange (Senior Economist, EPVGE). The team also benefited from great comments and advices from Emmanuel Skoufias (Lead Economist, EA1PV), Theresa Osborne (Senior Economist, EA1PV), and Emilie Jourdan (Sr. Operations Officer, GTFOS) who peer-reviewed the Project Concept Note. The team would like to extend its sincere thanks to Nicolas Rosemberg (Economist, HAWH2), Zacharie Ngueng (Education Specialist, HAWE2), Patrick Hoang-Vu Eozenou (Senior Economist, HAWH2), Loredana Luisa Horezeanu (consultant, HAWH2), Fulbert Tchana Tchana (Senior Economist, EAWM1) for their useful feedback and support through the preparation of the report. The team gratefully acknowledges guidance from Johan A. Mistiaen (Practice Manager, EAWPV), Andrew Dabalen (Practice Manager ESAPV and former Practice Manager, EWWPV), Rasit Pertev (Country Manager, AWMTD), Francois Nankobogo (Manager AFWDE and former country Manager, AFMTD), Soukeyna Kane (Country Director, AWCW3), Kofi Nouve (Manager, operations, AWCW3), Abebe Adugna (Regional Director, EAWDR), Jean-Pierre Christophe Chauffour (Program Leader, EAWDR), and Christophe Rockmore (Program Leader, HAWDR). The team thanks Santosh Kumar Sahoo (Program Assistant, EAWPV), Arlette Sourou (Sr. Program Assistant, EAWPV), Senait Kassa Yifru (Operation Analyst, EAWPV), Fatime Mahamat Adoum (Executive Assistant, AWMTD), Nicolas Amadai (Program Assistant, AWMTD), and Kochikpa Abdou Raman Olodo (Operations Officer, AWMTD) for their valuable assistance during the preparation of the report. Finally, the team would like to thank Dr Baradine Zakaria Moursal, the General Director of the Institut National de la Statistique, des Etudes Economiques et Demographiques (INSEED), Ahmat Abderahim Abbo, the Director of Households survey of INSEED, and everyone at INSEED for the important support and critical feedback provided throughout the implementation of the survey and the preparation of the report. 4 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION TABLE OF CONTENTS Acknowledgments 3 Preface 10 Abbreviations and Acronyms 11 Glossary 12 Executive Summary 13 Chapter 1: Understanding Poverty in Chad 18 1.1 Poverty at the National and Subnational Levels 19 1.1.1 Rates and Trends 19 1.1.2 Gains in Poverty Reduction 22 1.1.3 Inequality and shared prosperity patterns 25 1.1.4 Drivers of Poverty reduction 29 1.2. Poverty Profile 31 1.2.1 Social and Demographic Characteristics of Poor Households 31 1.2.2 Food Security, Living Conditions and Asset Ownership 35 1.4 Poverty-Reduction Challenges 40 1.4.1 The Social Safety Net 40 1.4.2 Fragility, Conflict, and Violence 42 1.4.3 Forced Displacement 43 1.4.4 The Challenges Posed by the COVID-19 Pandemic 46 1.5 Conclusion 48 Chapter 2: Rural Income Analysis 50 2.1 Income and Productive Resources 55 2.1.1 Agriculture 55 2.1.2 Agricultural Market Participation 59 2.1.3 The Nonfarm Sector 60 2.1.4 Nonlabor Income: Remittances 63 2.2 Rural Income Growth: Opportunities and Constraints 64 2.2.1 Opportunities 64 2.2.2 Key Constraints on Rural Income Growth 66 2.3 Policy Options for Supporting Rural Income Growth 70 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 5 TABLE OF CONTENTS Chapter 3. Vulnerability to Shocks in Chad 73 3.1 Vulnerability to Poverty 75 3.2 Occurrence of Shocks 81 3.3 Coping Mechanisms 85 3.4 Preliminary Assessment of the Impact of the COVID-19 Outbreak 91 3.5 Conclusion and Policy Recommendations 94 Chapter 4: Human Capital in Chad 101 Achievements in Human Capital 103 4.1 Education 107 4.2 Health 116 4.3 Nutrition and Food Security 123 4.4 4.5 Education, Health, and Food Security during the Covid-19 Pandemic 126 Conclusion and Policy Recommendations 132 4.6 Data and Knowledge Gaps 134 References 136 Annex A: Technical note on poverty measurement based on ECOSIT 4 (2018/19) data 139 Annex B: Survey-to-Survey Imputation Methodology 151 Annex C: Multidimensional Poverty in Chad 155 Annex D: Drivers of Poverty Reduction 168 Annex E: Vulnerability and shocks 171 Annex F: Additional Data on Vulnerability in Chad 178 6 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION LIST OF TABLES Table 1.1. Poverty Rates in Chad under National poverty line, 2018-2018 19 Table 1.2. Multidimensional Poverty in Chad in 2003, 2011 and 2018 23 Table 1.3. Decomposition of Inequality by Household Attributes, 2011-18 (%) 27 Table 1.4. Determinants of change in consumption 2011-2018 at national level, 2011-18 30 Table 1.5. Social and Demographics Characteristics of Poverty in Chad 31 Table 1.6. Simulated Impact of COVID-19 at the Household Level 46 Table 2.1. Poverty, Agricultural, and Income Statistics by Agro-Ecological Zone 54 Table 3.1. Idiosyncratic and Covariate Shocks 75 Table 3.2. Covariate and Idiosyncratic Shocks Assessed in the High-Frequency Phone Survey 91 Table 3.3. Summary of policy recommendations 98 Table 4.1. Human Capital Index Scores by Component, Chad and Comparators, 2020 104 Table 4.2. Human Capital Index Scores by Component, Chad, 2010 and 2020 105 Table 4.3. Education Spending and Outcomes, Chad and Comparator Countries 108 Table 4.4. Food-Security Indicators across Benchmark Countries 123 Table C.1. MPI : Dimensions, Indicators, Deprivation, and Weights 155 Table C.2. Multidimensional Poverty in Chad, 2018 157 Table C.3. Multidimensional Poverty in Chad in 2003, 2011 and 2018 163 Table C.4. Multidimensional Poverty Indicators by Region 165 Table C.5. Multidimensional Poverty Indicators by Region 167 Table D.1. Detrminants of change in consumption 2011-2018 at national and Rural level (Endowment and Returns effects) 168 Table D.2. Detrminants of change in consumption 2011-2018 at Ndjamena and other urban level (Endowment and Returns effects) 170 Table E.1. Summary statistics for households and communities 172 Table E.2. Regression results of per capita consumption (two-level model) 173 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 7 LIST OF FIGURES Figure 1.1. Poverty Rates by regions in Chad 20 Figure 1.2. Distribution of the Poor Population by Area of Residence and Region 21 Figure 1.3. Poverty rate in Chad 22 Figure 1.4. Deprivation by Indicator among Multidimensionally Poor Households, 2003, 2011, and 2018 24 Figure 1.5. Inequality trend 25 Figure 1.6. Growth Incidence Curves, 2011–2018 (%) 26 Figure 1.7. Poverty Rates and Education Levels of Household Heads 32 Figure 1.8. Poverty and Employment Status by Characteristics of Households Heads (% of population) 33 Figure 1.9. Distribution of Consumption Shares by Region 35 Figure 1.10. Housing Conditions 36 Figure 1.11. Access to Basic Services at the Household Level 38 Figure 1.12. Household Asset Ownership 39 Figure 1.13. Poverty Rates and Social Assistance Coverage 41 Figure 1.14. Poverty Rates and FCV-Related Fatalities in Chad 42 Figure 1.15. Forced Displacement and Poverty Rates across Regions 44 Figure 1.16. Poverty and food insecurity of refugee households 45 Figure 1.17. The Impact of COVID-19 on the Chadian Population 47 Figure 2.1. The Share of the Population Living in Urban Areas, Chad and SSA, 2017 52 Figure 2.2. Chad’s Agro-Ecological Zones 53 Figure 2.3. Rural household Income Composition (%) 53 Figure 2.4. Nationwide Allocation of Farmland by Crop Type 56 Figure 2.5. Allocation of Farmland in the Soudanian AEZ by Poverty Status 57 Figure 2.6. Differences in Crop Yields between Poor and Nonpoor Households 58 Figure 2.7. Total Cash Income from Crop Sales (in FCFA) 59 Figure 2.8 Share of Households that Own Nonfarm Enterprise by Household Characteristics 60 Figure 2.9. Nonfarm Enterprises by Location and Head of Household 60 Figure 2.10. Number of Workers in Nonfarm Enterprises by Type of Labor 61 Figure 2.11. Share of Nonfarm Enterprises that Employ Formal Workers by Sector 62 Figure 2.12. Age Distribution across Nonfarm Enterprises 62 Figure 2.13. Net Migration, 1962-2017 63 Figure 2.14. Migrants as a Share of the Population (%) 63 Figure 2.15. Perceived Soil Quality (% of plots) 64 Figure 2.16. Differences in Yield between Plots Rated “Good Quality” and “Poor Quality” (percentage points) 64 Figure 2.17. Income Elasticity of Food Demand in Urban Areas 65 8 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION LIST OF FIGURES Figure 2.18. Access to Financial Services by Gender, Chad and Mali 69 Figure 3.1. Close to half of Chad’s population is vulnerable to poverty. 76 Figure 3.2. Idiosyncratic shocks have a marginally greater impact on household vulnerability than covariate shocks. 77 Figure 3.3. Households in the cash crop and cereals zone are the most vulnerable to poverty. 78 Figure 3.4. Vulnerability due to covariate shocks is prevalent across regions. 78 Figure 3.5. Household vulnerability is driven by low levels of consumption, as most nonpoor households remain close to the poverty line. 79 Figure 3.6. Low levels of consumption are the dominant source of vulnerability across all regions. 79 Figure 3.7. The presence of young children significantly increases the likelihood of household-level vulnerability to poverty. 80 Figure 3.8. Almost all Chadian households experience shocks. 82 Figure 3.9. The incidence of demographic idiosyncratic shocks is especially high. 82 83 Figure 3.10. Households in the cash crop and cereals zone are the most exposed to shocks. Figure 3.11. Female-headed households are more vulnerable to idiosyncratic shocks than their male counterparts. 83 Figure 3.12. Most Chadian households rely on help from family or friends in the event of a shock, a significantly larger share than in most comparable countries. 85 Figure 3.13. When coping with shocks, female-headed households are less likely to use savings and more likely to rely on help from family or friends. 86 Figure 3.14. Shocks lead to major reductions in income and livestock holdings. 87 Figure 3.15. Households resort to detrimental coping mechanisms to respond to natural covariate shocks. 88 Figure 3.16. The impact of shocks on agricultural production is more severe in rural areas. 89 Figure 3.17. The pandemic has heightened household exposure to economic covariate shocks. 92 Figure 3.18. A lack of money has prevented households from accessing essential food items since March 2020. 93 Figure 3.19. The impact of the pandemic on income has worsened over time. 94 Figure 3.20. Chadian households have primarily reduced their consumption and used savings to mitigate the effects of the pandemic. 95 Figure 4.1. Human Capital and Poverty Rates across Regions 106 Figure 4.2. School enrollment rates by gender and age group 109 Figure 4.3. School enrollment rates by age group and wealth quintile 109 Figure 4.4. Primary Enrollment Rates among Children Ages 7–12 by Region 110 Figure 4.5. Distribution of Reasons for Having Never Attended a Formal School among Respondents Ages 7–24 111 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 9 LIST OF FIGURES Figure 4.6. Hours of domestic work and work outside the home in the past week, disaggregated by gender, age group, and school enrollment (mean values) 112 Figure 4.7. Problems reported at schools, disaggregated by school type (%) 113 Figure 4.8. Correlates of the Likelihood of Being Enrolled in School (Logit marginal effects with 95% confidence intervals) 114 Figure 4.9. Literacy Rates by Gender and Age Group 115 Figure 4.10. Maternal Health Indicators across Subpopulations 117 Figure 4.11. Child Health Indicators across Subpopulations 118 Figure 4.12. Stunting Rates and GDP per Capita 119 Figure 4.13. Average Annual Change in Stunting Rates 119 Figure 4.14. Stunting Trends and Inequality 120 Figure 4.15. Stunting Rates at the Prefecture Level 120 Figure 4.16. Reported Incidence of Illness and Visits to a Health Facility by Gender and Age Group 121 Figure 4.17. Reasons for Not Visiting a Health Facility when Ill by Wealth Quintile 122 Figure 4.18. Problems Reported at Public and Private Health Facilities 122 Figure 4.19. Food Insecurity among Rural and Urban Households 124 Figure 4.20. Average Household Food Security across Regions (0 = insecure and 8 = secure) 125 Figure 4.21. Correlates of Being Food Secure (logit marginal effects with 95% confidence intervals) 125 Figure 4.22. Educational Activities during School Closures as of May 2020 127 Figure 4.23. Projected Effects of School Closures due to COVID-19 in the Sahel 128 Figure 4.24. Household Food Security during the Covid-19 Pandemic 129 Figure 4.25. Share of Food-Insecure Households that Attribute their Food Insecurity to COVID-19 130 Figure 4.26. Share of Households that Attribute their Inability to Meet Basic Health Needs to COVID-19 130 Figure 4.27. Correlates of Being Unable to Meet Basic Needs Due to COVID-19 (logit marginal effects with 95% confidence intervals) 131 Figure C.1. Multidimensional Poverty in Chad 158 Figure C.2. Multidimensional Poverty Rates by Region 160 Figure C.3. Overlapping Dimensions of Poverty 161 Figure C.4. MPI Headcount Poverty Rates at Different Poverty Thresholds 164 Figure C.5. Deprivation by Indicator among Multidimensionally Poor Households, 2003, 2011, and 2018 164 Figure C.6. Multidimensional Headcount Poverty Rates by Region 166 10 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION PREFACE This poverty assessment was prepared during a period marked by extraordinary uncertainty around Chad’s security situation and political future. As the report was being finalized, a coalition of rebel groups known as the Front for Change and Concord in Chad (Front pour l’Alternance et la Concorde au Tchad, FACT) began advancing toward N’Djamena, the capital. Chadian forces responded and the conflict escalated rapidly. On April 20 2021, Chad’s president, Idriss Deby, was killed in action while overseeing the deployment of government troops. President Deby had led the country for more than three decades and his sudden death was an enormous shock. Shortly thereafter, General Mahamat Idriss Deby, was named interim president. These additional shocks against the backdrop of the ongoing global Covid pandemic could jeopardize Chad’s limited and extremely fragile gains in poverty reduction and shared prosperity. Since achieving its independence from France in 1960, Chad has suffered numerous outbreaks of both large- and small-scale violence, as well as spillovers from regional conflicts, and the country hosts thousands of refugees and internally displaced persons. Yet the country has also played a key role in advancing regional security as a member of the G5 Sahel. Sustainable poverty reduction underpinned by broad-based gains in education, health, and food security will be vital to ensure a lasting peace, and the World Bank will continue to work closely with Chad and its development partners to establish the foundation for robust and inclusive growth. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 11 ABBREVIATIONS AND ACRONYMS CAR Central African Republic CFAF CFA francs DHS Demographic Heath survey ECOSIT Enquête sur les Conditions de vie des Ménages et la Pauvreté au Tchad HCI Human capital index HDI Human development index EHCVM Enquête harmonisée sur les conditions de vie des ménages FAO Food and agriculture organization GII Gender Inequality Index HFPS High frequency phone survey KNOMAD Global Knowledge Partnership on Migration and Development MPI Multidimensional poverty index UNESCO United Nations Educational, Scientific and Cultural Organization UNHCR United Nations High Commissioner for Refugees WAEMU West African Economic and Monetary Union WDI World Development Indicators 12 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION GLOSSARY Poverty headcount rate or monetary basic needs poverty Multidimensional poverty or multidimensional poverty rate: the percentage of the population living below the index (MPI): a measure of the different deprivations that national poverty line of 242,094 CFA francs (CFAF) per year. an individual or household faces at one time. A person is considered to be multidimensionally poor if she/ Extreme poverty headcount rate: the percentage of the he is defined as being deprived in at least 30 percent of population whose income is insufficient to meet the the indicators covering the six dimensions of wellbeing minimum nutritional requirement of 2,400 kilocalories described above. (kcal) per adult per day. Human capital index (HCI) score: denotes the expected Poverty gap or depth of poverty: the distance between the lifetime productivity of a child born today relative to what poverty line and the average consumption of people living her lifetime productivity would have been had she enjoyed below the poverty line. a complete education and full health. Severity of poverty: a measure of inequality among people Vulnerability to poverty: a measure of the likelihood that a living below the poverty line. nonpoor household will fall below the poverty line during a given period, reflecting both the probability of a negative shock and its potential impact on household welfare. International poverty rate: the percentage of the population whose daily consumption is below the international poverty line of US$1.90 per person per day in2011 purchasing- Poverty-induced vulnerability or structural poverty: power-parity terms. describes persistently low levels of consumption due to low levels of physical and human capital accumulation. Dimension-specific deprivation: the percentage of households (or individuals) defined as being deprived in Risk-induced vulnerability: describes the variability of one of six dimensions of wellbeing: (i) education, (ii) health, consumption, which reflects household-level exposure to (iii) childhood and youth, (iv) access to basic services, (v) shocks. housing conditions, or (vi) assets. For example, a household is defined as being deprived in electricity if it lacks access to an electricity grid, generator, solar panel, or other power source. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 13 EXECUTIVE SUMMARY This Poverty Assessment aims to inform poverty-focused A PROFILE OF POVERTY AND INEQUALITY policymaking in Chad. The report examines recent trends in poverty, inequality and other social indicators and Despite facing considerable structural obstacles, Chad has identifies key constraints on poverty reduction. Chad made significant progress in reducing poverty over the with a GDP per capita of US$ 709 in 2019 (a significant last 15 years. Between 2003 and 2018, the monetary poverty deterioration from US$ 1018 in 2014) is one of the poorest rate fell from 54 percent to 42 percent, and this decline in countries in the world. Overall, poverty remains prevalent monetary poverty was accompanied by improvements in in the country at above 42 percent and with demographic shared prosperity, as the growth of consumption among growth the number of poor Chadians is increasing by households in the bottom 40 percent of the population 200,000 people per annum. The country development outpaced that of the top 60 percent. Meanwhile, Chad’s pattern—in which an export-oriented extractive industry Multidimensional Poverty Index (MPI) poverty rate fell exists alongside an underdeveloped rural economy and from about 70 percent to 59 percent, and the share of the an anemic manufacturing base—is associated with deeply population identified as “deprived” in at least one-third negative long-term social and economic outcomes. The of the 13 MPI indicators fell by about 9 percentage points. analysis finds that broad-based income growth will require The observed improvement in multidimensional poverty sustained investment in health and education supported reflects gains in housing conditions, asset ownership, by increased productivity in the rural economy. Although nutrition, education quality, and access to basic services. agriculture, pastoralism, and related activities provide livelihoods for about 80 percent of the population, the capital-intensive oil sector drives macroeconomic growth, Nevertheless, Chad remains one of the poorest countries exports, and fiscal revenues. Eliminating poverty and in the world. In the 2018 Human Development Index (HDI), boosting shared prosperity in Chad will require robust Chad ranked 187th out of 189 countries and territories. and sustained interventions along three strategic axes: (i) With an HDI score of 0.401, Chad’s performance was economic diversification, with a focus on the rural sector; well below both the average for countries in the low (ii) building resilience to multidimensional shocks; and human development group (0.507) and the average for (iii) accelerating human capital formation. Chad faces Sub-Saharan Africa (0.541). enormous challenges, and the effectiveness of its poverty- reduction efforts will hinge on factors that extend beyond economic policy, including the reestablishment of peace and security through improved governance, as well as a comprehensive effort to address the country’s profound gender disparities. 14 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Poverty outcomes vary widely across regions, with a deep divide between rural and urban areas. Close to 89 percent CONSTRAINTS ON POVERTY REDUCTION of poor households are in rural areas, while only 3 percent Numerous factors contribute to Chad’s high levels of are in the capital city of N’Djamena. The country’s lowest monetary and nonmonetary poverty. While the causes MPI scores are in N’Djamena, followed by the Soudanian of poverty are complex and overlapping, four factors are zone in the south, while the Saharan zone in the north especially relevant: (i) a lack of economic diversification; has the highest average MPI scores. At the regional level, (ii) low productivity in the rural sector; (iii) exposure to the highest MPI scores are in the Lac Region, which is shocks; and (iv) low levels of human capital. Insecurity and experiencing widespread population displacement due inadequate public investment drive all four factors, and to the Boko Haram insurgency coupled with accelerating sustainable progress on poverty reduction will require environmental instability, followed by the Sila Region, fully stabilizing the security situation, consolidating which faces escalating tensions between farmers and the rule of law, and investing in both the physical and pastoralists, spillover effects from the recurrent conflicts in institutional infrastructure necessary to support long- neighboring Sudan, and a sharp decline in global prices for term productivity growth. cotton, the region’s chief export. Lack of Economic Diversification Due to the substantial differences in livelihood patterns within Chad, the analysis separately examines the Saharan, Soudanian, and Sahelian agro-ecological zones. Chad’s undiversified production structure severely limits Transhumance pastoralism is the dominant economic livelihood opportunities while intensifying budgetary activity in much of the sparsely populated Saharan zone. volatility and exacerbating household-level exposure to Households in this zone are less likely to experience shocks. Oil represents about 94 percent of total exports, monetary poverty than households in other areas, but they with cotton, livestock, and other agricultural products are highly likely to experience multidimensional poverty. accounting for the remaining 6 percent. 1 Consequently, the By contrast, monetary poverty is extremely pervasive in the government’s budget and the macroeconomic balances are densely populated Soudanian zone, where most households highly sensitive to oil prices. Agricultural exports consist derive their income from smallholder farming and related largely of unprocessed commodities, and a combination activities. Agro-pastoralism is common in the semi-arid of infrastructure deficiencies and fragile trade linkages Sahelian zone, where a favorable climate for crops and limits the potential for domestic value addition. Many rural livestock offers considerable potential for diversification, households are extremely vulnerable to fluctuating prices but environmental degradation is intensifying conflicts for a narrow range of agricultural commodities, especially between farmers and herders. Each zone faces unique cotton. The manufacturing sector accounts for only about challenges, and addressing poverty will require solutions 8 percent of GDP, and the share of services in total output tailored to address the specific circumstances of Chad’s is declining. diverse households. 1 International Monetary Fund. 2019. Chad: Selected Issues. IMF Country Report No. 19/259. Washington, DC: IMF INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 15 Low Productivity in the Rural Sector In addition to the covariate shocks induced by ecological pressure and security crises, Chadian households are exposed to a wide range of idiosyncratic shocks. The Smallholder agriculture and pastoralism are the main most severe idiosyncratic shocks arise from the severe sources of livelihoods in Chad. Agriculture and pastoralism illness, injury, or death of a household member. Because account for 73 percent of household income; nonfarm most smallholder farmers rely on family labor during activities contribute just 22 percent; and remittances make planting, cultivation, and harvesting, the incapacitation of up the remaining 5 percent. Despite its critical importance a single member can threaten the livelihood of the entire to household welfare, Chad’s rural sector performs far household. Moreover, many households operate nonfarm below its potential. Only 6 percent of the country’s arable microenterprises that rely on family labor, including special land is cultivated, and more than 80 percent of farms skills possessed by individual members. Idiosyncratic cultivate fewer than two hectares. Just 9 percent of the and covariate shocks are related, as climate change may country’s available water resources are being used for accelerate the spread of endemic diseases and provoke or irrigation, and irrigation networks cover less than 1 percent exacerbate conflicts over resources, leading to increased of agricultural land. Low productivity limits the scope for illness, injury, and death at the household level. diversification into cash crops or activities that would generate income during the off season, aggravating the challenges posed by undiversified production and exposure The COVID-19 pandemic threatens to reverse the modest to shocks. Six key constraints diminish the productivity of but important gains in poverty reduction achieved in Chad the agricultural sector: (i) general insecurity and the threat over the last decade. Disruptions in supply chains and of conflict over natural resources; (ii) the risks posed by commercial activities may adversely impact employment unpredictable weather patterns and climatic shocks; (iii) and income levels, and the government’s fiscal policies low levels of human capital; (iv) a severe infrastructure gap; and social protection systems have little capacity to offset (v) a lack of complementary services such as input supply, these effects. While data constraints limit the scope for storage, transportation, and logistics; and (vi) deep gender macroeconomic projections, the impact of the COVID-19 disparities, which reduce women’s access to land and crisis on employment, remittances, and inflation could productive resources. increase the national poverty rate by as much as 5.5 percentage points. Exposure to Climatic Shocks Low Levels of Human Capital Declining levels of rainfall and rising temperatures pose a severe risk to Chadian households, particularly in rural Chad’s score on the World Bank’s 2020 Human Capital areas. In recent decades, a combination of unsustainable Index (HCI) was the second lowest in the world at 0.30. exploitation and the burgeoning effects of climate change This score implies that a child born today in Chad can has lowered the water level of rivers and lakes across the expect to achieve only 30 percent of her lifetime productive country, especially Lake Chad, which has shrunk by 90 potential due to inadequate education and adverse health percent since the 1960s. Persistent drought has accelerated outcomes. Chad’s education system suffers from challenges desertification in the northern part of the country, reducing in both access and quality. Enrollment, attendance, and the size of agro-pastoral areas and spurring a southward completion rates are low; schools are not equitably shift in transhumance patterns, which has heightened distributed across the country; and there are large tensions between farming and herding communities. disparities in access to primary and secondary education Recurrent droughts and falling water levels are also by gender, area of residence, and household income level. having a direct negative impact on agricultural production, While improvements have been observed in certain health threatening the livelihoods of millions of people. indicators, such as maternal and under-five mortality rates, health outcomes remain generally poor, reflecting limited access to quality healthcare. Chad has just 0.4 doctors per 10,000 people, far below the World Health Organization standard of 1 doctor per 10,000 people. An estimated 80 percent of all newborn deaths could be prevented by the presence of a skilled healthcare provider. 16 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Severe infrastructure Gap land titling and property issues was drafted in 2014 but has yet to be adopted. Meanwhile, there are currently no laws governing the grazing rights of itinerant pastoralists Chad’s poor state of infrastructure deters private or mechanisms for resolving land disputes involving investment, inhibits commercialization, and limits them, which heightens tensions between livestock owners connectivity to input and output markets. Chad’s vast land and farmers. area and dispersed population raise the marginal cost of infrastructure. In 2018, the country’s population density was estimated at just 12 people per square kilometer, Deep Gender Disparities compared to 18 in Niger and 72 in Burkina Faso. The difficulty of connecting small, remote communities limits Gender-related inequalities limit the opportunities to market opportunities in agriculture and other sectors. Just increase rural income sustainably. Chadian women have 3.2 percent of households in Chad regularly use asphalt less access to human capital and productive assets, such roads, and only 3.4 percent are within reach of a permanent as land and large livestock, due to social and structural market. In addition, lack of storage capacity and poor access barriers. Early marriage is among the most harmful social to electricity limit the capacity for processing agricultural norms, as it keeps girls out of school and contributes to low goods, forcing households to sell their crops and livestock levels of human capital among women. High fertility rates with little or no value added. and low maternal health indicators limit women’s livelihood options, increase their risk of poverty and vulnerability, Lack of complementary services and contribute to a range of deeply negative health consequences, including early death. In some communities, social norms limit the free movement of women, thereby The ability of the Chadian population to access market constraining their access to economic opportunities. The opportunities and invest in productive activities is gender-related inequalities also extend to access to tenure restricted due to limited insurance markets, inadequate security, financial services, and knowledge. Only 11 percent public investment in information and communications and 5 percent of women over the age of 15 have a mobile- technology (ICT) infrastructure, and weak land rights. money account and a financial account, compared to 20 The extension of mobile coverage has been shown to percent and 13 percent of men, respectively. significantly reduce transaction costs and price dispersion in rural areas in low- and lower-middle-income countries. However, only about 58 percent of Chadians own a A WAY FORWARD cellphone, and this share drops to 46 percent among poor households. Mobile-money services are also limited, with a Eliminating extreme poverty and promoting shared penetration rate of only 16 percent, compared to 24 percent prosperity in Chad will require concerted interventions in Mali. In Chad, over 80 percent of NFEs use personal by policymakers and their international partners. These funds as startup capital, underscoring their limited access efforts must strive to: (i) expand livelihoods opportunities to credit. through productivity growth and diversification; (ii) build resilience to both idiosyncratic and covariate shocks; and The absence of crop or livestock insurance heightens (iii) accelerate human capital formation by improving the production uncertainty arising from weather-related access to high-quality education and healthcare. Gender shocks. The country’s community-based land systems and equity is a critical cross-cutting issue that should inform overlapping land rights, weak tenure security, and costly policies and programs in each of these areas. land certification procedures create uncertainty around long-term ownership and access rights. In 2018, fewer than 3 percent of the country’s cultivated plots had formal titles, and parallel customary and formal land-tenure systems exacerbate uncertainty and discourage investment in fixed capital. In rural areas, informal customary tenure systems, including Islamic land rights, are the dominant model, and obtaining formal land titles is often a lengthy, complex, and costly process. A new land code that would address INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 17 Expanding Livelihood Opportunities through Productivity Investing in Human Capital Growth and Diversification Economic diversification requires a healthy, educated Chad’s unbalanced development model—in which a workforce capable of seizing new opportunities. In small, capital-intensive oil sector drives economic output Chad, limited educational attainment and adverse health while a vast, low-productivity rural economy provides outcomes diminish lifetime productivity and earnings the vast majority of employment—severely weakens potential, contributing to cycles of intergenerational the connection between growth and poverty reduction. poverty. To protect and develop human capital, policymakers A shift toward robust, broad-based growth will require a must simultaneously improve the quantity and quality combination of increased productivity and diversified of education and health services. According to the World production. Diversification, in turn, will require a conducive Bank’s 2018 World Development Report, successful learning business environment, especially in agriculture and related requires: (i) prepared students; (ii) effective teachers; (iii) sectors. Chadian farmers and pastoralists have the potential appropriate inputs; and (iv) skilled school management and to accelerate diversification both by expanding into new good governance across the educational system. Efforts to forms of production and by increasing value addition in improve public health should focus on improving health existing production chains. Enhancing the productivity of coverage and quality among marginalized communities, smallholder farming, increasing market connectivity, and enhancing nutrition, and expanding access to sanitation. creating mechanisms to reduce vulnerability to shocks are crucial to support diversification in the rural economy. Addressing Gender Disparities In addition to directly enhancing the welfare of rural households, the production of larger agricultural surpluses Gender is a critical cross-cutting issue in Chad and and a wider range of both food and nonfood commodities should be mainstreamed into all development efforts. will create new opportunities in other sectors, including Chad ranked 160th out of 162 countries and territories on agro-processing, trade, and logistics. the United Nations’ 2018 Gender Inequality Index. Gender disparities in education and health indicators, land access, Building Resilience to Shocks the use of financial services and technology, and the ability to participate in the full range of socioeconomic activities Despite their extreme exposure to shocks, few Chadian have profound consequences for household productivity, households have access to effective coping mechanisms. income diversification, and intergenerational economic To enable households to protect their physical, financial, mobility. Addressing the constraints to growth and poverty and especially human capital, policymakers should develop reduction identified in this report will require interventions adaptive social protection systems that are able to: (i) that effectively: (i) address gender gaps in human capital swiftly provide income support and essential information accumulation; (ii) increase demand for female labor; and during shocks and crises; (ii) facilitate access to education, (iii) improve women’s access to productive assets, including healthcare, nutrition, and other critical services; (iii) ensure land, credit, and physical inputs. Prioritizing maternal equal access to productive resources and economic health, encouraging girls to complete school before opportunities; (iv) support the development of skills marriage, proactively recruiting women as agricultural that generate greater labor-market returns; and (v) build extension workers and lead farmers, and implementing delivery platforms for coordinated multisectoral initiatives. both legal and administrative reforms to equalize access Limited insurance markets, inadequate public investment to land and financial services could enable rural women to in infrastructure, and weak land rights restrict the ability realize far more of their productive potential, with positive of the rural population to access market opportunities and effects on their economic security, health, and education. invest in productive activities. Improved communication and transportation links could allow rural extension services and international partners to form farmer organizations capable of leveraging economies of scale in input supply and output marketing. Enhancing social protection through adaptive safety nets and creating a framework for index- based insurance can enable households to build resilience and transition out of poverty. 18 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION CHAPTER UNDERSTANDING 01 POVERTY IN CHAD INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 19 1.1. POVERTY AT THE NATIONAL AND SUBNATIONAL LEVELS 1.1.1 RATES AND TRENDS of poor people are in rural areas, where low-productivity agriculture and livestock herding are the main livelihoods, while just 3 percent are located in the capital city, Poverty in Chad is pervasive and severe. The Fourth N’Djamena (Figure 1.2a). Poverty rates are highest in regions Household Consumption and Informal Sector Survey that border on the Central Africa Republic, Cameroon, (Quatrième Enquête sur les Conditions de vie des Ménages Sudan, and Nigeria. These areas are affected by conflict et la Pauvreté au Tchad, ECOSIT 4), conducted in 2018, found and instability in neighboring countries, and they host that 3.4 million women and 3.1 million men—representing thousands of refugees and internally displaced people. about 42 percent of Chad’s population—live below the The regions of Mandoul and Logone Oriental, which border national poverty line of 242,094 CFA francs (CFAF) per on the Central Africa Republic, are home to 8 percent and year. Approximately 15 percent of the population, or 2.4 9 percent of Chad’s poor population, respectively, while million people, are unable to meet the basic nutritional Mayo-Kebbi Est and Mayo-Kebbi Ouest, on the border with requirement of 2400 kilocalories per day.2 Cameroon, are home to a combined 17 percent (Figure 1.2b). Due to the concentration of poverty in the rural Soudanian Chad’s poor population is concentrated in rural areas, and zone, policies designed to strengthen social protection poverty rates vary widely by region.3 Almost 89 percent systems, empower women and youth, and encourage Table 1.1. Poverty Rates in Chad under National poverty line, 2018-2018 Headcount Depth Severity N’Djamena 13.8 2.9 0.9 Other urban 23.0 6.4 2.5 Urban 19.4 5.0 1.9 Rural 49.7 15.1 6.3 Chad 42.3 12.6 5.2 Source: World Bank staff calculation using data from ECOSIT 4 2 The poverty-measurement methodology is described in Annex A. 3 Chad is divided into 23 administrative regions (régions). Terms such as “regional level” therefore refer to subnational jurisdictions within Chad, not to the larger multi-country region in which Chad is located. 20 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure 1.1 : Poverty Rates by regions in Chad 70.0 63.1 60.0 57.8 58.7 60.0 60.1 52.9 48.4 50.0 41.1 41.7 42.3 42.3 43.5 44.3 40.0 37.5 38.4 34.6 30.9 30.0 27.7 21.7 22.3 24.4 20.0 13.8 10.0 0.0 N’djamena city Chari - Baguirmi Ennedi Borkou Kanem Hadjer - Lamis Barh-El-Gazal Wadi Fira Ouaddai Moyen-Chari Salamat Sila Chad Logone Occidental Lac Batha Logone Oriental Mandoul Mayo-Kebbi Est Tandjilé Guera Mayo-Kebbi Ouest Source: WAEMU and World Bank staff calculation using data from ECOSIT 4 entrepreneurship should target the country’s southern poverty rates to rise as high as 60 percent in Tandjilé. regions and reflect the unique needs of agricultural During the 2017/18 agricultural year, the Sahelian area households (Figure 1.2c). experienced a late rainy season and low total rainfall, which caused agricultural production to drop by more than In recent years, multiple shocks have exacerbated poverty 20 percent in the regions of Kanem, Wadi Fira, and Bahr El at the regional level. Falling international cotton prices Ghazal, increasing rates of poverty and vulnerability among have severely impacted household incomes in cotton- rural households. growing areas such as Tandjilé, Logone, and Sila, causing INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 21 Figure 1.2: Distribution of the Poor Population by Area of Residence and Region (a) Share of Poor Population by Area of Residence (%) (b) Poverty Rates by Region (%) 3.1 8.1 53.0 - 63.1 44.4 - 52.9 34.7 - 44-3 24.5 - 34.6 13.8 - 24.4 88.8 Ndjamena Other Urban Rural © Vemaps.com (c) Share of Poor Population by Region (%) 10.00 9.7 8.6 8.9 8.00 7.7 7.8 6.5 6.00 5.9 5.6 4.9 5.2 4.1 4.1 4.00 3.4 3.7 3.1 2.7 2.7 1.9 2.0 2.00 0.6 0.8 0.00 Borkou/Tibesti Ennedi Barh-El-Gazal Kanem Chari-Baguirmi Salamat Ville de N’djamena Sila Hadjer-Lamis Wadi Fira Lac Batha Moyen-Chari Guéra Ouaddaï Logone Occidental Mayo-Kebbi Ouest Mandoul Tandjilé Logone Oriental Mayo-Kebbi Est Source: World Bank staff calculation using data from ECOSIT 4 22 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 1.1.2 GAINS IN POVERTY REDUCTION average of 6 percentage points percent between 2011 and Poverty rates at the national level have declined 2018. By contrast, the severity of urban poverty declined by substantially over the last 15 years. 4 The share of Chadians just 1 percentage point over the same period. Nevertheless, living below the national poverty line fell by 12 percentage poor rural households experience greater consumption points between 2003 and 2018, with especially sharp deprivation than do their urban counterparts. declines observed during 2003-2011 and 2014-2018, when the poverty rate fell by an average of 1 percentage point and 0.8 percentage points per year, respectively (Figure 1.3). While The quality of poverty reduction is confirmed by the poverty rates have declined in urban areas, most of the declining trend in the multidimensional5 poverty rate. The overall reduction in poverty has occurred in rural areas. In national MPI score fell from 70.3 percent to 58.7 percent addition, rural poverty has become significantly less severe over the period, while the multdimensionnal poverty overall, with the measure of severity of poverty falling by an headcount rate declined by 8.7 percentage points (Table 1.2). Figure 1.3: Poverty rate in Chad (a) National poverty rate 70 58.4 60 52.5 52.6 49.7 50 54.8 40 46.7 45.5 42.3 30 20 24.4 20.9 20.9 19.4 10 0 2003 2011 2014 (Est.) 2018 Urban Rural Chad (b) Depth of poverty (b) Severity of poverty 15.0 11.7 12.6 25.0 23.1 22.6 20.0 21.5 16.5 10.0 19.7 15.1 10.8 10.8 7.0 15.0 6.3 14.0 12.6 5.9 10.0 5.0 3.2 5.2 5.0 7.4 6.6 3.0 2.1 5.6 5.0 1.9 0.0 0.0 2003 2011 2014 2018 2003 2011 2014 (Est.) 2018 (Est.) Urban Rural Chad Urban Rural Chad Source: World Bank staff calculation using data from ECOSIT 4 4 The poverty-trend analysis presented in this section is based on consumption surveys conducted in 2003, 2011, and 2018 and on the 2014 Demographic and Household Survey. While the 2003 and 2011 consumption surveys are directly comparable to each other, they are not comparable to the 2018 Household Consumption and Informal Sector Survey. Data from the 2014 Demographic and Household Survey are utilized, via the survey-to-survey imputation methodology, to estimate the poverty rate in 2014 and identify trends between 2014 and 2018. The methodology is presented in Annex B. 5 More details on multidimensional poverty in Chad are presented in Annex C. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 23 Table 1.2. Multidimensional Poverty in Chad in 2003, 2011 and 2018 MPI H A 2003 70.3 97.5 72.1 2011 66.3 95.2 69.6 2018 58.7 88.8 66.1 Change 2003-2011 (in percentage points) -4.0 -2.2 -2.5 Change 2011-2018 (in percentage points) -7.7 -6.4 -3.6 Source: World Bank staff calculation using data from ECOSIT 4 Meanwhile, the intensity of multidimensional poverty fell Pro-poor nutrition and education policies also contributed by 6 percentage points, from 72.1 percent in 2003 to 66.1 to the decline in multidimensional poverty. The share percent in 2018, indicating that living conditions improved of multidimensionally poor households suffering from among households experiencing multidimensional poverty. challenges related to nutrition, school attendance, and grade Multidimensional poverty reduction accelerated over the repetition fell significantly between 2003 and 2018. While period, and the national MPI score fell faster between education indicators improved among multidimensionally 2011 and 2018 than between 2003 and 2011. Moreover, the poor households, Chad’s poor population continued to face decline in MPI rates was significant at all poverty thresholds high levels of deprivation in terms of years of schooling (see Figure C.4 in Annex C). While the incidence of (75.7 percent) and literacy rates (88.7 percent) in 2018. multidimensional poverty declined between 2003 and 2018, much of the country continues to face severe deprivation. Progress in improving access to basic services among multidimensionally poor households has been mixed. The decline in MPI scores reflected significant progress The deprivation rate for healthcare fell by 4.5 percentage in improving housing conditions and asset ownership points between 2003 and 2018, but this improvement was among the country’s poorest households. The material more than offset by a drop in access to health specialists, quality of housing and the rate of asset ownership are resulting in an overall deterioration in the health the two MPI indicators that experienced the largest gains component of the MPI. While the deprivation rate for over the last 15 years. Between 2003 and 2018, the share access to improved water fell by 9.3 percentage points, the of multidimensionally poor households with low-quality deprivation rate for sanitation and cooking fuel increased. walls and roofs declined by 20 percentage points and 13 Finally, the deprivation rate for electricity declined by percentage points, respectively, and the share that owned a mere 1 percentage point, highlighting the persistent at least some modern assets increased by 17 percentage structural constraints on electricity access in Chad. points (Figure 1.4). 24 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure 1.4: Deprivation by indicator among Multidimentionally Poor Households, 2003, 2011, and 2018 37.0 74.1 Nutrition 59.6 Asset 63.7 91.9 49.7 28.9 room 32.7 49.6 34.3 Health Specialist 29.0 43.2 71.8 Housing roof 85.3 85.0 31.5 affordable healthcare 51.7 36.0 68.2 Housing wall 95.7 88.9 21.5 96.2 Child labour 25.2 95.7 Housing floor 12.3 95.0 62.0 96.4 Electricity 98.5 Child grade 66.4 97.3 73.8 71.3 Sanitation 74.4 47.3 68.4 School attendance 51.5 55.9 42.6 Water 62.8 52.0 88.7 Literacy 95.6 97.7 92.3 99.3 Cooking fuel 89.2 75.7 94.5 75.2 98.4 years of schooling Waters 77.6 95.0 0.0 20.0 40.0 60.0 80.0 100.0 0.0 20.0 40.0 60.0 80.0 100.0 2018 2011 2003 2018 2011 2003 Source: World Bank staff calculation using data from ECOSIT 4 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 25 1.1.3 INEQUALITY AND SHARED PROSPERITY PATTERNS households than among their wealthier counterparts. Between 2014 and 2018, consumption grew at an average annual rate of 1.1 percent among households in the bottom Inequality increased between 2003 and 2011 but declined 40 percent of the consumption distribution versus a rate of thereafter. From 2003 to 2011, as poverty rates fell, just 0.1 percent among households in the top 60 percent. inequality in the consumption distribution widened from In rural areas, the rate of consumption growth among 39.4 percent to 42.1 percent (Figure 1.5). However, inequality households in the bottom 40 percent has been consistent narrowed to 35 percent in 2014 and 33.6 percent in 2018, with the national average at 1.1 percent, but consumption with an especially significant decrease observed in urban growth among households in the top 60 percent has areas. The urban Gini coefficient fell from 37.4 percent in averaged 1.5 percent. However, consumption among 2014 to 33.7 percent in 2018. households at all income levels grew faster in 2003-2011 than in 2014-18. During recent years, indicators of shared prosperity have improved. Consumption has increased faster among poor Figure 1.5: Inequality trend 0.450 0.421 0.394 0.400 0.350 0.336 0.350 0.300 0.250 0.200 0.150 0.100 0.050 0.000 2003 2011 2014 (Est.) 2018 Chad Source: World Bank staff calculation using data from ECOSIT 4 26 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Box 1.1: Inequality and Pro-Poor Growth Patterns Inequality appears to have declined between 2011 and 2018, and signs of pro-poor growth are emerging. The consumption-based Gini coefficient fell from 42.1 percent in 2011 to 33.4 percent in 2018. Inequality decreased across the board, but it fell fastest in rural areas, dropping from 41.6 percent to 30.3 percent, compared with a more modest decline in urban zones from 36.2 percent to 33.6 percent. The growth incidence curve for 2011–18, which shows the percentage change in average consumption for each percentile of the distribution, is downwardly sloped, indicating faster consumption growth among the poorest segments of the population (Figure 1.6). Again, rural areas drove the pro-poor pattern, while pro-poor growth was limited in urban areas, especially N’Djamena. Figure 1.6 Growth Incidence Curves, 2011–18 (%) Chad Rural 16 20 16 20 12 12 8 8 4 4 0 0 0 20 40 60 80 100 0 20 40 60 80 100 Growth rate by percentile Growth rate by percentile Growth rate in mean Growth rate in mean Other urban Ndjamena 20 20 16 16 12 12 8 8 4 4 0 0 0 20 40 60 80 100 0 20 40 60 80 100 Growth rate by percentile Growth rate by percentile Growth rate in mean Growth rate in mean Source: World Bank staffs calculation using data from ECOSIT3 and ECOSIT4. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 27 Table 1.3 : Decomposition of Inequality by Household Attributes, 2011-18 (%) 2011 2018 Share of inequality Share of inequality explained by… (%) explained by… (%) Theil-L Theil-T Theil-L Theil-T Education of household head 3.62*** 4.0*** 6.06*** 6.59*** Gender of household head 0.17 0.164 0.01 0.01 Family type 6.03*** 5.95*** 17.8*** 18.0*** Age of household head 1.45** 13.6** 0.75*** 0.74** Activity status of household head 0.14 0.131 0.15 0.15 Employment sector of household head 9.46*** 9.34*** 15.9*** 15.6*** Urban/rural status 8.15*** 8.75*** 13.5*** 13.6*** Regional location 10.1*** 10.2*** 10.1*** 10.2*** Source: World Bank staffs calculation using data from ECOSIT3 and ECOSIT4. Note: * Significant at the 10 percent level; ** significant at the 5 percent level; *** significant at the 1 percent level. Results are based on bootstrap 100 replications. 28 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Survey-to-survey imputation based on the small-area estimation approach can assess changes in poverty in cases where the 2011 and 2018 survey designs and methodologies are incompatible. The results confirm the robustness of the finding that poverty rates fell over the period, consistent with the observed decline in multidimensional poverty. However, the Gini coefficient remained stable at a moderate level of 33 percent. Moreover, the imputed pro-poor growth pattern is less pronounced than that observed from actual data. The high level of inequality registered in 2011 is probably due to noise in the 2011 survey, which seems to have affected the measurement of inequality but not of poverty. Consumption-based estimates of inequality tend to change slowly, and the rapid decline in inequality observed between 2011 and 2018, particularly in rural areas, appears to be unrealistic given the absence of redistribution programs or structural economic transformation. Moreover, the decomposition of inequality between and within population subgroups reveals significantly smaller between-group shares in 2011 than in 2018, which cannot be explained by structural changes in the socio- demographic characteristics of the population. Large inequalities between households are based on their demographic composition and the sector of employment of household heads, suggesting that poverty reduction could be accelerated through a faster demographic transition and economic transformation. In 2018, differences in the number of children per household accounted for about 18 percent of total inequality (Table 1.3). The per capita consumption level of households with fewer than three children below age 15 was, on average, 1.8 times higher than that of households with five or more children below age 15. This finding suggests that efforts to reduce the fertility rate and catalyze the demographic transition could accelerate poverty reduction. Similarly, differences in the employment sector of the household head account for about 16 percent of overall inequality. Households headed by a worker in the services and industrial sectors have average consumption levels about 1.7 times and 15 times higher than those headed by an agricultural worker, respectively. Consequently, speeding the reallocation of labor to more-productive sectors could foster income growth and poverty reduction. Inequality between geographic regions is also relatively high. Differences between urban and rural areas explain about 14 percent of total inequality, while differences between geographic locations explain about 10 percent. The persistence of large spatial inequalities can exacerbate social tensions and intensify fragility, hindering inclusive growth and shared prosperity. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 29 1.1.4 DRIVERS OF POVERTY REDUCTION in rural areas and among younger household heads. Meanwhile, some of the poorest households, including those headed by older workers, have been least able to invest Chad’s recent decline in poverty rates has been driven in education and remained confined to low-productivity by improvements in both asset endowments and returns sectors, have experienced a decline in the returns to among poor households. Changes in consumption reflect years of schooling. To accelerate poverty reduction, gains in human and physical capital, access to basic investment in human capital must be accompanied by the services, and employment opportunities, as well as the creation of higher-value-added jobs in the industrial and returns to education, employment, and physical capital. services sector. The technical decomposition of these changes is presented in Annex D. Increased returns to the ownership of physical assets, especially cellphones, have also contributed to welfare Improved access to employment has bolstered the improvements and poverty reduction. Expanded living standards of Chadian households. Recent poverty cellphone ownership and increased economic returns to reduction has been partially explained by the rising shares the productive use of cellphones have generated positive of households headed by an employed worker and by a changes in consumption. Rates of cellphone ownership worker with a secondary job (Table 1.4). Slight increases have risen fastest among poorer and rural households, in nonfarm self-employment and in the average share yielding a significant impact on poverty reduction. Wealthy of working household members have also contributed and middle-class households have also experienced gains to positive changes in consumption. However, rising in the returns to motor-vehicle ownership, but this seems employment rates have not been accompanied by a to have had only a limited impact on poverty reduction. shift to more-productive sectors, and almost 70 percent of households continue to earn their livelihood from Improvements in access to electricity and improved agriculture. Between 2011 and 2018, a modest increase in drinking water also contributed to poverty reduction, the share of households headed by a service worker had a albeit to a lesser extent. While rates of electricity access limited but positive impact on the poverty rate. remain low nationwide, rapid gains in N’Djamena have slightly reduced urban poverty. Broader progress in rates Persistently low levels of human capital among household of access to improved water sources, which rose from heads has slowed the reallocation of labor to more- 35 percent in 2011 to 55 percent in 2018, appears to have productive sectors. Between 2011 and 2018, the share of enabled women to spend less time on domestic activities household heads with education beyond the primary level and increase their engagement in nonfarm economic rose by 3 percentage points, while the share of household activities, contributing to a modest improvement in living heads with no formal education decreased by 4 percentage standards among poor households. points. Gains in educational attainment were concentrated 30 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Table 1.4 : Determinants of Changes in Consumption at the National Level, 2011-18 Extreme poor Poor Middle class Richest Total 0.808*** 0.480*** 0.408*** 0.271*** Endowments 0.117 0.269** 0.230** 0.196 Head employed 0.012* 0.007 0.009* 0.007 Head second job 0.022*** 0.017*** 0.006 0.008 Head self employed 0.032** 0.015 0.004 0.009 Head farm employment -0.039*** -0.026*** -0.027*** -0.026** Livestock 0.008*** 0.007*** 0.005*** 0.007*** Head nonfarm employment -0.004 0.003 0.004* 0.003 Head education 0.001 0.003* -0.003 0.005*** Own mobile phone 0.143*** 0.135*** 0.096*** 0.102*** Own car/motorcycle 0 0.001 0.001* 0.007* Access to electricity 0.001 0.005** 0.006*** 0.023*** Access to improved water 0.002* 0.003*** 0.004* 0.006*** Share of workers in the HH 0.007*** 0.008*** 0.004** 0.010*** Returns 0.692*** 0.212* 0.178 0.075 Head employed -0.039 0.016 0.001 -0.089 Head second job 0.002 0.028** -0.012 -0.007 Head self employed 0.028 -0.023 -0.024 -0.035 Head farm employment -0.025 -0.025 -0.018 0.046 Livestock 0.001 0.003* 0 0.005*** Head nonfarm employment -0.037** -0.016 -0.001 0.054*** Head education -0.086*** -0.040*** -0.012 0.005* Own mobile phone 0.018*** 0.012*** 0.001 -0.002 Own car/motorcycle -0.001 0.003 0.011** 0.024*** Access to electricity 0.006 0.003 0.002 0.004 Access to improved water 0.009 0.002 0.001 0.023*** Share of workers in the HH 0.042* 0.058*** 0.053*** 0.070*** Source: World Bank staff calculation using data from ECOSIT 3 and Ecosit 4 Note: Extreme poor denotes households in the bottom two deciles of the distribution; poor denotes households in the third and fourth decile; middle class denotes those in the fifth decile; and richest denotes those in the top decile. * Significant at the 10 percent level; ** significant at the 5 percent level; *** significant at the 1 percent level. Numbers in parentheses are bootstrap standard deviations based on 100 replications. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 31 1.2. POVERTY PROFILE 1.2.1 SOCIAL AND DEMOGRAPHIC CHARACTERISTICS OF experience shows that poor households often regard a large number of children as a form of insurance against infant POOR HOUSEHOLDS mortality and as means of ensuring support for the parents in old age. Household poverty status is also associated with In Chad, poor households tend to have many children, and early marriage, limited access to contraception, inadequate heads of poor households are likely to be self-employed health information, a lack of family-planning services, and and less educated than average. Across the country, having other factors that tend to increase the number of children a large number of children in the household is strongly per household. Moreover, large numbers of children also associated with poverty status. Poor households have an contribute to intergenerational poverty, as the increased average of 1.5 more children under the age of 17 than do demand for food and care strains the resources of poor their nonpoor counterparts. Poor households also have households and limits their ability to invest in the an average of 1.7 more total household members than human capital of each child. Chad’s score on the 2020 do nonpoor households. In N’Djamena, the average poor Human Capital Index was the second lowest in the world, household has 7.7 members, more the average for poor underscoring the severely constrained opportunities for households in rural and other urban areas. Polygamy is more productive employment and upward economic mobility prevalent among poor households, which contributes to facing Chadian children. their high average fertility rates (Table 1.5). The international Table 1.5: Social and Demographics Characteristics of Poverty in Chad Chad Non-poor Poor Poor N’Djamena Other Urban Urban Rural Household Size 5.3 4.7 6.4 7.7 6.5 6.8 6.3 Children under 5 1.1 0.9 1.4 1.4 1.4 1.4 1.4 Children Ages 5-17 2.0 1.7 2.7 3.5 2.9 3.0 2.7 Adults Ages 18-64 2.0 2.0 2.1 2.5 2.2 2.2 2.1 Elders over 64 0.1 0.1 0.1 0.2 0.1 0.1 0.1 Dependency Ratio 0.5 0.5 0.6 0.6 0.6 0.6 0.6 Age of Household Head 42.7 42.4 43.2 48.2 44.1 44.9 43.0 Male Household Head 74.9 74.8 75.1 71.1 67.6 68.3 75.9 Polygamous Household 14.5 12.8 17.5 18.8 9.0 11.0 18.2 Source: World Bank staff calculations using data from ECOSIT 4 32 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION The level of education of the household head is also headed by a person with complete or incomplete primary correlated with poverty status. Households headed by a education (55.1 and 52.8 percent, respectively) and person with tertiary or upper-secondary education have marginally lower among households headed by a person the lowest poverty rate at 9.5 percent (Figure 1.7a). Poverty with no education (42 percent). While education levels rates are far higher for households headed by a person with are correlated with other factors that influence poverty upper secondary education (37.5 percent) and higher still for status, the low overall quality of education in Chad limits those headed by a person with lower secondary education the impact of educational attainment on employment (41.4 percent). Poverty rates are highest among households opportunities, income, and poverty. Figure 1.7 : Poverty Rates and Education Levels of Household Heads Poverty by level of education of head of household 60.0 52.8 55.1 50.0 42.3 41.4 37.5 40.0 30.0 20.0 9.5 10.0 0.0 No education Primary Primary Lower Upper University uncompleted completed secondary secondary Poverty decomposition by education level of household head 1.4 Rural poor 80.0 2.9 4.8 10.8 Urban poor 44.2 9.1 11.9 19.9 15.0 Poor 76.2 3.6 2.5 6.4 11.3 Non poor 53.8 11.5 9.0 14.1 11.6 Chad 61.5 8.8 6.8 11.4 11.5 0.0 20.0 40.0 60.0 80.0 100.0 Agriculture Trade Services Other Unemployed Source: World Bank staff calculations using data from ECOSIT 4 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 33 Over half of households headed by agricultural workers headed by domestic workers is the highest in the country are poor. The Chadian agricultural sector suffers from at 48 percent, even exceeding the poverty rate of 45 low productivity and a high degree of exposure to shocks, percent among households headed by the unemployed and 52 percent of households headed by an agricultural (Figure 1.8b). By contrast, poverty rates are lowest among worker are poor. Employment in the construction sector households headed by salaried staff members at less is also associated with higher poverty rates, and three than 7 percent, followed by households headed by skilled out of 10 Chadians residing in a household headed by a workers at 12 percent (Figure 1.8c). The low poverty rate construction worker are below the poverty line (Figure among households headed by service workers suggests 1.8a). Self-employed and unskilled construction workers that employment in services may be a viable means to with low-paying, short-term contracts are especially escape poverty. vulnerable to poverty. The poverty rate among households Figure 1.8: Poverty and Employment Status by Characteristics of Households Heads (% of population) (a) Poverty rates by sector of employment of the household head Unemployed 44.9 Services 15.0 Education/ 19.3 Health Transport 19.7 Hotel 5.1 Trade 17.4 Construction 30.3 Manufacturing 28.6 Mining 24.4 Agriculture 51.9 0.0 10.0 20.0 30.0 40.0 50.0 60.0 Source: World Bank staff calculations using data from ECOSIT 4 34 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION (b) Poverty rates by employment status of the household head Unemployed 44.9 Home worker/trainee 47.9 Self employed 45.3 Unqualified worker 30.0 Qualified worker 12.1 Staff 6.8 0.0 10.0 20.0 30.0 40.0 50.0 60.0 (c) Total population by employment status of the household head 1.4 Rural poor 80.0 2.9 4.8 10.8 Urban poor 44.2 9.1 11.9 19.9 15.0 Poor 76.2 3.6 2.5 6.4 11.3 Non poor 53.8 11.5 9.0 14.1 11.6 Chad 61.5 8.8 6.8 11.4 11.5 0.0 50.0 100.0 Agriculture Trade Services Other Unemployed INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 35 Figure 1.9: Distribution of Consumption Shared by Region 100% 80% 60% 40% 20% 0% al an al ha ri-B kou i jer- a is em Log ciden c one tal Ma and l Ma Ke l Keb Est yen t Oua ri Sal ai Tan t Vill Wa jile e N Fira h-E a edi l st Sila irm a yo- ou Enn l-Gaza Mo Oues a Oc La Had Guer Bar amen a Lam ion Rur ent dd Oue am Urb Bat -Ch d Kan yo- bbi Cha Bor agu e d di Nat Ori bi 'dj M one Log Food Housing Rent Clothes Durable Goods Health Education Transportation Other 1.2.2 FOOD SECURITY, LIVING CONDITIONS AND ASSET produced and purchased food is not enough to meet the basic needs of much of Chad’s population, and inadequate OWNERSHIP nutrition prevents the food poor from escaping poverty. Chad’s high monetary poverty rates are coupled with Poor households tend to live in houses made of low-quality pervasive food poverty, particularly in rural areas. materials, and housing conditions in N’Djamena differ Approximately 15 percent of the population, or 2.4 million greatly from those in rural areas and other urban centers. people, are estimated to be “food poor,” which is defined Most poor households own their own homes, and renting as being unable to meet a basic nutritional requirement is common only among poor households in N’Djamena. of 2,300 kilocalories per day. Rates of food poverty range Almost nine out of ten poor households own their own from 18.5 percent in rural areas to 2.5 percent in N’Djamena homes, versus just seven out of ten nonpoor households and 6.9 percent in other urban areas. Chadian households (Figure 1.10). Housing costs are relatively high in N’Djamena, dedicate an average of about 60 percent of their budgets to and poor households are often unable to afford their food, versus just 3.4 percent to healthcare and 1.1 percent to own homes. One-third of poor households in N’Djamena education, yet many remain unable to satisfy their essential are renters, and they often rent low-quality houses that nutritional needs (Figure 1.9). In addition, an average of lack utilities such as piped water and electricity. Poor 19 percent of food consumption comes from households’ households in N’Djamena that own their homes primarily own production (23 percent in rural areas), and household live in peri-urban areas, many of which are marked by agriculture relies on rudimentary productive practices and limited access to utilities and a high risk of insecurity. The is highly vulnerable to shocks. The combination of self- outskirts of N’Djamena have recently experienced flooding, 36 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure 1.10: Housing Conditions (a) Housing type (b) Dwelling ownership 0.2 0.3 Rural poor 0.0 97.8 2.0 93.8 5.6 0.4 Rural poor Urban poor 0.6 49.1 48.3 2.0 Other Urban poor 74.6 8.4 17 0.0 Other Urban poor 0.4 38.5 58.9 2.2 Ndjamena poor 34.4 35.6 29.4 0.5 Ndjamena poor 1.0 82.7 14.9 1.5 0.3 Poor 90.7 1.9 7.1 0.3 Poor 5.2 92.5 2.0 Non-poor 1.6 19.5 Non-poor 79 8.1 12.4 0.6 76.8 2.1 Chad 1.1 14.6 82.2 2.1 Chad 83 5.9 10.6 0.5 0 20 40 60 80 100 0 20 40 60 80 100 Building Townhouse Owner Rent Individual house Other Free Other (c) Improved dwelling materials (d) Cooking Fuel 15.1 0.1 Rural poor 27.6 Rural poor 1.4 97.9 0.7 1.5 40.9 0.0 Other Urban poor 45.9 Other Urban poor 9.7 90.3 0.0 9.2 97.9 Ndjamena poor 45.4 Ndjamena poor 37.2 8.2 38.1 16.5 13.5 19.3 1.1 Poor 29.5 Poor 2.2 95.7 1.0 2.5 39.5 Non poor 38.3 Non poor 8.8 6.8 82.8 1.6 10.4 32.5 Chad 35.3 Chad 6.1 5.2 87.3 1.4 7.6 0.0 20.0 40.0 60.0 80.0 100.0 0% 20% 40% 60% 80% 100% Improved roof Improved wall Electricity/Gas Charcoal Improved floor Wood Other Energy Source: World Bank staff calculations using data from ECOSIT 4 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 37 which destroyed houses and left thousands of people water supply and improve water quality in N’Djamena, only without shelter or food. Due to these circumstances, poor 3 out of 10 poor households have access to the capital households in N’Djamena are more likely to experience city’s water grid, and even those poor households that have nonmonetary poverty than are their counterparts in rural access to the grid often cannot afford the cost of piped areas or other urban centers. water and instead rely on boreholes. Poor Chadians tend to live in homes made of wood, straw, The vast majority of Chadian households lack access sand, mud wattle, and metal sheets. At the national level, to adequate sanitation services. More than two-thirds only 7 percent of Chadians live in a house with an improved of households do not have a toilet, and 14 percent use floor, 35 percent live in a house with improved walls, and rudimentary latrines or open pits, which increase the risk 33 percent live in a house with an improved roof. Among of disease. While most poor households in N’Djamena poor households, these shares fall to 3 percent, 30 percent, have access to improved sanitation, the same is true for and 19 percent, respectively. Poor households in N’Djamena just 4 percent of their rural counterparts (Figure 1.11b). The have higher-quality housing than their counterparts in rural government’s efforts to expand sanitation access must areas and other urban centers: almost 98 percent and 45 consider rural/urban and regional disparities, as well as percent of poor households in N’Djamena have improved the needs of remote communities and vulnerable groups roofs and walls, respectively, versus just 15 percent and 28 such as women and people with disabilities. percent of poor rural households. Poor households are also frequently overcrowded, with an average of 3.4 persons per Few poor households have access to electricity. The room, well above the national average of 2.8. Overcrowding is national electrification rate is 11.8 percent among nonpoor especially common among poor households in N’Djamena, households and just 1.4 percent among poor households which average 4.5 persons per room. (Figure 1.11c). While 22.5 percent of poor households in N’Djamena are connected to the electricity grid, power Poor households have limited access to safe drinking outages are common: in nationwide surveys, about 75.5 water. Only 55 percent of the poor population has access percent of households with electricity access report to water from an improved source. Of these, 9 percent have experiencing a power outage during the previous seven access to piped water, and 46 percent use boreholes (Figure days. Nationwide, the average duration of a power outage 1.11a). Access to improved water sources is particularly is 4.5 days, and an average of 1.89 outages occur each day. limited among the rural poor, and one in ten poor rural The electricity supply is particularly unreliable in urban households relies on surface water. The use of unsafe water areas outside the capital, where all poor households report sources increases exposure to water-related diseases such experiencing power outages. Unreliable power access limits as cholera and diarrhea. In addition, many poor people, income-generating opportunities, compels households to particularly in the Saharan and Sahelian zones, spend use hazardous cooking fuels, and hinders efforts to improve substantial time and energy retrieving water from distant living conditions. sources. Despite the government’s efforts to expand the 38 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure 1.11: Access to Basic Services at the Household Level Water Sanitation 0.3 Rural poor 7.5 45.3 36.4 9.8 0.9 Rural poor 3.9 11.0 84.9 1.2 1.2 Urban-poor 19.7 49.9 20.8 8.4 Other Urban poor 25.3 21.4 52.1 1.6 Other Urban poor 16.5 49.3 27.0 5.7 0.5 Ndjamena poor 68.2 25.4 6 1.3 0.0 Ndjamena poor 29.7 52.0 16.9 0.3 Poor 7.3 12.2 80.2 Poor 8.8 45.8 34.8 8.9 1.7 1.8 Non poor 23.3 15.6 59.3 Non poor 17.8 42.4 27.4 6.1 6.4 1.3 Chad 14.7 43.5 29.9 7.0 4.8 Chad 17.8 14.4 66.5 0.0 20.0 40.0 60.0 80.0 100.0 0.0 20.0 40.0 60.0 80.0 100.0 Piped water Borehole Flushing toilet Latrine Unprotected water Surface water Rudimentary/open pit No toilet Other source Electricity and lighting 100.0 5.4 2.0 4.1 0.0 0.8 0.6 4.6 80.0 60.0 1.4 86.7 94.4 74.1 97.3 91.7 94.7 40.0 3.5 0.0 20.0 0.4 22.5 0.1 0.1 0.4 1.2 1.6 0.4 0.4 1.2 0.3 5.4 1.4 7.8 1.9 0.7 1.3 6.4 0.0 0.0 Chad Non poor Poor Ndjamena poor Other Urban Urban-poor Rural poor poor Electricity SNE Generator Solar Lamp Other Source: World Bank staff calculations using data from ECOSIT 4 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 39 Poor and nonpoor households differ greatly in terms of access to electricity outside the capital (Figure 1.12). In terms modern asset ownership. The ownership rate for modern of transportation assets, poor households are more likely assets, including computers, generators, refrigerators, fans, to own bicycles, and nonpoor households are more likely and televisions, is extremely low among poor households. to own motorbikes. Mobile phones are among the most Asset-ownership rates among poor households are higher popular household assets: 42 percent of poor households in N’Djamena than in rural areas or other urban centers, in rural areas and 91 percent of poor households in which is unsurprising given poor households’ low rates of N’Djamena report owning a mobile phone. Figure 1.12: Access to Basic Services at the Household Level (a) By poverty status (b) By location among the poor 45.1 42.3 Phone 65.8 Phone 60.9 91.3 58.6 12.9 12.3 radio 23.8 20.0 radio 15.1 24.7 14.2 14.3 bike 11.6 bike 10.1 14.5 12.5 3.1 2.6 motorbike 12.2 motorbike 6.3 10.9 9.1 0.3 0.0 tv 7.5 tv 0.5 10.9 5.0 0.3 0.0 parabole 6.8 parabole 0.3 4.6 9.1 0.2 0.0 ventilator 6.3 ventilator 0.3 4.2 9.3 0.1 0.0 refrigerator 2.6 refrigerator 0.0 3.1 1.8 0.3 0.3 generator 2.4 generator 1.8 0.7 1.7 0.0 0.0 Computer 2.1 Computer 0.0 1.4 0.0 0.0 0.0 car 2.0 car 0.0 1.3 0.8 0.1 0.0 Cooker 1.3 Cooker 0.0 0.9 3.2 0.0 20.0 40.0 60.0 80.0 0 20 40 60 80 100 Poor Rural poor Non poor Other Urban poor Chad Ndjamena poor Source: World Bank staff calculations using data from ECOSIT 4 40 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 1.4 POVERTY-REDUCTION CHALLENGES 1.4.1 THE SOCIAL SAFETY NET More effective targeting of social assistance programs is also necessary to address the country’s weak health indicators. In the ECOSIT 4, households were asked if Chad’s public social assistance programs are limited, and they participated in a healthcare program for pregnant they often fail to reach the poorest households. For example, women or in a program for children under five years old. many regions with high levels of food poverty receive less While households in almost all regions are covered by public assistance than regions with greater food security these two programs, beneficiary status does not seem to (Figure 1.13a). While social assistance programs may have be correlated with household poverty. For example, in the helped reinforce food security in the regions where they second-poorest region of Guéra, which has a poverty rate operate, the government should strengthen these programs of 60 percent and is home to 5.6 percent of the country’s and extend them to the poorest regions of the country, poor, less than 1 percent of households reported receiving particularly Mayo-Kebbi Est, Mayo-Kebbi Ouest, Mandoul, support from either program (Figure 1.13b). Moreover, Logone Oriental, Logone Occidental, and Tandjilé. Better to reduce the incidence of malaria, insecticide-treated targeting mechanisms could help improve the allocation of mosquito nets have been distributed to households in all support delivered via existing safety-net systems. regions, but the allocation of these nets does not appear to correlate with poverty status (Figure 1.13c). For example, in Mandoul, where 58 percent of the population is poor, only 11 percent of households have received an insecticide- treated mosquito net, while 38 percent of households have received nets in Chari-Baguima, where the poverty rate is less than half the level in Mandoul. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 41 Figure 1.13: Distribution of the Poor Population by Area of Residence and Region (a) Food Poverty and Food Assistance (b) Poverty Rates and Health Support Food Poverty Poverty rate 24.9 - 27.3 53.0 - 63.1 15.6 - 24.8 44.4 - 52.9 9.8 - 5.5 34.7 - 44-3 4.7 - 9.7 24.5 - 34.6 2.3 - 4.6 13.8 - 24.4 Food Assistance Health support 26.6 - 34.2 8.8 - 13.6 8.5 - 20.4 4.8 - 8.7 4.7 - 8.4 2.3 - 4.7 1.9 - 4.6 0.4 - 2.2 0.5 - 1.8 0.0 - 0.3 (c) Poverty Rates and Mosquito Net Distribution Poverty rate 53.0 - 63.1 44.4 - 52.9 34.7 - 44-3 24.5 - 34.6 13.8 - 24.4 Food Poverty 37.7 - 77.1 25.8 - 37.6 11.1 - 25.7 3.2 - 11.0 0.0 - 3.1 © Vemaps.com 42 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 1.4.2 FRAGILITY, CONFLICT, AND VIOLENCE high fatality rates. Border regions tend to have both high poverty rates and a large number of FCV-related fatalities. For example, in the Lac Region, where 44 percent of the Over the last five years, Chad has experienced numerous population is below the poverty line, about 385 FCV-related armed clashes, protests and riots, attacks against civilians, deaths were reported between 2018 and 2019 (Figure terrorist explosions, and other manifestations of fragility, 1.14b). The Lac Region is part of the Lake Chad area, where conflict, and violence (FCV). The number of deaths from attacks from extremist groups have resulted in the loss of FCV-related causes increased from 259 in 2018 to 567 in lives, the destruction of property, and massive population 2019, most due to armed conflict. Another 618 Chadians are displacement. The Ouaddaï Region, which borders Sudan, estimated to have died from FCV-related causes in 2020, has a poverty rate of 38 percent and reported 119 FCV- representing a 9 percent year-on-year increase (1.14a). related fatalities in 2018-19. Two of the country’s poorest regions are Tandjilé and Mayo-Kebbi Ouest, near the While all regions of Chad suffer from FCV-related fatalities, border with Cameroon; they reported 63 and 15 FCV-related some of the country’s poorer regions have especially fatalities, respectively, during 2018-19. Figure 1.14: Poverty Rates and FCV-Related Fatalities in Chad (a) Evolution of FCV-related fatalities 7 5 2020 488 30 88 26 2019 450 40 69 0 2 2018 212 8 37 2 0 1 2017 282 11 10 2 4 4 2016 22 0 2015 233 145 9 42 0 100 200 300 400 500 600 700 Battles Explosions/Remote violence Protests Riots Violence against civilians Source: Author’s graphic created using Acled et ECOSIT 4 data INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 43 1.4.3 FORCED DISPLACEMENT (b) Poverty Rates and Mosquito Net Distribution Chad has long been affected by forced displacement. The country currently hosts nearly 480,000 refugees and asylum seekers. Its largest refugee populations are from Poverty rate Number of Fatalities Sudan, Central African Republic (CAR), and Nigeria (UNHCR, 53.0 - 63.1 64 - 385 2020), and most have remained in Chad for over 15 years. 44.4 - 52.9 15 - 83 In addition, there are an estimated 240,000 internally displaced persons in Chad, as well as more than 30,000 34.7 - 44-3 5 - 14 Chadian former refugees who have recently returned to 24.5 - 34.6 3-4 Chad from neighboring countries. Forced displacement of 13.8 - 24.4 0-2 this magnitude has put enormous pressure on the fragile Chadian economy, while intensifying risks related to violent conflict and environmental degradation. 6 Most refugees reside in Chad’s border areas, often in isolated and impoverished regions with little infrastructure and few social services. Approximately 95 percent of the refugee population consists of Sudanese refugees located along the eastern border and CAR refugees located along the southern border. In eastern Chad, poverty rates range from 38 percent in Wadi-Fira to 42 percent in Sila, while poverty rates in southern Chad range from 52 percent in Logone Oriental to 58 percent in Mandoul (Figure 1.15). High levels of community- and household-level fragility, a lack of local development initiatives under the national development plan, a steep decline in humanitarian assistance, and widespread vulnerability to sexual and gender-based violence create profound challenges for both displaced populations and host communities. Source: Author’s graphic created using Acled et ECOSIT 4 data © Vemaps.com 6 OCHA, 2019. 44 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure 1.15 : Forced Displacement and Poverty Rates across Regions (a) Refugee camps along the Eastern and Southern border (b) Poverty Rates by region (%) Poverty rate 53.0 - 63.1 44.4 - 52.9 34.7 - 44-3 24.5 - 34.6 13.8 - 24.4 Source: UNHCR (2017) Source: World Bank staff calculations using data from ECOSIT 4 © Vemaps.com INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 45 Understanding the unique challenges faced by refugees Food insecurity poses a major threat to refugees in Chad, and host communities requires detailed data. Chad’s especially those from CAR. In 2018, approximately half ECOSIT 4 has been expanded to include a representative of Sudanese refugees and two-thirds of CAR refugees sample of Sudanese and CAR refugees and their host consumed fewer than 2,300 calories per day, the minimum communities. ECOSIT 4 captures key information on threshold for food security. By contrast, just 15 percent of the household poverty, food security, asset ownership, and Chadian population and about 40 percent of households in access to basic services. host communities are food insecure. Refugees also tend to have little dietary diversity, which reduces the nutritional ECOSIT 4 finds that roughly four out of five refugee quality of their already inadequate food intake. Persistent households are below Chad’s national poverty line, a far undernutrition poses serious risks to physical and cognitive higher rate than those of either the host communities or development, particularly among children. the general population. The poverty headcount rates for Sudanese and CAR refugees are estimated at 79.8 percent In addition to facing high levels of poverty and food and 83.7 percent, respectively. By contrast, the average insecurity, refugee households have limited access to poverty rate for host communities is about 70 percent, and productive assets. Fewer than 10 percent of CAR refugees the national poverty rate is 42.3 percent (Figure 1.16). The own livestock, and only 35 percent have access to land. high poverty rates among host communities underscore For Sudanese refugees, these figures are 30 percent and the preexisting deprivation, fragility, and lack of opportunity 50 percent, respectively. By contrast, over 60 percent of that characterize the regions in which most refugee households in Chadian host communities own livestock, groups are located. The COVID-19 pandemic has increased and 90 percent have access to land. Average herd sizes and the vulnerability of the refugee population, especially cultivated areas are far larger among host communities women, girls, and children, who are at high risk of sexual than among refugees. exploitation, and who may resort to prostitution or survival sex to cope with extreme poverty. Figure 1.16 : Poverty and food insecurity of refugee households Poverty rate Food insecurity 5.2 12.6 Chadian population Chadian population 15.2 42.3 14.8 28.6 Distant host communities Distant host communities 40.3 69.7 14 27.2 Nearby host communities Nearby host communities 38.1 71.4 15.9 32.1 Sudanese refugees 79.8 Sudanese refugees 47.4 29.8 45.9 CAR refugees CAR refugees 63.3 83.7 0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 Severity Depth Headcount Source: Author’s graphic created using Acled et ECOSIT 4 data 46 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 1.4.4 THE CHALLENGES POSED BY THE COVID-19 with highly uncertain income streams. Almost two-thirds of households have experienced a decline in their total PANDEMIC income due to the pandemic, and seven out of ten rural households have lost all or part of their income, increasing The COVID-19 pandemic has adversely affected public poverty rates and intensifying food insecurity (Figure 1.17b). health and living standards in Chad. As of November 23rd, the country had registered 1,642 COVID-19 cases and 101 Simulations using ECOSIT 4 high-frequency survey data and deaths, but the extent of the pandemic’s economic and the consumer price index suggest that the pandemic has social impact is unknown. The authorities undertook a caused the headcount poverty rate to rise by 5.5 percentage high-frequency survey to better monitor the pandemic’s points. This increase reflects an additional 849,574 people effects on the Chadian population, and the results of the falling below the poverty line due to the cumulative effect first two rounds show a loss of income, remittances, and of the loss of income, the decline in remittances, and the jobs, as well as diminished access to basic services and increase in prices associated with the pandemic. Household an increase in the price of essential goods. The pandemic consumption has dropped by an average of 10 percent and the collapse of global oil prices have contributed to nationally and by more than 20 percent in N’Djamena, a recession, which has primarily affected households increasing the poverty rate in the capital by 11 percent (Table with members working in the rural agricultural sector (48 1.6). Moreover, the intensity of poverty among the poorest percent) and in the urban informal sector, including trade households has increased, and many households in the (18 percent), services (6 percent), and transportation (7 intermediate deciles of the consumption distribution have percent). In Chad, many informal workers are day laborers fallen into poverty (Figure 1.17a). Table 1.6: Simulated Impact of COVID-19 at the Household Level Ndjamena Percentage Change in Consumption Percentage Change in Poverty Nominal Increase in the Poor Population Impact of reported decline in household income Ndjamena -18.9 9.63 143,321 Other urban -10.1 4.89 112,134 Rural -4.2 3.76 440,739 Chad -7.3 4.41 683,690 Impact of reported decline in domestic remittances Ndjamena -2.31 1.14 16,966 Other urban -3.94 2.66 60,997 Rural -0.32 0.48 56,265 Chad -1.37 0.88 136,428 Impact of consumer price inflation Ndjamena N/A 0.23 3,423 Other urban N/A 0.43 9,860 Rural N/A 0.74 86,741 Chad N/A 0.65 100,771 Response for households to the combine impact of decline in income, reduction in remittances and increase in prices Ndjamena -22.7 11.00 163,710 Other urban -15.6 7.40 169,691 Rural -6.1 4.50 527,480 National -10.2 5.48 849,574 Source: World Bank staff calculations using data from ECOSIT 4 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 47 Figure 1.17: The Impact of COVID-19 on the Chadian Population Distribution of the increase in the poor population caused by COVID-19 - 100 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.9 90 80 70 68.5 83.8 86.4 100 100 100 60 50 40 30 100 100 100 97.1 20 21.4 10 16.2 13.6 0.0 0.0 0.0 10.1 0.0 0.0 0.0 0.0 0.0 0 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 Structurally poor New poor Structurally rich During the last 30 days, share of households reporting that they: Spent an entire day without eating 31.9 27.1 Were hungry but did not eat 56.5 49.6 Ran out of food 55.5 53.4 Ate less you thought you should 82.6 78.6 Had to skip a meal 66.7 66.7 Ate the same kind of foods 78.8 77.6 Unable to eat healthy and nutritious foods 89.4 84.3 Worried about not having enough food 87.3 87.4 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 June-July August-September Source: World Bank staff calculations using data from ECOSIT 4 48 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Of the increase in poverty rates attributable to the 1.5 CONCLUSION COVID-19 pandemic and concurrent recession, 4.4 percentage points is due to the loss of household income. In the last fifteen years, Chad has achieved significant This factor alone caused the poor population to increase progress in reducing poverty and raising living standards, by 683,690 people. Two-thirds of the newly poor live in rural yet it continues to face enormous challenges. Monetary areas, where households have experienced a 4 percent and nonmonetary poverty rates have both declined decline in total consumption. The loss of household substantially, with gains observed across almost all income has been especially significant in N’Djamena, dimensions of the MPI. Meanwhile, Chad’s score on the where the poverty rate rose by 9.6 percentage points as Human Development Index (HDI) increased by about 35 143,321 people fell below the poverty line. Many vulnerable percent, from 0.298 in 2000 to 0.401 in 2018. Nevertheless, households in the capital city earn their livelihood from Chad remains among the poorest countries in the world. the informal sector, which has been disproportionately Despite considerable gains since 2000, Chad’s 2018 HDI affected by the pandemic, and informal workers are score was still well below both the average of 0.507 for low- especially vulnerable to poverty due to their lack of job income countries and 0.541 for Sub-Saharan Africa, and it security and limited savings. ranked 187th out of 189 countries and territories worldwide. The pandemic has reduced the amount and frequency of The unequal distribution of economic welfare constrains remittances in Chad. Data from the first round of the high- progress on poverty reduction and shared prosperity. frequency survey implemented in June-July 2020 shows that Almost 89 percent of Chad’s poor population lives in rural the amount of transfers fell by 57 percent and the frequency areas, while just 11 percent of the poor live in N’Djamena of transfers dropped by 61 percent. In 2018, remittances and other urban centers. Heads of poor households tend averaged FCFA 37,122 for non-poor households and FCFA to have little education, and most are farmers, unskilled 14,516 for poor households. While remittances represent a workers, or self-employed in the informal sector. Poor mere 5 percent of the income of poor households in Chad, households typically have little access to essential services, lower than 10 percent in Mali and 27 percent in Senegal, the own few modern assets, and live in overcrowded housing decline in remittances has led to 1.4 percent decline in total made of low-quality materials. All of these dimensions consumption and a 0.8 increase in the poverty rate. This of poverty are worse in rural areas, with the exception of corresponds to an absolute number of 136,428 additional overcrowding. The average monetary poverty rate in rural poor, 41 percent of whom live in rural areas. Chad is about 50 percent, and an estimated 15 percent of the rural population is food poor. Market disruptions due to COVID-19 increased prices, resulting in a 0.65-percentage-point increase in the Regional disparities are coupled with a highly unequal poverty rate. This reflects an increase in the poor distribution of economic welfare across households. population of 100,771 people, of whom 86,741 live in The wealthiest 20 percent of the population accounts for rural areas. Government measures to limit the spread of about 40 percent of total consumption, while the bottom the virus have led to shortages of essential goods and 20 percent accounts for just 8 percent. Gender inequality is contributed to the increase in prices, negatively impacting especially acute, and female-headed households are more poor and vulnerable households. Rising consumer prices likely to be poor than their male counterparts. In 2018, Chad have led to a 0.23-percentage-point increase in the ranked 160th out of 162 countries and territories on the poverty rate in N’Djamena, representing 3,423 additional Gender Inequality Index (GII). poor people, and a 0.43-percentage-point increase in the poverty rate in other urban areas, representing 9,860 additional poor people. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 49 Inadequate health and education services are a major on detrimental coping strategies—including child labor, the contributor to multidimensional poverty and pose a sale or consumption of productive assets in response to serious threat to long-term poverty reduction. Without shocks, and the use of fertility as an insurance mechanism substantial improvements in health and education, to compensate for child mortality and/or provide support especially in rural areas, many households will remain in old age—enabling households to achieve the modicum locked in a cycle of intergenerational poverty, as high of security necessary to invest in the future. While rural fertility rates and low levels of investment in human capital development offers a viable pathway to sustainable poverty formation sharply limit the productive opportunities of reduction, policymakers must alleviate binding constraints future generations. By extending health and education on income growth, including: (i) low levels of human capital, services into impoverished areas, the authorities can (ii) a vast infrastructure deficit, (iii) a lack of basic services, alleviate two key causes of deprivation while encouraging (iv) deep gender disparities, and (v) fragility and exposure poor households to invest in expanding their long-term to shocks. These constraints, along with strategies for productive potential. Strengthening social assistance addressing them, are discussed in the following chapters. programs can support this objective by lessening reliance 50 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION CHAPTER RURAL INCOME 02 ANALYSIS INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 51 Key Insights • Poverty in Chad is overwhelmingly concentrated in rural areas, where • Six key challenges inhibit rural income growth: (i) the infrastructure 89 percent of poor households are located. The urbanization rate is low gap; (ii) low levels of human capital; (iii) lack of complementary and stagnant. services; (iv) the gender gap; and (vi) shocks and fragility. • Most Chadian households rely on smallholder agriculture and • While the government continues to work closely with its international pastoralism, activities that are marked by low productivity and extreme partners to address the Boko Haram conflict, easing the risks of conflict vulnerability to shocks. nationwide will require broader security support and large-scale investments in economic infrastructure and ecological restoration. • Poor households cultivate more land, on average, than nonpoor households, but due to the larger size of poor households, the cultivated • Improving maternal health by reducing adolescent pregnancy and area per capita is significantly smaller. encouraging family planning could yield substantial benefits across generations, while expanding school feeding programs could boost • Male- and female-headed households cultivate similar amounts of land educational attainment while improving nutrition indicators. per capita, but average livestock holdings are much larger among male- headed households. • Targeted investments in roads and digital infrastructure could enable the expansion of cash crop production while enhancing access to • A pervasive lack of investment in improved inputs, mechanization, market information. and infrastructure keeps marginal agricultural productivity low and variable, while the livestock sector suffers from ecological vulnerability, • Fostering the development of farmer organizations can leverage limited uptake of veterinary medicine, and low levels of value addition. economies of scale in both input supply and output marketing while facilitating diversification and encouraging the adoption of new • Due to widespread food insecurity combined with inadequate physical technologies and production methods. and market infrastructure, commercial agriculture is extremely limited, and sales of crops and livestock account for only a small share of • Gender is a vital cross-cutting issue. Prioritizing maternal health, household income. encouraging girls to complete school before marriage, recruiting women into leadership roles, and equalizing access to land and • Livestock herds are often used as a store of value, especially among financial services could greatly enhance the productive potential of poor households with little access to financial services, and livestock women in rural Chad. sales tend to occur in response to economic distress rather than as a routine means of generating income. • Many households earn cash income by operating nonfarm enterprises, but these tend to be small, informal, and driven by necessity rather than opportunity. • Expanding the use of fertilizer, drought-resistant seed varieties, and other improved inputs, and promoting the adoption of new water-management techniques and other productive technologies could strengthen food security and enable the commercialization of agriculture; however, deficiencies in infrastructure, inadequate startup capital, and high levels of insecurity sharply constrain access to inputs and technologies. 52 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION In Chad, poverty is heavily concentrated in rural areas. percent of Chad’s poor population is located in rural areas, Rapid urbanization provides opportunities to accelerate fostering rural income growth will be vital to reduce poverty agricultural transformation, increase value addition, and and achieve shared prosperity. This chapter examines rural expand nonfarm income opportunities. However, Chad’s income sources, identifies opportunities for and constraints urbanization rate is very low compared to the average on accelerating rural income growth, and presents growth- for Sub-Saharan Africa (SSA) and has remained broadly oriented policy options. The analysis applies a version unchanged for 25 years. In 2017, the share of households of the World Bank’s rural income diagnostic framework living in urban areas was about 23 percent, well below the modified to distinguish between agricultural and nonfarm SSA average of 40 percent (Figure 2.1). As an estimated 89 labor income, as well as transfers. Figure 2.1. The Share of the Population Living in Urban Areas, Chad and SSA, 2017 50 40 30 20 10 0 1960 1963 1966 1969 1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 2014 2017 Chad Sub-Saharan Africa Source: WDI and World Bank staff calculation INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 53 The analysis presented in this chapter explores differences poverty rates (Figure 2.3). Therefore, the analysis identifies in livelihoods across Chad’s diverse agro-ecological five AEZs: Saharan, North Sahelian, South Sahelian, North zones (AEZs). Chad’s three main AEZs are the Saharan, Soudanian, and South Soudanian.e also variations between Sahelian, and Soudanian zones (Figure 2.2). Annual rainfall the northern and southern areas of the Sahelian and varies significantly across these zones, ranging from less Soudanian AEZs, especially in terms of crop cultivation and than 200mm in the Saharan zone to up to 1,000mm in poverty rates (Figure 2.3). Therefore, the analysis identifies the Soudanian zone. There are also variations between five AEZs: Saharan, North Sahelian, South Sahelian, North the northern and southern areas of the Sahelian and Soudanian, and South Soudanian. Soudanian AEZs, especially in terms of crop cultivation and Figure 2.2 Chad’s Agro-Ecological Zones Figure 2.3 Rural household Income Composition (%) 100 5.3 7.5 1.8 8.9 90 2.2 2 18.4 80 20.5 21.9 70 6 6.7 60 7.2 Saharian 50 40 65.1 54 58.4 30 20 10 Sahelian 3.4 5.4 4.9 0 Wage Non-poor National Soudanian Wage Enterprise profit Agriculture Remittances Livestock Other Source: World Bank staff calculations using data from ECOSIT 4 Source: World Bank staff calculations using data from ECOSIT 4 54 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Table 2.1 Poverty, Agricultural, and Income Statistics by Agro-Ecological Zone Saharan North South North South Sahelian Sahelian Soudanian Soudanian Poverty rates (national poverty line) 0.26 0.43 0.36 0.63 0.55 Total cultivated (ha) 0.72 1.30 1.79 1.72 2.13 Share of land cultivated to millet 0.56 0.82 0.51 0.22 0.18 Share of land cultivated to sorghum 0.00 0.12 0.21 0.32 0.31 Share of land cultivated to rice 0.00 0.00 0.00 0.16 0.07 Share of land cultivated to maize 0.00 0.03 0.06 0.07 0.04 Share of land cultivated to groundnut 0.00 0.01 0.13 0.13 0.22 Share of land cultivated to sesame 0.00 0.00 0.02 0.05 0.10 Share of land cultivated to cowpea 0.00 0.01 0.01 0.02 0.02 Share of land cultivated to other crops 0.44 0.01 0.04 0.05 0.06 Share of agricultural income in the total income 0.04 0.43 0.64 0.67 0.61 Share of livestock income in the total income 0.42 0.09 0.06 0.05 0.04 Share of wages in the total income 0.17 0.05 0.05 0.04 0.05 Share of enterprise income in the total income 0.16 0.24 0.17 0.17 0.24 Share of remittances in the total income 0.03 0.02 0.02 0.02 0.01 Share of other incomes in the total income 0.18 0.17 0.06 0.04 0.04 Median cash income from cash crops (FCFA) NA81 18,667 40,000 39,000 33,750 Median cash income from cereal crops (FCFA) NA 20,000 36,000 30,000 20,000 Median cash income from livestock 450,000 52,500 54,000 25,500 21,250 % of households with wage labor income 13.2 2.8 3.1 1.5 3.5 % of households at >=60 minutes to nearest permanent market NA 54.0 92.1 19.1 78.0 Source: World Bank staff calculations using data from ECOSIT 4 1 8 Not computed due to no observations (zero plots), as crop cultivation is not widely practiced in the zone. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 55 2.1 INCOME AND PRODUCTIVE RESOURCES 2.1.1 AGRICULTURE Unequal land access between poor and nonpoor households reflect differences in average household size. On average, nonpoor households cultivate 0.56 Crops and livestock are the primary sources of income for hectares (ha) per household member, while poor rural households in Chad. The agricultural sector represents households cultivate just 0.36 ha per household member 65 percent of total income, of which crop production despite having larger average landholdings. Although accounts for 58.4 percent, while livestock accounts for 6.7 male-headed households cultivate much more land percent (Figure 2.3). In Sub-Saharan Africa, the agricultural than female-headed households (an average of 2.23 ha sector represents an average of 68 percent of rural income, versus 1.18 ha), female-headed households tend to be only slightly higher than the sector’s share in Chad’s much small, and there is no significant difference in land total income. Moreover, agriculture’s contribution to rural cultivation on a per capita basis. income is much higher for households in the bottom 40 percent of the consumption distribution. Among these households, crops and livestock account for 71 percent The livestock sector is characterized by small herd sizes. of the total income, whereas they represent 61 percent of The national average herd size is just 2.4 tropical livestock income among households in the top 60 percent of the units (TLUs), but the average herd size in the Saharan AEZ, consumption distribution. where pastoralism is the main source of livelihoods, is close to 19 TLUs. Across regions, medium-size livestock such as goats and sheep are the most common. Broadly equal More than eight out of ten households in rural Chad shares of poor and nonpoor households engage in livestock depend on agriculture and related activities. About 88 and production, and their herd sizes are similar. However, there 62 percent of rural households are engaged in crop and is a large gender gap in livestock holdings, as the average livestock production, respectively. Livestock herding is by far herd size for male-headed households is approximately 3.3 the most important economic activity in the Saharan zone, times larger than that of female-headed households. This where only about 10 percent of households are engaged difference is even larger than that observed in Mali, where in crop production. 8Agro-pastoralism is also important the average herd size of male-head households is 3 times outside the Saharan zone: for example, more than 50 that of female-headed households.10 percent of households in the Soudanian zone engage in both crop and livestock production. While the combination of crop and livestock production provides opportunities The allocation of farmland in Chad is driven by a focus for agricultural income diversification, households in the on food security over revenue maximization. On average, Soudanian zone are more likely to be poor than their Chadian farmers devote almost 70 percent of their counterparts in the Saharan zone—consistent with patterns land to cereal production. The staple crops millet and observed in comparable areas such as the Sikasso region sorghum represent 32 percent and 25 percent of total in Mali.9 cereal production, respectively, while groundnuts are the 8 The Household Consumption and Informal Sector Survey in Chad (2018-2019) – ECOSIT4. 9 Beegle, K., & Christiaensen, L. (Eds.). (2019). Accelerating poverty reduction in Africa. World Bank Publications 10 Staff calculation using the ECOSIT4. 56 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION dominant cash crop, accounting for 16 percent of total south Sahelian zone is somewhat more diversified, with cash-crop production (Figure 2.4). Cropping patterns about 70 percent of land allocated to millet and sorghum among poor and nonpoor households are similar in and 14 percent to peanuts. The Soudanian zone is the the Saharan and Sahelian zones, but in the Soudanian most diversified, and it is the only zone with significant zone nonpoor households are significantly more likely differences between the types of crops cultivated by to cultivate higher-value crops. Likewise, both female- nonpoor and poor households. For example, high-value and male-headed households tend to have similar crop cereal crops such as maize and rice occupy a much larger portfolios. However, the amounts of land allocated to share of land owned by nonpoor households, while poor producing the country’s main crops vary widely across households dedicate more land to low-value staple crops AEZs. The Saharan and north Sahelian zones are the least such as millet and sorghum (Figure 2.5). diversified, with 80 percent of land allocated to millet. The Figure 2.4. Nationwide Allocation of Farmland by Crop Type 32.5 30 25 24.7 20 16.3 15 10 7.5 6.3 6.1 5 4.2 2.4 0 Millet Sorghun Maize Paddy rice Peanut Sesame CowpeaO thers Source: World Bank staff calculations using data from ECOSIT 4 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 57 Figure 2.5. Allocation of Farmland in the Soudanian AEZ by Poverty Status 6.4 6.0 7.3 8.6 22.8 25.3 36.9 35.2 34.2 23.3 11.4 6.4 45.4 44.3 49.8 36.6 Non-poor Poor Non-poor Poor North Soudanian South Soudanian Others Cash crops High-value cereals Staple cereals (maize & rice) (milet & sorghum) Source: World Bank staff calculations using data from ECOSIT 4 The limited use of improved inputs and production Farmers in Chad lack access to the necessary tools and practices contributes to low levels of marginal inputs to increase productivity and crop yields. Crop productivity. Millet and sorghum, the main staple crops, production suffers from low rates of mechanization and have the lowest yields among cereal grains, while sesame the limited use of improved inputs. Less than 0.5 percent and cowpeas are the cash crops with the lowest yields. of households report owning a tractor, and about 3 Poor households have lower yields than their wealthier percent report using one on their farm. 12The average counterparts across all crop types, except millet and cultivated plot size is under 1 hectare, which, combined cowpeas (Figure 2.6). Millet and cowpeas are drought- with the indivisibility of agricultural machinery and lack resistant, which likely contributes to their popularity of developed agricultural equipment rental markets, among poor households. 11Low yields in the agriculture limits mechanization. 13The use of variable inputs such sector, especially for staple crops, intensifies food as fertilizer is also extremely low in Chad compared to its insecurity: about 75 percent of rural households reported neighbor Mali. Between 1990 and 2014, the average amount having had to skip a meal at least once in 2018 due to a of synthetic fertilizer used in Mali was approximately lack of food. 80,640 metric tons, versus just 9,180 metric tons in Chad. 14 Inadequate infrastructure is also a major challenge, and low levels of rainfall in a context of very limited irrigation systems further reduced yields during the 2017- 18 agricultural season. 11 Fischer, H. W., Reddy, N. N., & Rao, M. S. (2016). Can more drought-resistant crops promote more climate secure agriculture? Prospects and challenges of millet cultivation in Ananthapur, Andhra Pradesh. World Development Perspectives, 2, 5-10. 12 This difference between ownership and use indicates some rental activities of agricultural equipment although such activities are limited. 13 Beegle, K., & Christiaensen, L. (Eds.). (2019). Accelerating poverty reduction in Africa. World Bank Publications 14 Fuglie, Keith O. 2015. “Accounting for Growth in Global Agriculture,” Bio-based and Applied Economics 4(3): 221-54. 58 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure 2.6. Differences in Crop Yields between Poor and Nonpoor Households 2000 1657 1501 1500 1300 1217 1000 934 866 780 738 751 652 500 348 410 363 227 0 Sorghum Millet Paddy Rice Maize Peanut Cowpea Sesame Non-poor Poor Source: World Bank staff calculations using data from ECOSIT 4 Like crop production, the livestock sector is dominated by that stretches across the border into the Central African traditional practices. Maintaining livestock requires access Republic is used to feed livestock during the dry season, to natural resources such as pastureland and water, the but this area is becoming increasingly difficult to access availability of which varies during the year and across AEZs. due to mounting insecurity. 17Uptake of veterinary medicine Water sources are highly vulnerable to drought, and the is limited, with only about 15 percent and 25 percent of limited availability of water restricts access to pastureland, households using deworming and vaccination services, especially during the dry season.15 Pastureland is especially respectively. 18The livestock sector is also characterized by critical in Chad, where only 20 percent of households report very little value addition, as most sales are of live animals buying animal fodder.16 The transhumance corridor rather than animal products. 15 Mahamat Guindé. 2018. Ouagal Mahamat Mahamat Abdallah. OIE – World Organization for Animal Health. 16 Staff calculations using ECOSIT4 17 FAO (2020). https://reliefweb.int/report/chad/strengthening-social-cohesion-among-communities-central-african-republic-and-chad 18 Author’s calculations using ECOSIT4 data. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 59 2.1.2 AGRICULTURAL MARKET PARTICIPATION production is especially low for millet (35 percent) and sorghum (45 percent). The uni-seasonal nature of crop production in a context of limited irrigation prevents Rural households derive little cash income from agriculture. households from commercializing their agricultural While the overwhelming majority of rural households activities. Fewer than 2 percent of the country’s cultivated engage in crop and livestock production, these activities plots were irrigated in 2018, and irrigation levels are just 2.5 generate only a small share of their cash income (Figure percent in the Hadjer Lamis and Lac Regions, despite their 2.7). Even in the north Soudanian AEZ, where households access to the Lake Chad Basin. receive an especially large share of their income in cash, average total household income from crop sales averages Income from selling livestock and animal products is just FCFA 204,216—well below the individual poverty line of similarly low. Approximately 28 percent of households FCFA 242,094. receive income from the sale of livestock. In rural areas, livestock is usually used as both a productive asset and a Low marginal productivity, a focus on subsistence store of value, especially among poor households, which over commercialization, and undiversified production often lack access to alternative savings options. Due to structures reduce the cash income of agricultural low livestock productivity, most livestock sales occur households. Over 80 percent of cereal output in Chad is under distress rather than to generate cash income. Low for consumption by the households that produced it.19 Even urbanization rates and poor access to processing and among the modest share of households that participate cooling infrastructure also limit income opportunities from in commercial agriculture, only about half of their total the sale of meat and dairy products. production is sold, and the proportion of commercialized Figure 2.7. Total Cash Income from Crop Sales (in FCFA) 300,000 250,000 242,094 204,216 200,000 168,679 150,000 120,031 100,000 50,000 48,380 - North Sahelian South Sahelian North Soudanian South Soudanian National poverty line Source: World Bank staff calculations using data from ECOSIT 4 19 This statistic includes households that do not sell any share of their output. 60 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 2.1.3 THE NONFARM SECTOR NFE ownership, and there is no difference in NFE ownership between male- and female-headed households. Rural households derive about one-fifth of their total The most common NFEs in rural areas are focused on income from nonfarm enterprises. The contribution of retail trade (33 percent), food production (24 percent), and nonfarm activities to total income ranges from 24 percent personal services (12 percent) (Figure 2.9). The popularity among rural households in the south Sahelian AEZ to just of these activities likely reflects their low startup costs20. 15.6 percent among households in the Saharan AEZ. Though NFEs are mainly informal: nationwide, less than 6 percent nonfarm activities are a substantial source of income in are formally registered, and this share falls to less than 1 rural areas, the average urban household earns almost 40 percent in rural areas. As in other low-income countries, percent of its total income from nonfarm activities. Nonfarm many NFEs in Chad are started out of necessity rather than income represents 18 percent of total income among as a response to business opportunities, and they tend to households in the bottom 40 percent of the consumption contribute little to job creation.21 distribution and 21 percent among households in the top 60 percent. About 36 percent of households in rural Chad own and operate at least one nonfarm enterprise (NFE). While Figure 2.9 Nonfarm Enterprises by Location similar shares of poor and nonpoor households own and and Head of Household operate NFEs, 45 percent of households in the top 20 percent of the consumption distribution derive income from an 100% 8.9 NFE, versus just 36 percent of households in the bottom 20 13.7 percent of the distribution (Figure 2.8). The education level 17.8 17.1 of the household head is positively correlated with 4.3 80% 11.8 7.6 10.6 60% Figure 2.8 Share of Households that Own Nonfarm 24.2 48 Enterprise by Household Characteristics 35.2 26.1 Share of Households that Own a Nonfamr Enterprise 40% by Household Characteristics 60 55.8 20% 49.4 33.4 29.4 50 26.6 26.8 40.7 42.9 40 36.9 37.5 0% Rural Urban Female Male HHHs 30 20 Retail Wholesale Foods and drinks Personal services 10 Carpenty, sewing Construction 0 Rural Urban Female Male Head with Head Other HHHs HHHs formal without education formal Source: World Bank staff calculations using data from ECOSIT 4 education Source: World Bank staff calculations using data from ECOSIT 4 20 Nagler, P., & Naudé, W. (2017). Non-farm entrepreneurship in rural sub-Saharan Africa: New empirical evidence. Food policy, 67, 175-191. 21 Poschke, M. (2013). ‘Entrepreneurs out of necessity’: a snapshot. Applied Economics Letters, 20(7), 658-663. Beegle, K., & Christiaensen, L. (Eds.). (2019). Accelerating poverty reduction in Africa. World Bank Publications. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 61 Over 90 percent of NFEs in Chad exclusively employ family The concentration of NFEs in a narrow range of sectors with labor, and most employ just one family member, though limited market opportunities leads to intense competition there are differences between rural and urban areas (Figure and high entry and exit rates. About half of NFEs report weak 2.10). Urban NFEs are more likely to hire non-family labor (16 demand as a constraint on their operations that threatens percent) than their rural counterparts (7 percent). Firms that their solvency,23 and only about half of NFEs operate year- hire labor tend to employ more than one non-family worker. round. Over 40 percent of NFEs in the retail trade and food NFEs in the construction and personal services sectors and drinks sectors in both rural and urban areas have been are the most likely to hire non-family labor, while NFEs in operating for fewer than five years, and their high entry and sectors such as food and drinks and retail are the least likely exit rates likely reflect a lack of profitability, financing, and to employ non-family labor. The lower rate of non-family mechanisms to cope with idiosyncratic shocks.24 NFEs that workers in restaurants and retail stores may reflect the less rely on skilled or semiskilled labor tend to be more stable. skill-intensive nature of these businesses, as well as their For example, NFEs involved in carpentry and sewing tend to small size and typically informal status (Figure 2.11).22 remain active for 12 years or more in both rural and urban areas, and urban NFEs in construction and personal services are also likely to remain in operation for 12 years or more. Figure 2.10. Number of Workers in Nonfarm Enterprises by Type of Labor 5.4 4.6 5.4 4.9 10.8 12.2 10.8 11.7 17.1 30.2 34.8 29.6 83.8 83.2 83.8 83.4 48.1 40.2 Urban Rural Urban Rural Urban Rural Any type of labor Family Hired 1 worker 2 workers 3 or more workers Source: World Bank staff calculations using data from ECOSIT 4 22 Nagler, P., & Naudé, W. (2017). Non-farm Entrepreneurship in Rural Sub-Saharan Africa: New Empirical Evidence. Food Policy, 67, 175-191. 23 ECOSIT4. 24 Nagler, P., & Naudé, W. (2017). Non-farm entrepreneurship in rural sub-Saharan Africa: New empirical evidence. Food policy, 67, 175-191. 62 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure 2.11. Share of Nonfarm Enterprises that Employ Formal Workers by Sector 80 70 68% 60 50 46% 40 30 24% 22% 20 12% 12% 13% 10 6% 6% 9% 6% 1% 3% 2% 0 ail e s nks g n er ice sal win ctio Oth Ret dri erv ole , se stru d& Wh al s try Con Foo son pen Per Car Urban Rural Source: World Bank staff calculations using data from ECOSIT 4 Figure 2.12. Age Distribution across Nonfarm Enterprises 100% 20 24 34 16 34 30 21 16 16 23 19 27 14 15 80% 12 15 14 14 16 16 15 27 21 60% 21 23 13 23 21 22 17 25 24 27 24 20 38 40% 25 22 24 29 25 31 20% 43 29 28 49 24 23 40 45 38 32 44 21 35 42 0% Personal services Food & drinks Carpentry, sewing… Personal services Food & drinks Carpentry, sewing… Retail Wholesale Others Retail Wholesale Others Construction Construction Urban Rural <=4 years 5 to 8 years 9 to 12 years >12 years Source: World Bank staff calculations using data from ECOSIT 4 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 63 2.1.4 NONLABOR INCOME: REMITTANCES The average share of remittances in household income varies based on household poverty status and the gender of the household head. Remittances represent 3 percent of Remittances play a limited role in rural household income. household income among poor households and 7 percent The contribution of remittances to total household income among nonpoor households. Female-headed households ranges from 2 percent in rural areas to 6.3 percent in urban are almost twice as likely to receive remittances as male- centers. Chad’s net migration rate has remained low and headed households (34 percent versus 18 percent), and stable over time (Figure 2.13), and the population of Chadian they receive about twice the level of remittances as their migrants is much smaller than those of most regional peers male counterparts (CFAF 45,125 versus CFAF 23,267). Chad’s (Figure 2.14 ). In Mali and Senegal, remittances finance a adoption of the Global Compact for Migration represents an significant share of consumption, but in Chad they represent opportunity to boost nonlabor income through remittances, just 5 percent of household income. However, the share of which have been shown to increase the probability of small- remittances in household income varies widely in Chad, scale self-employment and could catalyze the development from 3 percent in the north and south Soudanian AEZs to 12 of NFEs. percent in the north Sahelian AEZ. Figure 2.13. Age Distribution across Nonfarm Enterprises Figure 2.14 Migrants as a Share of the Population (%) 10 8.0 0 1962 1967 1972 1977 1982 1987 1992 1997 2002 2007 2012 2017 5.4 -10 3.6 3.8 -20 1.6 -30 Burkina Chad Mali Niger Senegal -40 Chad SSA Faso Source: United Nations Population Division. World Population Prospects: 2019 Revision. Source: KNOMAD 2019. 64 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 2.2 RURAL INCOME GROWTH: OPPORTUNITIES AND CONSTRAINTS 2.2.1 OPPORTUNITIES perceptions are correlated with differences in observed yields, especially for millet, sorghum, and paddy rice (Figure 2.16), and they may impact crop choices and investment Efforts to increase the productivity of staple crops could decisions.25 Marginal yields could be significantly increased enhance food security and facilitate commercialization. by enhancing soil quality through improved water Measures to promote the use of improved inputs such as management, the adoption of soil-conservation techniques, fertilizers and drought-resistant seeds and to encourage the introduction of complementary inputs, and the use of the adoption of new productive technologies could greatly climate-smart agricultural technologies to increase labor increase marginal productivity. Households perceive about productivity. Greater productivity could generate a surplus one-third of their plots to be of poor quality, with the of staple crops, which in turn could increase the rate of largest share of poor-quality plots reported in the south commercialization. Soudanian AEZ (Figure 2.15). Though subjective, these Figure 2.15 Perceived Soil Quality (% of plots) Figure 2.16 Differences in Yield between Plots Rated “Good Quality” and “Poor Quality” (percentage points) 17.6 16.1 13.5 11.9 Sesame 62% 26.6 Groundnut -4% 50.4 53.7 Cowpea 38% 51.3 62.7 45.8 Maize 30% Paddy rice** 49% 31.2 36.1 34.3 Sorghum** 49% 21.2 27.7 Millet*** 30% National North South North South Sahelian Sahelian Soudanian Soudanian -10% 0% 10% 20% 30% 50% 60% 70% Good Medium Poor Source: World Bank staff calculation using data from ECOSIT 4 Source: World Bank staff calculation using data from ECOSIT 4 Note: ** and *** indicate statistical significance at the 5 percent and 1 percent confidence levels, respectively. 25 Yield differences between plots with “poor” and “good” soil quality are statistically different from zero. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 65 Increasing the area under cultivation could promote high- with large yield gaps, underscores the agriculture value crop production. Although Chad’s total agricultural sector’s potential to generate income growth among poor land area exceeds 49 million hectares, only about 6 percent households, especially in the Sahelian and Soudanian AEZs. of the country’s arable land is cultivated, and more than In addition to boosting farm income, diversification into 80 percent of farms cultivate fewer than 2 hectares.26 cash crops can also benefit the production of staple crops Increased production through area expansion would need by easing liquidity constraints and allowing for greater to be accompanied by sustainable cultivation practices and investment in improved inputs and new technologies.28 infrastructure development to enable access to land. The resulting surpluses could encourage further commercialization, providing farmers with greater liquidity and enabling further investment in a virtuous cycle. Expanding irrigation networks could help to mitigate vulnerability to drought, increase diversification into cash crops, and provide opportunities to generate The prevalence of agro-pastoralism in the southern AEZs income during the off season. Despite its considerable offers a unique opportunity to develop complementary potential, irrigation in Chad remains underdeveloped and and inclusive value chains for crops and livestock. For underutilized. Based on data from 2002, only 9 percent of example, the low domestic production and high income the country’s available water resources are being used, and elasticity of rice indicate that there is scope to increase less than 1 percent of its agricultural land is irrigated.27 productivity and promote inclusive value chains (Figure 2.17). Furthermore, the low use of both preventive and curative veterinary services indicates the potential to Diversifying into cash crops and raising marginal increase income from livestock by improving animal productivity could increase the rate of commercialization. health. The development of agricultural value chains would A full 80 percent of Chadian households cite the need to also generate more NFE opportunities in areas where produce food for their own consumption as a key obstacle agricultural businesses are rare. to commercialization. However, the high commercialization rate for sesame, over 50 percent of which is sold, combined Figure 2.17. Income Elasticity of Food Demand in Urban Areas 3.0 2.4 2.5 2.1 1.5 0.9 1.0 0.9 0.4 Sorghum Rice Millet Maize Wheat Meat Dairy Fish & Fruits & Beverages seafood vegetables -1.4 Source: World Bank staff calculations using data from ECOSIT 4 26 Masters, W. A., Djurfeldt, A. A., De Haan, C., Hazell, P., Jayne, T., Jirström, M., & Reardon, T. (2013). Urbanization and farm size in Asia and Africa: Implications for food security and agricultural research. Global Food Security, 2(3), 156-165 27 FAO. AQUASTAT geospatial information. http://www.fao.org/aquastat/en/geospatial-information/global-maps-irrigated-areas/irrigation-by-country/ country/TCD 28 Beegle, K., & Christiaensen, L. (Eds.). (2019). Accelerating poverty reduction in Africa. World Bank Publications. Govereh, J., & Jayne, T. S. (2003). Cash cropping and food crop productivity: synergies or trade-offs? Agricultural economics, 28(1), 39-50. 66 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Emerging international demand for non-timber forest of infrastructure. In 2018, Chad’s population density was products (NTFPs) such as gum arabic, shea, and oasis tree estimated at just 12 people per square kilometer, compared to fruits (e.g., dates) could accelerate rural income growth. 18 in Niger and 72 in Burkina Faso. The difficulty of connecting Chad is the world’s second-largest producer of gum arabic, small, remote communities limits market opportunities in which is the country’s fourth largest export product. Gum agriculture and other sectors. Just 3.2 percent of households arabic is mainly produced in the central part of the Sahelian in Chad regularly use asphalt roads, and only 3.4 percent AEZ and is the leading commodity in the NTFP sector and are within reach of a permanent market.30 In addition, lack has been identified as a resilient income source for rural of storage capacity and poor access to electricity limit households along with other NTFPs such as shea, which is the capacity for processing agricultural goods, forcing primarily produced in the Soudanian AEZ. The development households to sell their crops and livestock with little or of NTFP value chains could support inclusive income growth no value added. Estimates based on ECOSIT 4 data indicate for vulnerable groups such as poor and female-headed that 25 percent of households that sell livestock receive cash households, which often have limited access to land. income from selling live animals, while less than 3 percent receive cash income from selling meat. The availability of date palms in the Saharan AEZ represents an important opportunity to accelerate income The authorities can address these infrastructure constraints growth and poverty reduction. In 1990, 67 percent of the by developing integrated input-output markets capable of estimated 2,010,000 date palms in Sub-Sahelian countries connecting farmers to urban and peri-urban commercial were located in Chad, primarily in the Borkou, Ennedi, and centers while creating an enabling environment for the Tibesti regions.29 Dates are mainly grown for commercial adoption of new technologies.31 The expansion of contract sale and are highly marketable, both domestically and farming could increase access to inputs while generating internationally. Date cultivation is viable in arid regions, cash income. However, these arrangements can entail where it offers households a valuable opportunity to significant risks for farmers, especially for the country’s generate cash income and exit poverty. poorest households, and a lack of technical knowledge might impede their ability to meet quality requirements. Furthermore, long distances between farming households 2.2.2 KEY CONSTRAINTS ON RURAL INCOME GROWTH increase transaction costs, including search costs, which complicates the development of value chains. Six key challenges constrain rural income growth. These include: (i) the infrastructure gap; (ii) low levels of human Low Levels of Human Capital capital; (iii) lack of complementary services; (iv) the gender gap; and (vi) shocks and fragility. Low educational outcomes and a lack of technical knowledge discourage the uptake of new technologies The Infrastructure Gap and may limit the effectiveness of agricultural inputs and investments.32 Approximately 57 percent of male plot The poor state of Chad’s infrastructure deters private managers and 81 percent of female plot managers have investment, inhibits commercialization, and limits no formal education, and less than 2 percent of all plot connectivity to input and output markets. Chad’s vast land managers have any education beyond the primary level. area and dispersed population raise the marginal cost Low educational achievement contributes to poor technical expertise, which constrains technological adoption and 29 Salah, M. B. (2015). Date Palm Status and Perspective in Sub-Sahelian African Countries: Burkina Faso, Chad, Ethiopia, Mali, Senegal, and Somalia. In Date Palm genetic resources and utilization (pp. 369-386). Springer, Dordrecht. 30 Ecosit 4. 31 Ton, G., Desiere, S., Vellema, W., Weituschat, S., & D’haese, M. (2017). “The effectiveness of contract farming in improving smallholder income and food security in low- and middle-income countries: a mixed-method systematic review.” 3ie Systematic Review. 32 Results from a random field experiment in Niger show that training farmers in water management and soil conservation techniques increased uptake by up to 60 percentage points. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 67 diversification into high-value crops such as fruits and The poor state of the healthcare system further vegetables. Compared to traditional crops, high-value crops undermines human capital and deepens the impact are often especially vulnerable to weather-related shocks, of idiosyncratic health shocks on labor income and and managing those shocks requires technical knowledge household productivity. The public healthcare system, that may not be present in local communities. Additionally, which is characterized by poor accessibility and quality, is limited technical skills may prevent households from the main source of care for the country’s poor population. accessing income opportunities from NTFPs such as Rural households in remote areas are especially affected gum arabic and shea. In the gum arabic subsector, for by inadequate healthcare infrastructure. Combined with instance, the current cultivation method is mainly based liquidity constraints that reduce access to hired labor, a lack on propagation, which yields low-quality gum destined for of health services places the rural poor at especially high a narrow range of export markets.33 risk of income loss in the event of a health shock. According to ECOSIT 4, up to 50 percent of rural respondents report More effective use of extension agents could build having a sick member who did not seek medical services agricultural knowledge. However, the absence of reliable even though their condition prevented them from working. communication mechanisms between agents and farmers impedes the effectiveness of extension services. Lack of Complementary Services Furthermore, extension initiatives may be ineffective if they ignore context-specific variables (e.g., different soil types) Limited insurance markets, inadequate public investment or do not fully address the needs of farmers (e.g., utility in information and communications technology (ICT) maximization versus profit maximization). The country’s infrastructure, and weak land rights restrict the ability low level of human capital is also reflected in the poor of the rural population to access market opportunities quality of extension services. and invest in productive activities. The extension of mobile coverage has been shown to significantly reduce Low levels of human capital also contribute to the absence transaction costs and price dispersion in rural areas in of farmer organizations, which are critical to successful low- and lower-middle-income countries.34 However, only value chains. The lack of farmer organizations prevents about 58 percent of Chadians own a cellphone, and this the formation of economies of scale and limits the use of share drops to 46 percent among poor households. Mobile- digital technologies and financial services, which are vital to money services are also limited, with a penetration rate of connect farmers to markets. Results from an ordinary least only 16 percent, versus 24 percent in Mali and 33 percent squares regression of the correlates of selling a share of in Burkina Faso.35 The rate of active users of mobile money output suggest that the presence of a farmer organization is likely very low, especially in rural areas, where mobile- has a positive and significant effect on sales. Farmers with money agents are extremely rare. In Chad, over 80 percent a permanent market located less than an hour from their of NFEs use personal funds as startup capital, underscoring farms sell a greater share of their output than famers who their limited access to credit. must travel for over an hour to reach a permanent market. By contrast, access to a periodic market does not appear to As in other low-income countries, the absence of crop increase the share of output sold, indicating that proximity or livestock insurance in Chad heightens the production to a permanent market is a key determinant of agricultural uncertainty arising from weather-related shocks. The lack commercialization. Access to an asphalt road, even if it is of insurance discourages farmers from making productive more than an hour away from the farm, also matters more investments, taking risks on new crop types and production than access to a nearby laterite road. models, or accessing new market opportunities. While 33 Rahim, A. H., Ierland, E. C. V., & Weikard, H. P. (2010). Competition in the gum arabic market: a game theoretic modelling approach. Quarterly Journal of International Agriculture, 49(892-2016-65207). 34 Aker, J. C. (2010). Information from markets near and far: Mobile phones and agricultural markets in Niger. American Economic Journal: Applied Economics, 2(3), 46-59. 35 Demirgüç-Kunt, Asli, Leora Klapper, Dorothe Singer, Saniya Ansar, and Jake Hess. 2018. The Global Findex Database 2017: Measuring Financial Inclusion and the Fintech Revolution. Washington, DC: World Bank. doi:10.1596/978-1-4648-1259-0. License: Creative Commons Attribution CC BY 3.0 IGO 68 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION there is evidence that increased tenure security supports options, increase their risk of poverty, and contribute to a long-term investment,36 Chad’s community-based land range of deeply negative health consequences, including systems and overlapping land rights, weak tenure security, early death. In some communities, social norms limit the and costly land certification procedures create uncertainty free movement of women, thereby constraining their access around long-term ownership and access rights.37 The latest to economic opportunities.40 household survey shows that fewer than 3 percent of the country’s cultivated plots have formal titles, and parallel Gender-related inequalities also extend to access to customary and formal land-tenure systems exacerbate tenure security, financial services, and knowledge. In uncertainty and discourage investment in fixed capital.38 In Chad, 11 percent of women over the age of 15 have a mobile- rural areas, informal customary tenure systems, including money account, compared to 20 percent of men, and only Islamic land rights, are the dominant model, and obtaining 5 percent of women have a financial account, compared to formal land titles is often a lengthy, complex, and costly 13 percent of men (Figure 2.18). The gender gap in financial process. A new land code that would address land titling access in Chad is much wider than in comparable countries and property issues was drafted in 2014 but has yet to be such as Mali. Moreover, most extension agents and lead adopted. Meanwhile, there are currently no laws governing farmers are male,41 which increases information frictions the grazing rights of itinerant pastoralists or mechanisms and tends to exclude women from agricultural programs, for resolving land disputes involving them, which heightens even those that are specifically designed to reach them.42 tensions between livestock owners and farmers. In August 2019, a state of emergency was declared the provinces of Shocks and Fragility Sila and Ouaddaï, as disputes between pastoralists and farmers threatened to erupt into widespread conflict.39 Chad is located in a vulnerable region, and Chadian households regularly face both natural and manmade The Gender Gap shocks. The closure of the border with Nigeria since August 2019 has dealt a devastating blow to the Chadian Sustainably increasing rural income will require measures economy, as Nigeria provides a critical trade link both to address the wide gender disparities that prevail in rural for imports and exports. Repeated border closures areas. Compared to men, women tend to have less access have negatively impacted farmers’ output prices while to human capital and productive assets, such as land and increasing input prices,43 with a disproportionate impact large livestock, due to social and structural barriers. Early on the populations of the Lac and Hadjer-Lamis regions. marriage is among the most harmful social norms, as it The Recovery and Peacebuilding Assessment reported a keeps girls out of school and contributes to low levels of slowdown in business activities in the area due to reduced human capital among women. In Chad, high fertility rates purchasing power and weakening demand.44 and low maternal health indicators limit women’s livelihood 36 Goldstein, M., Houngbedji, K., Kondylis, F., O’Sullivan, M., & Selod, H. (2018). Formalization without certification? experimental evidence on property rights and investment. Journal of Development Economics, 132, 57-74. 37 World Bank (2020). Land property rights, investments, and agricultural productivity in Chad: Evidence from the 2018 LSMS-ISA in Chad. Volume I – Main synthesis report. 38 World Bank (2020). Land property rights, investments, and agricultural productivity in Chad: Evidence from the 2018 LSMS-ISA in Chad. Volume I – Main synthesis report. 39 World Bank (2020). Land property rights, investments, and agricultural productivity in Chad: Evidence from the 2018 LSMS-ISA in Chad. Volume I – Main synthesis report. 40 According to the DHS STAT Compiler, Chad is among the countries with the largest share of women who agree that domestic violence is justified when a woman goes out without telling her husband. 41 Kondylis et al. (2016) and Cohen and Lemma (2011). 42 Kondylis, F., Mueller, V., Sheriff, G., & Zhu, S. (2016). Do female instructors reduce gender bias in diffusion of sustainable land management techniques? Experimental evidence from Mozambique. World Development, 78, 436-449 43 WFP (2016). Lake Chad basin crisis regional market assessment preliminary observations. 44 FAO (2017). Mitigating the impact of the crisis and strengthening the resilience and food security of conflict-affected communities. http://www.fao. org/3/a-bs126e.pdf. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 69 Figure 2.18. Access to Financial Services by Gender, Chad and Mali 29% 27% 20% 20% 13% 11% 10% 5% Male Female Male Female Chad Mali Source: FINDEX 2017. Mobile money account (% age 15+) Financial account (% age 15+) Rural livelihoods are increasingly at risk from drought next 20 years at its current rate of use.46 The intensifying due to declining levels of rainfall and rising temperatures scarcity of water is negatively impacting agriculture and linked to climate change. In the northern part of the food security, threatening to drive even more households country, desertification is threatening grazing lands, forcing into extreme poverty.47 pastoralists to migrate to the south, where pressure on land is rising.45 High rainfall variability linked to climatic The Boko Haram insurgency in the Lake Chad region has change is also making households more vulnerable to severely disrupted economic activity. The conflict has crop-production shocks. For example, drought or irregular sharply reduced the incomes of many pastoralist, agro- rainfall affects about 20 percent of households, making it pastoralist, and fishing households that rely on cross- the largest source of natural covariate shocks. Drought or border trade with neighboring countries such as Cameroon, irregular rainfall is also the major cause of vulnerability CAR, Nigeria, Libya, and Sudan. Violent clashes have led to due to covariate shocks, affecting about 48 percent of the forced displacement of households, resulting in losses households. Climate change and the unsustainable use of crops, livestock, and productive assets. Meanwhile, an of natural resources also threaten the livelihoods of influx of refugees into Chad from neighboring countries has crop producers and fishermen. Lake Chad has shrunk further strained livelihoods and diminished access to land.48 dramatically and is predicted to vanish entirely in the 45 World Bank Climate Knowledge Portal. https://climateknowledgeportal.worldbank.org/country/chad/vulnerability 46 NASA (2017). The Rise and Fall of Africa’s Great Lake. https://earthobservatory.nasa.gov/features/LakeChad 47 FAO (2017). Lake Chad Basin: A Crisis Rooted in Hunger, Poverty and Lack of Rural Development. http://www.fao.org/news/story/en/item/880741/icode/ 48 WFP (2018). Executive Board. Chad Draft Country Strategic Plan (2019-2023). https://docs.wfp.org/api/documents/6ddef21988944069ae3d908a9cec1d20/ download/; Watson, C., Dnalbaye, E., & Nan-guer, B. (2018). Refugee and Host Communities in Chad: Dynamics of Economic and Social Inclusion. 70 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 2.3 POLICY OPTIONS FOR SUPPORTING RURAL INCOME GROWTH Chad’s poor population is overwhelmingly concentrated the country’s underexploited agricultural potential in rural areas, and poverty rates are especially high offers valuable opportunities to accelerate its economic among smallholder farming households. At the national development. Only 6 percent of Chad’s arable land is level, agriculture accounts for 65 percent of household cultivated, and more that 80 percent of farms cultivate income, while nonfarm income represents 30 percent, fewer than two hectares, underscoring the possibilities and remittances make up the remaining 2 percent. Among to increase crop production through area expansion. poor households, agriculture accounts for 71 percent of Moreover, only 9 percent of Chad’s available water household income, while nonfarm income represents resources are being used to irrigate less than 1 percent of about 19 percent, and remittances make up just 4 percent. its agricultural land. Investing in irrigation networks could Moreover, an estimated 85 percent of the rural population enable farmers to diversify into cash crops and generate depends on agriculture and related activities, as about 88 additional income during the off season. The prevalence percent and 62 percent of rural households are engaged of agro-pastoralism in the southern AEZs offers a unique in crop and livestock production, respectively. The relative opportunity to develop complementary and inclusive value importance of these activities varies across the Saharan, chains for crops and livestock. Given rising international Sahelian, and Soudanian AEZs. demand for gum arabic, shea, dates, and other NTFPs, linking producers to national and global markets could Due to the composition of the Chadian economy and the greatly accelerate rural development. distribution of poor households, rural development will be critical to poverty reduction and shared prosperity. The Six key challenges constrain rural income growth. These international experience shows that in countries such as include: (i) insecurity and the risk of conflict over natural Chad, which have largely agricultural economies and high resources; (ii) the risk of climatic shocks; (iii) low levels rates of rural poverty, agriculture can act as the engine of human capital; (iv) a wide infrastructure gap; (v) a lack of economic growth and a key instrument of poverty of complementary services; and (vi) a deep gender gap. reduction.49 Moreover, improving agricultural productivity To make agricultural development an effective pathway can stimulate growth in other parts of the economy. out of poverty, policymakers must address each of these Accelerating agricultural growth requires increasing the constraints. productivity of smallholder farming while protecting the welfare of vulnerable households. Though Chad’s rural sector faces considerable challenges, Security is the most fundamental prerequisite for rural 49 World Bank (2002) World Development Report. Washington DC: The World Bank INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 71 development in Chad. Greater international efforts to Risk management can build household resilience to adverse address the Boko Haram conflict and enable the return of shocks and enable rural workers and entrepreneurs to take refugees and internally displaced persons will be critical advantage of new opportunities. Building resilience entails to reestablish international trade linkages and enable developing financial markets, insurance schemes, and investment in the rural economy. The Chadian government social-protection systems capable of alleviating reliance is currently working with the African Union, France, and on costly coping mechanisms and enabling households neighboring countries to improve the security situation. to invest more in risky but more profitable agricultural However, even a permanent resolution to the Boko Haram activities. Index insurance products, such as agricultural conflict would not be sufficient to alleviate the broader index insurance or index-based livestock insurance, have security risks stemming from increased competition for emerged as a suitable instrument for bolstering resilience scarce and dwindling resources, especially water and among vulnerable agricultural households in developing pastureland. Sustainably addressing conflict risks in rural countries with limited institutional capacity and weak Chad will require a coordinated multi-stakeholder effort information systems. Policymakers can begin constructing that combines direct security assistance with large-scale the institutional framework for implementing these investments in economic infrastructure and ecological mechanisms in Chad. restoration. Over the longer term, a robust capacity- building program focused on strengthening the legal Accelerating human capital development will be critical system, enhancing land-tenure security, and formally to support productivity growth and diversification. establishing the land-use rights of pastoralist groups could Chad’s extremely poor health and education indicators further attenuate the risk of conflict. directly reduce the income-generating capabilities of rural households. The government faces a very tight resource Chad is highly vulnerable to climatic shocks. Rural envelope, and Chad’s diffuse, low-density population livelihoods are increasingly at risk from drought due to increases the marginal cost of providing social services, declining levels of rainfall and rising temperatures linked to making the prioritization of cost-effective interventions climate change. Desertification is threatening grazing lands especially important. Improvements in maternal health in the northern part of the country, forcing pastoralists to can offer high returns, as efforts to reduce adolescent migrate to the south, where pressure on land is rising. High pregnancy and encourage family planning are often less rainfall variability linked to climatic change is also making costly than other health interventions and yield benefits households more vulnerable to crop-production shocks. The at the household level that extend across generations. pursuit of high profitable opportunities necessarily entails Meanwhile, expanding school feeding programs could risk taking. Poor people tend to be risk averse because they boost educational attendance while improving nutrition fear the potential negative consequences of failure. The indicators, with positive implications for the lifetime absence of crop or livestock insurance heightens production productivity of children. uncertainty due to weather-related shocks. The lack of insurance discourages farmers from making productive investments, taking risks on new crop types and production models, or accessing new market opportunities. Effective risk management can be a powerful instrument for building resilience to risk, which is essential to achieving and sharing prosperity, and reducing poverty.50 50 World Development Report 2014: Risk and Opportunity: Managing Risk for Development. Washington, DC: The World Bank Group. The report emphasizes that this characterization of risk management is true whether risk stems from natural disasters, pandemics, financial crises, a wave of crime at the community level, or severe illness of a household’s main provider. In general, building resilience to risk entails a balanced approach that includes structural policy measures, community-based prevention, insurance, education, training, and effective regulation. 72 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Infrastructure investment can increase commercialization bring down input costs while making input sales viable in and create new opportunities for value addition. Access remote locations that cannot sustain a permanent market. to paved roads and permanent markets is strongly Similarly, collective marketing can reduce reliance on correlated with participation in commercial agriculture. spot transactions, alleviate information asymmetry, and While budgetary constraints limit the government’s ability encourage farmers to diversify into cash crops. Farmer to invest in costly road networks, targeted investments in organizations can also facilitate the provision of training areas with underutilized potential to produce export crops and extension services and encourage the adoption of new such as gum arabic, shea, and dates could contribute to technologies and production methods. income growth and diversification at a manageable fiscal cost. In addition, modest public investments in ICT-related Gender is a vital cross-cutting issue in Chad and should be infrastructure could enable the expansion of digital services, mainstreamed into all rural development efforts. Gender enhancing access to market information and attenuating disparities in education and health indicators, land access, the high transaction costs imposed by the low quality of the use of financial services and technology, and the the road network. Expanded ICT access would also increase ability to participate in the full range of economic activities the uptake of digital financial services, easing capital have profound consequences for household productivity, constraints that inhibit investment and entrepreneurship. income diversification, and intergenerational economic Policymakers can work with private ICT firms to identify mobility. Gender norms are socially ingrained, and while bottlenecks to digital connectivity and develop cooperative some disparities can be addressed through investments in strategies to address them. health, education, and social services, others will require sustained outreach efforts designed to highlight the broad- Limited insurance markets, inadequate public investment based gains that households and communities can realize in ICT infrastructure, and weak land rights restrict the ability by eliminating gender gaps. Prioritizing maternal health, of the rural population to access market opportunities encouraging girls to complete school before marriage, and invest in productive activities. As communication proactively recruiting women as agricultural extension and transportation links improve, rural extension services workers and lead farmers, and implementing both legal can collaborate with international partners to form farmer and administrative reforms to equalize access to land and organizations capable of leveraging economies of scale financial services could enable rural women to realize far in both input supply and output marketing. These groups more of their productive potential, with positive effects on can be organized geographically or around the production their economic security, health, and education. of specific crops. In either case, collective purchasing can INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 73 CHAPTER VULNERABILITY TO 03 SHOCKS IN CHAD 74 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Key Insights • Chad’s national poverty rate is 42 percent, but 48 percent of households are vulnerable to • Following a shock, key welfare indicators deteriorate to a similar extent among different falling into poverty in the event of a shock. demographic groups, male- and female-headed households, households headed by people with different occupations, and households that do and do not receive transfers. • Poverty and vulnerability rates are both highest in rural areas, where a negligible share of households is financially secure. • In response to shocks, households rely primarily on informal coping mechanisms, such as support from family and friends and their own savings. A sizeable share of households, • Vulnerability to poverty is driven by high levels of consumption variability. especially in rural areas, also reduce their consumption of food, buy cheaper food, and/or sell their livestock to mitigate the effects of shocks. • Households headed by a white-collar worker tend to be among the least vulnerable to poverty, while rural households—especially those with children under the age of six—are most likely to • Female-headed households are less likely to draw on savings than are their male peers, likely fall into poverty in the event of a shock. due to their lower rates of savings. Additionally, many households in the cash crop and cereals zone reduce their food consumption in the event of a shock. • A large share of households experience shocks, including the severe illness or injury of a household member (39 percent), the death of a household member (24 percent), and drought • The COVID-19 pandemic has resulted in higher food prices, the bankruptcy of nonfarm or irregular rainfall (20 percent). Households in the cash crop and cereals zone are the most enterprises, and the loss of wage employment, and many households have suffered from the exposed to shocks. illness of an income earner. As a result, household income has declined, and many households have seen their ability to access essential food items diminish. Households have mainly • At the national level, households’ exposure to covariate and idiosyncratic shocks is broadly reduced their consumption and used their personal savings to mitigate the effects of the crisis. similar. However, poor households are more exposed to natural covariate shocks, and those in the cash crop and cereals zone are more exposed to demographic idiosyncratic shocks. • The authorities need to strengthen formal safety nets to ease reliance on informal social The incidence of demographic idiosyncratic shocks is also higher among female-headed support systems and prevent the use of detrimental coping strategies. The government should households. consider a combination of short-term interventions to provide immediate relief following a shock, such as cash transfers or temporary food subsidies, and longer-term schemes to • Exposure to shocks is associated with the depletion of assets, including livestock, and/or increase the overall resilience and capacity of households to respond to shocks, including declines in income. Many households adopt detrimental coping strategies, such as selling or efforts to accelerate human capital development. consuming assets, especially in rural areas. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 75 3.1 VULNERABILITY TO POVERTY Vulnerability to poverty is defined as the likelihood that assets to escape poverty.52 Moreover, static measures are a household will fall below the poverty line during a especially unsuitable for gauging vulnerability to poverty, given period. Vulnerability reflects both the probability which is influenced by multiple factors that change over that a negative shock will occur and its potential impact time.53 Chadian households are especially likely to face on household welfare.51 The methodology for measuring fluctuations in income and consumption levels, as Chad vulnerability is presented in Annex E. While a static time- is a landlocked Saharan-Sahelian country that faces bound measure of poverty enables the identification numerous shocks in any given year. Relying on a static of households with a per-person consumption level measure in this context would elide critical nuances in below the national poverty line, it does not distinguish poverty dynamics that determine both a household’s between households that are structurally poor and those current poverty status and its vulnerability to poverty. that are transitionally poor but have access to sufficient Table 3.1. Idiosyncratic and Covariate Shocks Idiosyncratic Shocks Covariate Shocks Severe illness or injury of a household member Drought or irregular rainfall Flooding Demographic Death of a household member Fire Divorce/separation Natural High rate of crop disease High rate of animal disease End of regular transfers from other households Locusts or other pests Important loss of nonfarm income Landslide Important output price drop Bankruptcy of nonfarm enterprise Economic Economic High input prices Important loss of salary incomes High food prices Loss of wage employment Farmer/pastoralist conflict Violence Armed conflict, violence, or insecurity Theft of money, assets, production, or other goods 17.5 Source: World Bank staff calculations using data from ECOSIT 4 51 Calvo and Dercon, 2013; Hoddinott and Quisumbing, 2003. 52 Baulch and Hoddinott, 2000; Carter and Barrett, 2006. 53 Gunther and Harttgen, 2009. 76 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Vulnerability to poverty can be decomposed into poverty- of consumption, which reflects household-level exposure induced and risk-induced vulnerability. Poverty-induced to shocks. These households are at risk of falling below vulnerability, or structural poverty, reflects persistently the poverty line because of their sensitivity to idiosyncratic low levels of consumption, which are rooted in low levels shocks, such as the illness or death of an income-earning of physical and human capital accumulation. Households household member, and covariate shocks, such as a experiencing poverty-induced vulnerability are prone natural disaster (Table 3.1).54 The sum of poverty-induced to falling below the poverty line because they lack the vulnerability and risk-induced vulnerability is the overall productive capacity to achieve economic security. By vulnerability rate. contrast, risk-induced vulnerability is driven by the volatility Figure 3.1 Close to half of Chad’s population is vulnerable to poverty. 42% National 53% 49% Rural 60% 20% Urban 29% 0% 20% 40% 60% 80% Poverty rate Vulnerability rate (total) Source: World Bank staff calculation using data from ECOSIT 4 54 Covariate shocks affect everyone within a defined geographic region at the same time, while idiosyncratic shocks have individual-specific and isolated effects. Covariate shocks include natural, economic, and conflict-driven events, while idiosyncratic shocks include household-level financial and demographic events. The incidence of shocks is based on household surveys, which asked respondents whether they had experienced any of the events listed in Table 3.1 over the past three years. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 77 This poverty assessment estimates the expected mean distinction, and it overlooks the many Chadian households and variance in consumption using both household- that are at risk of falling into poverty in the future (Figure and community-level characteristics.55 This approach 3.1). The vulnerability rate compensates for this limitation makes it possible to analyze both the level and sources by capturing the probability that a nonpoor household will of vulnerability. The assessment begins by quantifying drop below the poverty line over the near term. Exposure vulnerability to poverty and examining its sources. It then to idiosyncratic shocks represents close to 52 percent of estimates the relationship between various household the vulnerability rate, while exposure to covariate shocks characteristics and vulnerability to poverty. represents 45 percent (Figure 3.2). Data from the 2018/2019 ECOSIT 4 suggest that vulnerability due to covariate shocks is primarily attributable to drought or irregular rainfall Among nonpoor households in Chad, 53 percent have (24 percent), while vulnerability to idiosyncratic shocks a 20 percent or greater chance of falling into poverty arises largely from the risk of severe illness or injury of a in the next year.56 Chad’s poverty rate is extremely high household member (39 percent). at 42 percent,57 but the poverty rate is a simple binary Figure 3.2 : Idiosyncratic shocks have a marginally greater impact on household vulnerability than covariate shocks. 53% National 60% Rural 29% Urban 0% 20% 40% 60% 80% Vulnerability rate (total) Vulnerability rate (idiosyncratic) Vulnerability rate (covariate) Source: World Bank staff calculation using data from ECOSIT 4 55 The analytical methodology is based on the work of Gunther and Harttgen (2009). Please refer to Table E.1 in the appendix for a summary statistic of the household and community level variables used in the estimation. 56 The measure of vulnerability to poverty and poverty rate should not be added as they capture different metrics of poverty. In the estimation, to identify households that are vulnerable to poverty, we chose a threshold of 20 percent above which a household is categorized as vulnerable to poverty hence the interpretation. 57 Chad’s poverty rate is based on the national poverty line of FCFA 242,094. 78 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Across livelihood zones, households in the cash crop covariate shocks (Figure 3.4). However, an analysis of the and cereals zone are the most vulnerable to poverty. idiosyncratic/covariate ratio for vulnerability reveals that This zone has the highest vulnerability rate at about 61 the relative importance of covariate shocks is greatest in percent, followed by the cereal and market gardening zone the pastoralism and transhumance zone. at 50 percent and the agro-pastoralism zone at 46 percent (Figure 3.3). The pastoralism and transhumance zone has Most rural households are vulnerable to poverty. the lowest vulnerability rate at 30 percent, which is likely Approximately 60 percent of rural households are due to the relative importance of livestock in this area, as vulnerable to poverty, compared to just 29 percent of urban herds are less sensitive to erratic weather patterns than households. Rural households also have a significantly crops. Livelihood zones also exhibit large differences in higher average poverty rate at 49 percent, versus a poverty poverty rates, which range from 50 percent in the cereals rate of 29 percent in urban areas. As rural households zone to 28 percent in the pastoralism and transhumance account for approximately 77 percent of the population, zone. Across all zones, vulnerability due to idiosyncratic these high rates indicate that both poverty and vulnerability shocks is moderately higher than vulnerability due to to poverty are heavily concentrated in rural areas. Figure 3.3 : Households in the cash crop and cereals zone Figure 3.4 : Vulnerability due to covariate shocks is are the most vulnerable to poverty. prevalent across regions. Pastoralism and 30% Pastoralism and 30% transhumance 28% 23% transhumance Cereal and market 50% Cereal and market 49% 40% 40% gardenic gardenic 46% 46% Agro-pastoralism 34% Agro-pastoralism 39% 61% 60% Cash crop and cereals Cash crop and cereals 50% 53% 0% 20% 40% 60% 80% 0% 20% 40% 60% 80% Vulnerability rate (total) Poverty rate Vulnerability rate Vulnerability rate (idiosyncratic) (covariate) INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 79 Covariate shocks have a slightly stronger effect on In Chad, vulnerable households primarily contend with the vulnerability of rural households than on urban poverty-induced vulnerability, especially in rural areas. households. The magnitude of vulnerability to poverty due Approximately 43 percent of households are subject to to idiosyncratic shocks is greater than that of covariate poverty-induced vulnerability, while just 10 percent face shocks both in rural and urban areas (Figure 3.2). The risk-induced vulnerability (Figure 3.5). The ratio between idiosyncratic vulnerability rate is 59 percent in rural areas poverty-induced and risk-induced vulnerability is 4.5 and 29 percent in urban centers, while the covariate in rural areas versus 2.3 in urban areas, indicating that vulnerability rate is 52 percent and 22 percent, respectively. poverty-induced vulnerability is 4.5 times greater than risk- However, the idiosyncratic/covariate ratio for vulnerability induced vulnerability among rural households. In other in rural areas (1.1) and urban centers (1.3) reveals that the words, many rural households are unable to rise far enough relative size of covariate shock is greater in rural areas. above the poverty line that they are no longer vulnerable to Covariate shocks such as drought have an adverse impact falling below it. on agricultural production, and idiosyncratic shocks such as the severe illness of a household member can be particularly damaging to agricultural households that depend on family labor. Figure 3.5: Household vulnerability is driven by low levels Figure 3.6: Low levels of consumption are the of consumption, as most nonpoor households remain close dominant source of vulnerability across all regions. to the poverty line. 70% 70% 60% 60% 10% 50% 11% 11% 12% 50% 40% 10% 30% 40% 51% 11% 20% 38% 36% 30% 10% 20% 9% 0% als lism nic e 20% anc e rde cer ora hum t ga ast and ans rke o-p rop tr ma 10% Agr nd hc and ma Cas 20% 49% 43% eal alis Cer tor 0% Pas Urban Rural National Risk induced vulnerability Risk induced vulnerability Poverty induced vulnerability Poverty induced vulnerability Source: World Bank staff calculation using data from ECOSIT 4 80 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Poverty-induced vulnerability is prevalent across Vulnerability rates are lowest among households headed livelihood zones, but rates of poverty-induced vulnerability by a skilled office worker. The results of a multivariate are highest in the cash crop and cereals zone (Figure 3.6). analysis (logistic model) show that households headed by The poverty-induced vulnerability rate ranges from 20 to a skilled office worker are the least likely to be vulnerable 51 percent nationwide, while the risk-induced vulnerability to poverty, though only about 7 percent of Chadian rate ranges from 10 to 12 percent. Among households in households fall into this category (Figure 3.7). Moreover, the cash crop and cereals zone, the poverty-induced having completed tertiary education, engaging in livestock vulnerability rate is 5.3 times the risk-induced vulnerability production, and being employed in the private sector rate. By contrast, in the pastoralism and transhumance are all correlated with a lower likelihood of household zone the poverty-induced vulnerability rate is just 1.8 times vulnerability. By contrast, heads of household engaged the risk-induced vulnerability rate. in the agricultural sector and those with young children (especially ages five and under) are more likely to be vulnerable to poverty. Across regions, rural households are significantly more likely to be vulnerable to poverty than urban households. Figure 3.7: The presence of young children significantly increases the likelihood of household-level vulnerability to poverty. Residence in rural areas Household size Age of household head (years) percent aged under 5 percent aged btw 5 & 10 percent aged btw 11 & 15 percent aged btw 16 & 24 percent aged 45+ Head - primary school Head - secondary school Head - tertiary school Participation in livestock Female-headed household Works in culture/livestock/fihing (HH head) Works in a management position (HH head) Works in private sector (HH head) -.2 0. 2. 4. 6 Source: World Bank staff calculation using data from ECOSIT 4 Notes: The figure presents logit marginal effects of the correlates of being vulnerable to poverty. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 81 3.2 OCCURRENCE OF SHOCKS Almost all households in Chad have been affected by The incidence of covariate and idiosyncratic shocks some type of shock. Between 2014 and 2017, nearly 90 differs between rural and urban areas. Both covariate percent of the population experienced a shock, including and idiosyncratic shocks affect approximately 63 percent 89 percent of rural households and 86 percent of urban of households in Chad. Households in urban areas are, households (Figure 3.8). This finding is consistent with however, more likely than their counterparts in rural areas previous analyses of household-level shocks in the Sahel to be affected by idiosyncratic shocks (71 percent), as rural region, which indicate that most households in that region households predominantly encounter covariate shocks (67 are exposed to repeated idiosyncratic and covariate percent). This rural-urban difference in the prevalent types shocks.58 However, Chadian households are more exposed of shocks potentially calls for location-based contextual to shocks than their counterparts in Sahelian comparators safety-net policies to meet the different needs of urban countries (73 percent) and in Senegal (48 percent). The and rural households. most common shocks in Chad are: (i) the severe illness or injury of a household member (39 percent), (ii) the death Most poor households in Chad are affected by natural of a household member (24 percent), and (iii) drought or covariate shocks. About 90 percent of both poor and irregular rainfall (20 percent). nonpoor households report experiencing shocks. However, about 70 percent of poor households experience covariate The relatively large share of households affected by the shocks, while 59 percent experience idiosyncratic shocks. illness or death of a household member has a direct As in other Sahelian countries, the incidence of climate- impact on agricultural productivity. About 80 percent and weather-related shocks is greatest among the poorest of Chadian households are involved in the agriculture households, underscoring the importance of drought- sector, with a majority relying on family labor at different responsive adaptative social protection systems.59 By stages of the planting season. In this context, the loss contrast, 60 percent of nonpoor households are subject of labor due to illness or death can significantly reduce to covariate shocks, while 65 percent are affected by productivity and undermine food security. In addition, idiosyncratic shocks. Overall, 54 percent of poor households many households operate nonfarm microenterprises that are vulnerable to natural covariate shocks, especially also rely on family labor. drought and irregular rainfall. 58 Brunelin et al., 2020. 59 Ibid. 82 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Demographic idiosyncratic shocks affect just over half of In urban areas, about 30 percent of households face Chadian households. In both rural and urban areas, close economic idiosyncratic shocks as well as economic and to 40 percent of the population report being affected by natural covariate shocks. Urban households are less the severe illness or injury of a household member, and likely to suffer from drought or irregular rainfall than about 23-26 percent of households report the death of a their rural counterparts (11 percent versus 27 percent), but household member (Figure 3.9). More than half of rural urban households are more likely to suffer from high food households experience natural covariate shocks: 27 percent prices (25 percent versus 20 percent). The most common are affected by drought or irregular rainfall, 11 percent by economic idiosyncratic shocks facing urban households flooding, 10 percent by a high rate of crop disease, and 9 are the theft of money, assets, production, or other goods percent by a high rate of animal disease. Violence-related (17 percent) and the loss of wage employment (5 percent). covariate shocks, especially farmer/pastoralist conflicts, By contrast, theft is reported by only about 11 percent of are reported by 14 percent of rural households and just 5 rural households, suggesting that security concerns differ percent of urban households. between rural and urban areas. Figure 3.8: Almost all Chadian households Figure 3.9: The incidence of demographic experience shocks. idiosyncratic shocks is especially high. 67% 13.8% Violence 5.3% Covariate shock 50% 11.8% 63% 22.9% Covariate shock Economic 26.4% 23.7% 61% 50.6% Natural 29.6% Idiosyncratic shock 71% 45.7% 64% 15.3% 89% Economic 30.6% Idiosyncratic 18.9% shock Any shock 86% 88% 53.1% Demographic 55.1% 53.6% 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% Rural Urban National Rural Urban National Source: World Bank staff calculation using data from ECOSIT 4 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 83 Households in the cash crop and cereals zone are the is the main type of natural covariate shock and affects most exposed to shocks. About 95 percent of households about one-quarter of households. Economic idiosyncratic in this zone report experiencing a shock in the three years shocks and violence-related and economic covariate preceding the 2018/2019 ECOSIT 4 survey, with no significant shocks affect approximately 20 percent of households in differences in the rate of covariate and idiosyncratic shocks the cash crop and cereals zone. In the pastoralism and (Figure 3.10). The leading type of covariate shock is drought transhumance zone, 30 percent of households experience or irregular rainfall (25 percent), while major idiosyncratic natural covariate shocks, while 15 percent and 5 percent of shocks include the severe illness or injury of a household households are affected by drought/irregular rainfall and member (46 percent) and the death of a household member flooding, respectively. (28 percent). About 87 percent of households in the agro- pastoralism zone are exposed to shocks, as are 83 percent Exposure to covariate shocks is greater among households of households in the cereal and market gardening zone, headed by men than among those headed by women. while just 55 percent of households in the pastoralism and About 65 percent of male-headed households report transhumance zone are exposed to shocks. experiencing covariate shocks, versus 57 percent of female- headed households (Figure 3.11). Male-headed households Demographic idiosyncratic shocks are widespread across are most likely to experience natural and economic all livelihood zones. In the cash crop and cereals zone, covariate shocks, with 23 percent and 20 percent reporting demographic idiosyncratic shocks are the most common, drought/irregular rainfall or high food prices, respectively. followed by natural covariate shocks. The severe illness Given that men often oversee farming activities, shocks or injury of a household member is the leading type of like drought significantly affect their livelihood, although demographic idiosyncratic shock and affects close to drought also makes it more challenging for women to 50 percent of households. Drought or irregular rainfall collect water. Figure 3.10 Households in the cash crop and cereals zone Figure 3.11 Female-headed households are more vulnerable are the most exposed to shocks. to idiosyncratic shocks than their male counterparts. 36.6% 59.2% 57% Covariate shock 59.5% Covariate shock 65% 68.9% 33.1% 51.8% 69% Idiosyncratic shock 59.2% Idiosyncratic shock 62% 73.1% 55.0% 83.2% 88% Any shock 87.1% Any shock 94.6% 88% 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% Pastoralism and transhumance Gender of head of household female Cereal and market gardening Gender of head of household male Agro-pastoralism Cash crop and cereals Source: World Bank staff calculation using data from ECOSIT 4 84 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Female-headed households are more likely than their experience shocks, among households headed by a person male counterparts to experience idiosyncratic shocks.60 with primary, secondary, or no education this share rises About 69 percent of female-headed households report to 90-95 percent.61 Regardless of educational attainment, experiencing idiosyncratic shocks, compared with 62 households in Chad are slightly more likely to experience percent of male-headed households. Moreover, 61 percent idiosyncratic than covariate shocks, except those headed of households headed by a woman report demographic by a person with no education, which exhibit a higher rate idiosyncratic shocks, with 42 percent affected by severe of exposure to covariate shocks. Consistent with national illness or injury of a household member and roughly 27 trends, demographic idiosyncratic shocks frequently affect percent by the death of a household member. The share of all households, regardless of the education level of the female-headed households affected by drought/irregular household head. However, about 35 percent of households rainfall and high food prices is comparable to that of male- headed by someone with higher education also experience headed households. economic idiosyncratic shocks, including: (i) theft of money, assets, production, or other goods (16 percent); (ii) loss of salary income (8 percent); and (iii) loss of wage employment Heads of household with higher education are the least (7 percent). likely to be affected by shocks. Although a strikingly high 80 percent of heads of households with higher education 60 Households self-identified as female headed or male headed households. The survey does not allow us to state whether a household became female headed as a result of one of the shocks that affected a household. 24 percent of households are headed by female against 76 percent headed by male. 61 Only about 5 percent of heads of household have higher education. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 85 3.3 COPING MECHANISMS Households in Chad primarily rely on family or friends Poor households are more likely than nonpoor households to cope with the impact of shocks. Approximately 63 to reduce their consumption and/or procure cheaper food percent of households report receiving help from family or in the event of a shock. While approximately 38 percent friends, and 47 percent report using savings to manage the of poor households report reducing their consumption impact of shocks (Figure 3.12). Among Sahelian comparator following an adverse event, 34 percent of their nonpoor countries, about 34 percent of households mitigate the counterparts do the same. Similarly, 36 percent of poor effects of shocks with the help of friends and relatives, households cope with shocks by procuring cheaper food, while 41 percent rely on savings. In Senegal, these figures compared to 32 percent of their nonpoor counterparts. are 26 percent and 22 percent, respectively. In Chad, other This is aligned with cross-country findings indicating coping strategies that households use during times of that the adoption of detrimental coping strategies tends crisis include: (i) reducing the quantity or number of meals to be most common among the poorest households.62 As consumed in a day (36 percent); (ii) procuring cheaper food expected, poor households rely on savings (40 percent) and (33 percent); and (iii) selling livestock (17 percent). These family or friends (58 percent) at a lower rate than nonpoor findings underscore the threat that household-level shocks households (51 percent and 65 percent, respectively). pose to food security Only 3.9 percent of the country’s households report not adopting any coping strategy in the event of a shock. Figure 3.12: Most Chadian households rely on help from family or friends in the event of a shock, a significantly larger share than in most comparable countries. Help from family or friends 63% Others 51% Using savings 47% Reducing consumption 36% Buying cheap food 33% Selling food stock 22% Selling livestock 17% Selling assets 15% Borrowing money 11% Help from NGOs 4% Help from government 3% Putting children in foster care 2% Source: World Bank staff calculation using data from ECOSIT 4 62 Brunelin, 2020 86 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Rural households are more likely than their urban 58 and 67 percent of those in the agro-pastoralism, cereal counterparts to sell stored food and livestock in response and market gardening, and pastoralism and transhumance to shocks. In rural areas, about 26 percent and 21 percent zones use savings. As a result, households in the cash crop of households respond to an emergency by selling livestock and cereals zone are both the most exposed to shocks and and stored food, respectively, while only 11 percent and 6 the most likely to resort to detrimental coping mechanisms percent of urban households do the same. However, rural/ in the event of a shock. urban differences are smaller for other coping strategies: about 63 percent of rural and urban households receive Female-headed households are somewhat less likely help from family or friends, roughly 35 percent reduce their to use savings as a coping strategy than male-headed consumption, and about 33 percent buy cheaper food. households. While relying on family or friends and using Rural residents are slightly less likely than their urban savings are the dominant types of coping strategies counterparts to use savings in the event of a shock. among all households, there are important differences in their use between female- and male-headed households. Households in the cash crop and cereals zone are especially Approximately 70 percent of female-headed households likely to reduce their consumption or sell stored food to report receiving help from relatives or friends in the cope with shocks. Close to 43 percent of these households event of a shock, compared to 60 percent of male-headed report reducing their consumption, compared to about 28 households, while about 52 percent of male-headed percent of households in the agro-pastoralism and cereal households use savings, versus just 47 percent of female- and market gardening zones. In addition, 31 percent of headed (Figure 3.13). This disparity suggests that female- households in the cash crop and cereals zone report selling headed households are less likely to have savings to draw stored food in the event of a shock. Coping strategies that upon and are therefore more likely to rely on the help involve reducing caloric intake are likely to adversely affect of others during emergencies. Finally, similar shares of the growth and development of younger children in the female- and male-headed households report changing household.63 While 36 percent of households in the cereal their consumption patterns and buying cheaper food in the and cash crop zone use savings in times of crisis, between event of a shock. Figure 3.13: When coping with shocks, female-headed households are less likely to use savings and more likely to rely on help from family or friends. 80% 70% 60% 60% 52% 50% 47% 40% 40% 37% 35% 35% 33% 17% 24% 20% 20% 16% 15% 9% 12% 10% 5% 4% 3% 2% 1% 3% 0% s ers s on od ets y s are nt k ck nd ing GO one toc me sto fo pti ass Oth c rie mN ds sav gm ter ap ern um live or f ling foo che fro fos ng ons gov in ling ily ow Sel ng Usi p n in g gc Hel fam om yin Sel li r Bor Sel e cin p fr Bu ldr om u chi Hel Red p fr g Hel tin Put Source: World Bank staff calculations using data from ECOSIT 4 63 Hoddinott, 2006 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 87 Households whose members work in the services sector with and without small children are roughly equally likely tend to fall back on their savings in the event of a shock. to reduce their consumption or procure cheaper food, with Approximately 61 percent of households with members about 35 percent of both types of households reporting working in the services sector (e.g., hospitality and the former and 33 percent reporting the latter. However, commercial services) use savings as a coping strategy, households with small children are marginally more likely compared to about 43-47 percent of households with to draw on savings or borrow money in the event of a shock members working in agriculture or industries (including than households with no small children. mining and other extractive industries). This divide could reflect differences in earning levels across sectors, as wages Most households report a decline in income and in the in the services sector are often higher than wages in other number of livestock following a shock. Approximately 90 sectors. The shares of households that receive help from percent of Chadians report a sharp drop in income and family or friends, reduce their consumption, or buy cheaper the number of livestock in the aftermath of a shock (Figure food during a crisis are comparable across the three major 3.14). The adverse impact of a shock on income is also employment sectors. evident, though markedly less severe, among households in Sahelian comparator countries (85 percent) and in There are no major differences between the primary Senegal (75 percent). While shocks have a non-negligible coping mechanisms of households with and without small impact on livestock holdings in Sahelian comparators and children.64 Although reduced caloric intake can have in Senegal, reductions in income are consistently greater deeply negative effects on child development, households than the depletion of livestock. Figure 3.14. Shocks lead to major reductions in income and livestock holdings. Food stock 78% 12% 8% 2% Number of livestock 89% 2% 7% 2% Agricultural production 82% 2% 8% 8% Assets 82% 3% 11% 4% Income 93% 2% 5% 1% 0% 20% 40% 60% 80% 100% Decreased Unchanged Increased Not concerned Source: World Bank staff calculations using data from ECOSIT 4 64 Nationwide, 68 percent of households have at least one child under the age of six. 88 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Other key welfare indicators in Chad are also adversely Natural covariate shocks are correlated with the adoption affected by the occurrence of shocks, further indicating of detrimental coping strategies, especially in rural that households adopt negative coping strategies. In the areas.65 In Chad, 32 percent of households respond to wake of a shock, more than 80 percent of households in shocks by using detrimental coping mechanisms such Chad report a decline in agricultural production, food stock, as selling livestock and other durable assets or reducing and assets. Just 8-11 percent of households do not report consumption (Figure 3.15). These coping strategies reduce any change in assets, food stocks, or agricultural production the long-term productivity of households and hinder in the event of a shock. These findings suggest that many their efforts to escape poverty and vulnerability.66 This is of the country’s households lack access to savings or other especially true for rural households, about 21 percent of benign coping mechanisms and are instead forced to adopt which sell their livestock in the event of a shock, compared strategies that reduce their long-term productivity and to just 6 percent of their urban counterparts. consumption. Figure 3.15. Households resort to detrimental coping mechanisms to respond to natural covariate shocks. 12% Violence covariate shock 4% 11% 11% Economic covariate shock 11% 11% Economic idiosyncratic 10% shock 21% 12% 26% Demographic 33% idiosyncratic shock 27% 39% Natural covariate shock 26% 37% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% Rural Urban National Source: World Bank staff calculations using data from ECOSIT 4 65 The full set of detrimental coping strategies considered in this text are child marriage, child labor, pulling children out of school, migration of household members, reduction in health and education expenditures, sale of durable assets, sales of land/real estate, sale of livestock, putting up children in foster care. Besides the sale of livestock or assets or putting children in foster care, the other strategies are adopted by less than 1 percent of households. 66 Brunelin et al., 2020 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 89 In rural areas, the occurrence of shocks has an adverse implement detrimental coping strategies such as selling impact on agricultural production. Approximately 90 assets, reducing food consumption, or removing children percent of rural residents experience a fall both in from school. This finding highlights the general absence agricultural production and in the number of livestock in of unemployment compensation, disability benefits, life the aftermath of a shock (Figure 3.16 ). By contrast, about insurance, or other formal mechanisms for offsetting the 63 percent of urban households see their agricultural sudden loss of income caused by the illness or death of a production fall in the event of a shock, owing to their household member. greater involvement in nonfarm activities. Comparable shares of urban and rural households—between 80 and Households affected by violence-related covariate shocks 90 percent—report a reduction in their income, assets, and are less likely to adopt coping mechanisms. This finding food stocks following a shock. may reflect the limited coping strategies available to these households. During periods of conflict, people are often Demographic idiosyncratic shocks are strongly associated forced to leave their homes unexpectedly and immediately, with the adoption of negative coping strategies. About 36 leaving little to no time to gather their assets. Conflict can percent of urban residents and 34 percent of rural residents also displace or disperse entire communities, disrupting who experience demographic idiosyncratic shocks traditional social mechanisms for coping with shocks. Figure 3.16 The impact of shocks on agricultural production is more severe in rural areas. 100 94% 89% 90% 90 87% 84% 83% 82% 80 78% 78% 70 63% 60 50 40 30 20 10 0 Urban Rural Income Assets Agricultural production Number of livestock Food stock Source: World Bank staff calculations using data from ECOSIT 4 90 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Shocks cause a similar deterioration in the key welfare There are also minimal differences in key welfare indicators of various groups. Shock-induced changes in measures between households that benefit from transfers household income, assets, agricultural production, livestock and those that do not. In the event of a shock, households herds, and food stores vary only slightly among different that receive and do not receive transfers experience similar types of households. Except for the 5 percent of households drops in income, assets, livestock numbers, and food stores. headed by someone with completed higher education, However, about 79 percent of households that receive which are less likely to experience a decline in welfare transfers experience a decrease in agricultural production, indicators than their less-educated counterparts, there are lower than 83 percent of households that do not receive no clear differences in post-shock welfare indictors among transfers, which may indicate that households that receive households based on the gender or occupation of the transfers have more liquidity to invest in agricultural inputs. household head. Likewise, there is no significant difference in changes in welfare indicators between households that The 5 percent of households headed by a person with receive transfers and those that do not. completed higher education are less likely to report a drop in agricultural production, number of livestock, and Comparable shares of male- and female-headed food stores vis-à-vis their less-educated counterparts. In households experience a decline in welfare indicators the event of a shock, close to 58 percent of households following a shock. In the aftermath of a shock, approximately headed by a person with higher education report a decline 90 percent of both types of households report a decline in agricultural production, far below the 82-86 percent of in their income levels and in the size of their livestock households headed by a person with primary, secondary, herds, and roughly 80 percent report a drop in agricultural or no education. Shock-induced declines in livestock production or dwindling assets. While female-headed numbers are reported by 80 percent of households headed households are affected by shocks to the same extent by a person with tertiary education and by 90 percent of as their male counterparts, they lack access to the same those headed by a person with less education. Similarly, coping strategies. For example, 17 percent of households 70 percent of the former report declines in food stores headed by a woman draw down their food stores following following a shock, compared with 80 percent of the latter. a shock, compared to 24 percent of households headed However, 90-95 percent of households report shock- by a man, and female-headed households are also less induced declines in income, while 80-85 percent report likely to sell livestock following a shock. Moreover, female- diminished assets, regardless of the education level of the headed households are more likely to reduce their food household head. consumption or purchase cheaper food. These findings underscore the need for policies that improve women’s access to benign coping strategies. There are no clear differences in the deterioration of welfare indicators based on the occupation of the household head. Indeed, the reduction of income and food stock following a crisis is similar regardless of the occupation of the household head. However, 89 percent of agricultural households experience a decline in agricultural production following a shock, higher than 76 percent of households engaging in industries and for 65 percent of households in the services sector.67 67 71 percent of household heads work in the agricultural sector; 7 percent work in the industrial sector; and 22 percent work in the services sector. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 91 3.4 PRELIMINARY ASSESSMENT OF THE IMPACT OF THE COVID-19 OUTBREAK To support data-driven decision-making in the wake of the June 2020, and the second round followed between July COVID-19 pandemic, the World Bank has collaborated with and August 2020. Reflecting the impact of the pandemic, Chad’s National Institute for Statistics and Economic and the types of shocks assessed in the phone survey differ Demographic Studies (Institut National de la Statistique somewhat from those included in the 2018/19 ECOSIT 4 et des Études Économiques et Démographiques, INSEED) survey, and the survey begins by asking respondents “Has to implement a nationally representative high-frequency your household been negatively affected by the following phone survey. The survey targets a subsample of 2,833 issue since the beginning of the COVID-19 pandemic?”. households that were included in ECOSIT 4, which was Nevertheless, various idiosyncratic and covariate shocks implemented in 2018/19. The first round of the high- assessed by the phone survey are comparable with the frequency phone survey took place between May and data from the ECOSIT 4 survey discussed above (Table 3.2). Table 3.2: Covariate and Idiosyncratic Shocks Assessed in the High-Frequency Phone Survey Idiosyncratic Shocks Covariate Shocks Death or disability of an active adult household member Demographic Death of an individual who sends money to the household Natural Locusts or other pests Illness of an income earner in the household Loss of an important acquaintance Important output price drop Loss of wage employment Economic Bankruptcy of nonfarm enterprise Economic High input prices Theft of money, assets, production, or other goods High food prices Bad harvest owing to lack of labor 92 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION The high-frequency phone survey reveals that the Following the outbreak, households have struggled COVID-19 pandemic has increased the rate of economic primarily with high food prices. In the 2018/19 EHCVM, 39 covariate shocks, particularly in rural areas. About 92 percent of households reported the severe illness or injury percent of households report experiencing a shock during of a household member, making this the most common the pandemic, and rural households have been especially shock recorded by the survey. Since the start of the affected (Figure 3.16). An estimated 69 percent of pandemic, 69 percent of households report being affected households have been impacted by covariate shocks and by shocks related to high food prices. Rising food prices 45 percent by idiosyncratic shocks. All covariate shocks have adversely affected 75 percent or rural households and have been economic, and rural households (75 percent) 50 percent of urban households (see Table F.3 in Annex have been more affected than their urban counterparts F). Other important shocks recorded during the pandemic (52 percent). Meanwhile, urban residents have been more include the illness of an income earner (18 percent) and the affected by economic idiosyncratic shocks (39 percent) bankruptcy of a nonfarm enterprise (14 percent), with no than rural residents (25 percent). These shocks likely reflect major differences observed between rural and urban areas. the lockdown measures implemented by the government A larger share of households reported the loss of wage since March 19, 2020 to contain the spread of COVID-19, employment during pandemic than in the 2018/19 EHCVM. precautionary behaviors adopted by firms and consumers in response to the pandemic, and extensive disruptions that continue to affect global markets and supply chains. Figure 3.17 The pandemic has heightened household exposure to economic covariate shocks. 75% 69% Economic covariate Covariate shock 52% 52% shock 69% 75% 44% 28% 49% Economic idiosyncratic Idiosyncratic shock 39% shock 45% 25% 94% 22% Demographic Any shock 85% 18% idiosyncratic shock 92% 23% 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% Rural Urban National Rural Urban National Source: HFPS 2020. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 93 Economic covariate shocks have had an especially Households’ ability to procure essential food items has significant impact on poor and female-headed households. not improved since March 2020. Since the onset of the Many of the shocks reported during the pandemic appear pandemic, Chadian households have been significantly to be largely unaffected by the gender of the household less able to access staple grains such as rice and maize, as head or by household poverty status, but economic well as other basic food items such as oil and sugar. Most covariate shocks have disproportionately impacted households cite lack of money as their primary obstacle poor and female-headed households. Approximately 49 to procuring adequate food, likely due to the slowdown in percent of poor households have experienced economic economic activity during the pandemic (Figure 3.18 ). Few covariate shocks during the pandemic, compared to 42 households attribute their inability to access essential percent of nonpoor households. Similarly, 54 percent of foods to COVID-19 directly. While food shocks affect female-headed households have reported being exposed comparable shares of urban and rural households, female- to this type of shock, versus 44 percent of male-headed headed households across the country are especially likely households. In addition, the spike in food prices has to report a lack of money as a key obstacle to obtaining affected 73 percent of poor households but just 66 percent basic foods. In addition to reducing the purchasing power of their nonpoor counterparts. for essential food items, the economic shocks induced by the pandemic have eroded the ability of households to pay for healthcare and to save for the future: approximately 81 percent of households report that the pandemic has hindered their ability to pay for healthcare, and 73 percent report that it has negatively affected their ability to save. Figure 3.18 A lack of money has prevented households from accessing essential food items since Mach 2020. 2% COVID-19 4% 3% Round 2 92% Lack of money 86% 91% 4% COVID-19 4% Round 1 4% 90% Lack of money 88% 89% 0% 20% 40% 60% 80% 100% Rural Urban National Source: HFPS 2020. 94 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure 3.19 The impact of the pandemic on income has worsened over time. Rural 72% 64% Urban 75% 67% National 73% 65% 58% 60% 62% 64% 66% 68% 70% 72% 74% 76% Round 2 Round 1 Source: HFPS 2020. The pandemic has had a negative impact on household this share had ticked up to 88 percent. Even larger shares income. Between May and August 2020, the share of of rural residents and poor households reported that households that reported a reduction in income increased their wellbeing had deteriorated. However, the share of sharply (Figure 3.19). The pandemic has had an especially households that believed that they were at risk of losing negative effect on the incomes of workers in agricultural their employment or main source of income within the next businesses, nonfarm enterprises, and salaried positions. four weeks dropped from 52 percent in May to 11 percent In both rounds of surveys, urban households fared worse in August 2020. than their rural counterparts, indicating that their sources of income are more at risk in the current environment. Reducing consumption and using savings have been the Among household members who were working prior to main coping strategies adopted during the pandemic, and the lockdown, 20 percent reported that they had stopped rural households have been more likely than their urban working due to the pandemic. Moreover, 34 percent of counterparts to adopt detrimental coping strategies. female heads of household reported having stopped Approximately 27 percent of all households have reduced working, compared to just 18 percent of male heads of their consumption or used savings to mitigate the effects household. of the crisis (Figure 3.20). Another 14 percent of households have received help from family or friends, and 9 percent Self-reported assessments of wellbeing indicate that have sold livestock. However, the choice of coping strategy most households became less able to meet their basic varies widely between rural and urban areas: 43 percent needs between the start of the pandemic and the second of urban households drew on their savings, versus just 22 round of the survey in August 2020. About 84 percent of percent of rural households; 18 percent of urban households respondents reported that their household wellbeing reduced their consumption, compared to 30 percent of had worsened between March and May, and by August rural households; and only 2 percent of urban households INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 95 sold livestock, versus 11 percent of rural households. These percent during the pandemic. In effect, the pandemic is a figures suggest that rural households have been more nationwide covariate shock that has weakened informal likely to adopt detrimental coping strategies. Less than 1 safety nets by impacting a large majority of households percent of all households have received assistance in the simultaneously. Female-headed households have been form of food or cash transfers from the government, non- less likely to reduce their consumption, and they have governmental organizations, or other groups. relied more on family or friends. Fewer households have resorted to selling livestock during the pandemic than in The COVID-19 crisis has diminished the ability of Chadian the period covered by the ECOSIT 4 survey, which may also households to rely on family or friends to cope with reflect the generalized nature of the shock, as the crisis shocks. While the 2018/2019 EHCVM found that 63 percent is affecting both potential sellers and potential buyers of of households turned to family or friends for support livestock. Poor and nonpoor households have adopted in the event of a shock, this share has plummeted to 27 broadly similar coping strategies. Figure 3.20: Chadian households have primarily reduced their consumption and used savings to mitigate the effects of the pandemic. Taking a loan 1% Reducing education/health expenses 1% Selling agricultural assets 1% Selling durable goods 2% Renting/pawning land 2% Selling food stock 3% Did nothing 4% Buying cheap food 8% Selling livestock 9% Help from family or friends 14% Using savings 27% Reducing consumption 27% 0% 5% 10%1 5% 20% 25% 30% Source: HFPS 2020. 96 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 3.5 CONCLUSION AND POLICY RECOMMENDATIONS The analysis presented in this chapter highlights the Many of the idiosyncratic shocks experienced by Chadian large share of Chadian households that are vulnerable households stem from the severe illness, injury, or death to poverty due to their exposure to both covariate and of a household member. This finding has important idiosyncratic shocks. Vulnerability rates are estimated implications for agricultural productivity and food security: at 50 percent for covariate shocks and 45 percent for about 80 percent of households in Chad are involved in idiosyncratic shocks. Furthermore, more than 80 percent agriculture, and the majority rely on family labor for planting, of households in 17 out of Chad’s 21 regions are exposed cultivation, and harvesting. Moreover, many households to shocks. At the national level, exposure to covariate and own nonfarm microenterprises that rely on family labor. idiosyncratic shocks is broadly similar, but in the three Climate-related covariate shocks may amplify idiosyncratic regions where households are most likely to be affected shocks, as extreme weather events can increase the by shocks, the incidence of covariate shocks is significantly population’s susceptibility to endemic diseases. Across all greater than that of idiosyncratic shocks. AEZs, and in both rural and urban areas, households are more vulnerable to covariate than to idiosyncratic shocks. Chad has experienced recurrent floods, suffered major Households in the cash crop and cereal zones are the most outbreaks of domestic and cross-border conflict, and exposed to shocks and hence most vulnerable to poverty. has a high prevalence of endemic diseases, but the most frequent covariate shocks faced by Chadian households Households in Chad rely on a range of coping mechanisms are drought or irregular rainfall. In recent years, water to respond to shocks. The most common responses include: levels have fallen dramatically in many rivers and lakes, (i) help from family or friends (63 percent), (ii) drawing on particularly Lake Chad. Persistent drought has accelerated savings (47 percent), (iii) reducing consumption (36 percent), the desertification of the northern part of the country, (iv) procuring cheaper food (33 percent), (v) selling stored shrinking the size of agro-pastoral areas and spurring a food (22 percent), (vi) selling livestock (17 percent), and southward shift in livestock grazing patterns.68 Drought (vii) selling assets (15 percent). These coping strategies fall and irregular rainfall also have a deeply negative impact within three main categories: consumption-based, asset- on agricultural production, threatening the livelihoods of based, and assistance-based. While consumption-based millions of people. and asset-based strategies can help households weather a crisis in the short run, they entail a heavy opportunity cost, as they tend to deplete productive assets and human capital, which limits the household’s capacity to improve its living standard in the future. 68 Source: https://climateknowledgeportal.worldbank.org/country/chad/vulnerability. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 97 The COVID-19 pandemic has severely constrained the Drought and irregular rainfall are the most frequent ability of Chadian households to turn to family and friends covariate shocks facing rural households in Chad, and thus as a coping strategy. The main coping mechanisms adopted climate variability is a major source of risk to smallholder by Chadian households include reducing consumption and farmers and pastoralists. Disseminating improved drawing down savings, both of which exacerbate long- production technologies and practices and implementing term vulnerability. Cash and in-kind transfers to affected institutional risk-management interventions can boost households could provide immediate short-term relief. The the marginal productivity of farming households challenge will be to open adequate fiscal space for these and strengthen their resilience to climate variability. transfers while enabling people to safely resume their Improved production technologies and practices include economic activities. the use of drought-resistant seeds, diversified farming systems, and conservation-focused agricultural practices. Household-level vulnerability is driven by a combination Proper targeting is critical to the design of climate-risk of climate-related risks and health risks, and existing management interventions, as the risk-reduction and coping mechanisms are overwhelmingly informal and resilience effects of improved agricultural technologies inadequate. Household vulnerability to shocks is a and practices are context-specific and will vary depending function of both exposure and coping ability. Most Chadian on local bio-physical and socioeconomic factors. households have a limited assets and human capital, and Policymakers must calibrate interventions to reflect they frequently employ consumption-based and asset- these conditions and the livelihood strategies available based coping responses that further deplete their assets to local households. Moreover, the effectiveness of these and erode their human capital. Chadian policymakers must interventions will hinge on the scope and quality of design and implement risk-management interventions that agricultural extension services. build resilience and enable households to escape poverty. The intervention theory underlying most risk-management Index-based insurance is the institutional intervention interventions targeting rural households focuses on most often promoted to manage climate risk in developing smoothing production and consumption while protecting countries. Rather than basing payouts on an assessment of productive assets and human capital.69 Helping households actual crop or livestock losses for an individual household,70 overcome risk-related barriers to adopting improved index-based insurance triggers payouts based on an index production technologies and practices can reinforce of agricultural losses for the larger area in which that household food security and build long-term wealth, which household is located. The index is usually based on factors are vital to mitigating vulnerability and escaping poverty. such as an area’s rainfall or vegetation growth,71 and 69 Hansen et al. 2019. 70 A distinction is commonly made between index-based insurance for crops and livestock. 71 Stoeffer et al., 2018. Agricultural Index Insurance Has Big Impacts for Farmers in Burkina Faso. Innovation Lab for Assets and Market Access Policy Brief No. 2018-05. 98 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION index values tend to be closely correlated with individual support sustainable poverty reduction.73 Social protection outcomes.72 Index insurance has lower transaction costs mechanisms involving cash or in-kind transfers are crucial than conventional insurance mechanisms that require to mitigate vulnerability, and they can alleviate credit, individual losses to be reported, verified, and indemnified, savings, and liquidity constraints, enabling households to making it a more viable option in areas with limited physical exit poverty. The international literature provides evidence and institutional infrastructure. that cash transfers are more effective at mitigating the adverse impacts of drought when combined with Strengthening social protection by establishing adaptive vocational training or productive investment grants. This safety nets can complement the use of index-based type of integrated approach incorporates both livelihood insurance to further strengthen household resilience and promotion and protection. Table 3.3 Summary of policy recommendations Constraints Policy Action Time Horizon Potential Evidence welfare gains and uses Vulnerability to Strengthen the response mechanisms available Long term Positive impact on – Sahel Adaptive Social Protection Program: poverty to households through investments in human education, health, Annual Report 2019. Available at http:// capital (e.g., education) and physical capital income, and food documents.worldbank.org/curated/ (e.g., community infrastructure). security. en/680361585895594749/Sahel-Adaptive-Social- Protection-Program-Annual-Report-2019 Reduce consumption high variability in Medium term Positive impact on consumption through the a diversification of consumption and income sources. welfare. Implement public works programs to further Short term Positive impact on – Gehrke, E., & Hartwig, R. (2018), “Productive effects reduce the volatility of household consumption. income and food of public works programs: What do we know? security. What should we know?”. World development, 107, 111-124. Increase in productive – Zimmermann, L. (2014), “Public works programs investments. in developing countries have the potential to reduce poverty” IZA World of Labor. Introduce basic universal health insurance Long term Positive impact schemes. on health and income. 72 Another desirable property of index insurance is that the index is based on data that is promptly available, collected inexpensively and reliably (e.g., satellite-based imagery). In addition, the data are not manipulable by either the insurer or the insured. For details, see Barrett et al., 2008. Altering Poverty Dynamics with Index Insurance: Northern Kenya’s HSNP+. Basis Brief No. 2008-08. 73 Hansen et al., 2019. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 99 Constraints Policy Action Time Horizon Potential Evidence welfare gains and uses Exposure to Increase the availability of affordable health Medium term Positive impact demographic care services. on health and idiosyncratic income. shocks Establish early warning systems and seasonal Medium term Positive impact – Sahel Adaptive Social Protection Program: forecasts to inform households’ agricultural on agricultural Annual Report 2019. Available at http:// decisions production and document s .worldbank . org/curated/ income. en/680361585895594749/Sahel-Adaptive-Social- Protection-Program-Annual-Report-2019 Exposure Increase access to agricultural technologies Medium term Positive impact – Hansen, J., Hellin, J., Rosenstock, T., Fisher, E., to natural and practices adapted to drought and irregular on agricultural Cairns, J., Stirling, C., ... & Campbell, B. (2019), covariate rainfall. production and “Climate risk management and rural poverty shocks income. reduction”, Agricultural Systems, 172, 28-46. Introduce index-based insurance to help Short term Positive impact – Greatrex, H., Hansen, J., Garvin, S., Diro, R., Le Guen, mitigate the effects of drought or irregular on agricultural M., Blakeley, S., Le Guen M, Rao KN, & Osgood, D. rainfall. investment and (2015). Scaling up index insurance for smallholder production, farmers: Recent evidence and insights. CCAFS health, and Report No. 14. education. – Hansen, J., Hellin, J., Rosenstock, T., Fisher, E., Cairns, J., Stirling, C., ... & Campbell, B. (2019), “Climate risk management and rural poverty reduction”, Agricultural Systems, 172, 28-46. – Stoeffler, Q., Carter, M., Guirkinger, C., & Gelade, W. (2020). “The spillover impact of index insurance on agricultural investment by cotton farmers in Burkina Faso.” NBER Working Paper w27564. Develop social protection schemes, such as Short term Positive impact on food distribution programs and child nutritional food security. support programs, to decrease the likelihood of households reducing their consumption as a coping mechanism. 100 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Constraints Policy Action Time Horizon Potential Evidence welfare gains and uses Prevalence of Develop livestock insurance programs to Medium term Reduction in – Jensen, N. D. & C. B. Barrett (2017). “Agricultural detrimental support pastoralists during periods of drought detrimental Index Insurance for Development.” Applied coping or irregular rainfall. coping strategies, Economic Perspectives and Policy, 39.2, pp. 199- mechanisms including the 219. emergency sale of livestock. Provide cash transfers to help households Short term Positive impact on – Asfaw, S., Davis, B., Dewbre, J., Handa, S., & Winters, purchase essential food items. food security. P. (2014), “Cash transfer programme, productive activities and labour supply: evidence from a randomised experiment in Kenya”, The Journal of Development Studies, 50(8), 1172-1196. Pandemic- Provide in-kind transfers of essential food items Short term Positive impact on related such as rice, maize, sugar, and oil. household food pressures security. on income sources and access to essential food items Extend favorable credit lines to nonfarm Short term Positive impact enterprises. on household income. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 101 CHAPTER HUMAN CAPITAL 04 IN CHAD 102 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Key Insights • Chad score on the Human Capital Index (HCI) was the second lowest in the world in 2020 at 0.30, • Rates of paid and unpaid work are lower among children enrolled in school, but girls spend indicating that a child born today can expect to attain only 30 percent of her lifetime productive twice as much time on domestic chores as boys. Among older students who work, both boys potential. Chad also lags its comparators in terms of gender parity on the HCI. Households in the and girls are more likely to work outside the home. southern regions tend to have higher HCI scores than their northern counterparts. • Over the last decade, maternal mortality rates have declined, but early marriage remains • Many households contend with chronic food insecurity, and more than half of the population is prevalent. Urban households fare better than their rural peers on maternal and child health estimated to be severely food insecure. Food insecurity is especially prevalent among female- outcome indicators. Northern regions have especially poor vaccination coverage, and headed households and rural communities. Across regions, households in northern and central households in the lowest wealth quintiles experience the highest rates of under-five mortality. zones are generally more food secure relative to those in the south. • Girls continue to face enormous challenges, including child marriage, teenage pregnancy, and • In recent years, net primary enrollment rates have improved, reflecting the government’s higher dropout rates than boys. commitment to strengthening the educational system. However, in a country dominated by public schools, education spending as a share of GDP remains extremely low, and weak • The COVID-19 pandemic has left many students without alternative learning options, and the educational outcomes and large gender disparities persist at the primary and secondary levels. resulting crisis threatens food security , especially among the poorest households. Declining income levels have further diminished access to medical, and many households are struggling • School enrollment rates are higher for boys than girls, and this gap widens among older age to meet their basic needs. groups. Literacy rates are also far higher among men than among women. • To accelerate human capital development, the government must increase public spending and • Enrollment gaps are evident across wealth quintiles and between rural and urban areas. build institutional capacity in the education and health sectors. Policies that reduce school However, early childhood education indicators are low across all wealth quintiles. dropout rates are especially critical. Moreover, due to the pandemic, many households require immediate relief to meet their basic needs. • The unavailability of schools and excessive distance to schools are the leading reasons for never attending school. Dilapidated classrooms, inadequate equipment and supplies, and high rates of teacher absenteeism further weaken the quality of the school system. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 103 This chapter assesses the state of human capital in Chad as measured by education, health, and nutrition indicators. This assessment is based on data from the ECOSIT 4, which enables both a descriptive analysis and an econometric analysis of human capital. The COVID-19 pandemic is having a devastating impact on human capital across the globe, and this chapter includes a preliminary analysis of the effects of the crisis on education and food security in Chad. The preliminary analysis draws on data collected from two rounds of a high-frequency phone survey jointly implemented by the World Bank and INSEED. 4.1 ACHIEVEMENTS IN HUMAN CAPITAL In 2018, the World Bank introduced the Human Capital In 2020, Chad ranked second lowest in the global HCI.75 Index (HCI), which presents an aggregate measure of Chad’s HCI score was estimated at 0.30, indicating that a human capital at the national level. The HCI score reflects typical child born today can expect to attain only 30 percent the expected lifetime productivity of a child born today of her lifetime productive potential.76 Nearby Sahelian relative to what her lifetime productivity would have been countries Niger and Mali had HCI scores of 0.32, and Burkina had she enjoyed a complete education and full health. The Faso led the regional comparator group with an HCI score HCI includes three components—survival, education, and of 0.38 (Table 4.1). Across these four benchmark countries, health—and applies five key measures that global research Chad’s HCI ranked lowest in indicators of child survival, has linked to productivity. These five measures are child years of school, and adult survival. In Chad, 88 percent of survival, school enrollment, quality of learning, healthy children survive to age five; a child born today can expect to growth, and adult survival. Each measure has corresponding receive 5.3 years of schooling; and 64.6 percent of 15-year- indicators, including the under-five mortality rate, average olds survive to age 60, far below the average for comparator years of school attendance by age 18, harmonized test countries. Chad outperformed Niger on the rate of stunting, scores, under-five stunting rate, and survival rate between and it performed better than both Mali and Niger on age 15 and age 60. The index calculates a single value harmonized test scores. Between 2010 and 2020, Chad’s HCI between 0 and 1, with a score of 1 indicating that a child score for girls improved but did not close the gap with the born today is expected to achieve her maximum lifetime score for boys.77 productivity.74 74 World Bank, 2018. 75 The Central African Republic ranked lowest with an HCI score of 0.29. 76 World Bank, 2020a. World Bank, 2020a. 77 Ibid. 104 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Table 4.1. Human Capital Index Scores by Component, Chad and Comparators, 2020 Sub-Saharan Indicator Burkina Faso Chad Mali Niger Low Income Africa HCI Component 1: Survival Probability of Survival to Age Five 0.924 0.881 0.902 0.916 0.934 0.928 HCI Component 2: Education Expected Years of School 7.0 5.3 5.2 5.5 8.3 7.6 Harmonized Test Scores 404 333 307 305 374 356 HCI Component 3: Health Survival Rate from Ages 15 to 60 0.761 0.646 0.750 0.767 0.735 0.747 Share of Children Under Five Not Stunted 0.751 0.602 0.731 0.515 0.688 0.654 Human Capital Index Score 0.38 0.30 0.32 0.32 0.40 0.37 Source: World Bank (2020) Note: Scores from international tests are converted into harmonized learning outcomes, with values ranging from approximately 300 to 600 across countries (World Bank 2020). Between 2010 and 2020, Chad made very limited progress A comparison of HCI scores across regions within Chad in improving its HCI score. A marginal increase in expected reveals similar trends, though the southern regions years of schooling among boys boosted Chad’s HCI score performed slightly better than the northern ones. from 0.29 to 0.30 (Table 4.2). As a result, a typical child born N’Djamena had the highest HCI score, but neighboring today can expect to attain an additional 1 percent of her regions such as Chadi-Barguirmi, Lac, and Hadjer-Lamis lifetime productive potential compared to a child born in had some of the lowest scores in the country (Figure 4.1a). 2010. Moreover, the improvement in HCI scores was limited The southern regions tended to have slightly higher scores to boys, causing the gap between boys and girls to widen than their northern counterparts, with values ranging from by 2 percentage points. 0.31-0.32 in the former to an average of 0.30 in the latter. Of Chad’s three major climatic zones, average HCI across were highest in the Soudanian zone and lowest in the Saharan zone, though these differences were modest. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 105 Table 4.2: Human Capital Index Scores by Component, Chad, 2010 and 2020 Indicators Male Female Male + Female 2010 2020 2010 2020 2010 2020 HCI Component 1: Survival Probability of Survival to Age Five 0.870 0.875 0.884 0.888 0.877 0.881 HCI Component 2: School Expected Years of School 5.6 6.2 4.3 4.4 5.0 5.3 Harmonized Test Scores 338 338 323 323 333 333 HCI Component 3: Health Survival Rate from Age 15 to Age 60 0.620 0.625 0.665 0.667 0.642 0.646 Share of Children Under Five Not Stunted 0.591 0.591 0.614 0.614 0.602 0.602 HCI Score 0.30 0.31 0.29 0.29 0.29 0.30 Uncertainty Interval [0.28,0.31] [0.29,0.33] [0.27,0.30] [0.27,0.31] [0.28,0.31] [0.28,0.32] Source: World Bank (2020) Note: Scores from international tests are converted into harmonized learning outcomes, with values ranging from approximately 300 to 600 across countries (World Bank 2020). Poverty is associated with low levels of human capital in is 35. The gap between poor and nonpoor households is Chad. Although average poverty rates are highest in the not consistent across HCI components: the difference in the Soudanian zone and lowest in the Saharan zone, at the probability of survival to age five is 2 just percentage points, regional level high poverty rates tended to correlate with while differences in average years of schooling (5.4 years low HCI scores (Figure 4.1b). For example, the third-poorest versus 9.4 years), harmonized test scores (322 versus 357), region in the country, Tandjilé, also had the third-lowest and the share of children not stunted (59 percent versus 68 human capital index score. According to the latest HCI percent) are much wider. The HCI report urges governments report, which covers 50 countries, the average HCI score for to embrace policies that reduce inequality in access to a child born in the richest 20 percent of households is 45, healthcare between poor and nonpoor households. while the average for a child born in the poorest 20 percent 106 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure 4.1: Human Capital Iand Poverty Rates across Regions (a) Human capital across regions (b) Human capital and poverty rates HCI 0.33 - 0.35 0.31 - 0.32 0.30 0.27 - 0.29 0.26 Source: World Bank staff calculations using data from ECOSIT 4 © Vemaps.com Source: World Bank staff calculations using data from ECOSIT 4 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 107 4.2 EDUCATION Recurrent conflicts, both within Chad and in neighboring net primary enrollment rate is now estimated at 73.2 countries, have created a large population of refugees and percent, outperforming some neighboring countries like internally displaced persons (IDPs) while also disrupting Mali and Niger (Table 4.3). access to education. Shortly after attaining independence from France in 1960, Chad suffered a devastating civil war Despite recent progress, Chad continues to experience and a protracted conflict with Libya that lasted until the both low enrollment rates and wide gender disparities in 1990s. Chad has continued to experience periods of civil primary and secondary education. More than half of all unrest, including ethnic and regional conflicts, as well as students enrolled in primary school do not continue to the armed rebellions in the north and east and violence along secondary level (Table 4.3), and in 2016 the share of female on the border with Sudan’s Darfur region. While Chad’s students transitioning from primary to secondary school internal conflicts largely ceased in January 2010, political was 12 percentage points lower than that of their male unrest and terrorist attacks have continued. Recurrent counterparts. Similar patterns are observed across Chad’s violence compounds Chad’s other socioeconomic four closest regional comparators, for which the average challenges and contributes to its low HCI scores. primary enrollment rate is over 50 percent, while the average secondary enrollment rate is less than 33 percent. Resource dependence has intensified budgetary volatility. Chad ranks lowest in both net secondary enrollment rates Following the discovery of oil, Chad’s budget has become and gender parity at the secondary level. In other words, heavily dependent on oil revenues, leaving public spending Chad has the region’s smallest share of total secondary vulnerable to oil-price volatility. After the 2008 financial students and female secondary students, as well as the crisis led to a sharp decline in oil prices, the Chadian largest share of female students who fail to transition government was forced to sharply reduce spending on from the primary to the secondary level. In addition to low education, health, and other public services. enrollment rates and large gender gaps, Chad faces serious challenges in terms of education access and quality, as just 4 percent of students who complete primary school have In recent years, net primary enrollment rates have adequate competency in math and reading.79 increased due to government programs aimed at strengthening the education system and building national capacity. The National Education for All Action Plan was Though the government has made successive efforts designed to improve the quality of human resources in to increase the education budget, public spending on the education system and to seek to integrate the network education remains low. At 2.5 percent of GDP, Chad’s of refugee-camp schools into the national school system, education spending is the lowest in the region in relative among other objectives. Meanwhile, the government terms (Table 4.3). Low rates of education spending improved the distribution of teaching staff across the weaken compensation incentives for teachers, including country by offering contracts to community teachers, who government-recruited teachers, trained community make up an estimated 54 percent of all teachers in Chad. teachers, and untrained community teachers. The student/ These efforts helped increase the net primary enrollment teacher ratio in Chad stands at 56:1, which further rate from 42 percent in 2003 to 44 percent in 2011.78 Chad’s undermines academic achievement.80 78 Government of Chad, 2019 79 World Bank, n.d. 80 African Development Bank, 2019. 108 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Table 4.3 Education Spending and Outcomes, Chad and Comparator Countries Public spending Public spending Net primary Net secondary Share of female Share of female on education on education enrollment rate enrollment rate primary students secondary students (% of GDP) (% of total public (%) (%) (%) (%) spending) Burkina Faso 6.4 21.4 76.4 29.1 48.8 48.4 Chad 2.5 16.4 73.2 18.9 43.4 31.3 Mali 3.8 16.5 61.3 29.4 46.4 44.1 Niger 3.5 13.2 65.1 20.1 45.6 41.9 Source: UNESCO Institute for Statistics (2020) Notes: These numbers refer to the year 2017 except for enrollment numbers in Chad, which are from 2016. School enrollment rates are higher for boys than for girls, Households in the highest wealth quintiles are more likely and the gender gap widens among older age groups. to send their children to school, but enrollment in early Public schools account for 66 percent of all schools in Chad, childhood education is low across all wealth quintiles. In while private schools make up 30 percent, and community 2016, school enrollment rates were highest for the 12-15 age schools account for the remaining 4 percent. In 2016, group, and children from wealthier households were the enrollment rates were highest among the 12-15 age group at most likely to be enrolled in school (Figure 4.3). Enrollment 55 percent and lowest among the 3-5 age group, indicating rates for the 12-15 age group were highest among major deficiencies in early childhood education for both households in the top wealth quintile at 68 percent, and girls and boys (Figure 4.2). Among the 12-15 age group, 61 lowest among the households in the lowest and second- percent of boys and 48 percent of girls were enrolled in lowest quintiles at 54 percent and 51 percent, respectively. school—a gender gap of 13 percentage points. Among the Enrollment in early childhood education is low across all 16-18 age group, the gender gap in enrollment widens to wealth quintiles at 5-17 percent. The enrollment gap across 23 percentage points. Enrollment rates for both boys and wealth quintiles is lowest for the 16-18 age group, indicating girls decline at age 19, likely reflecting a combination of that access to secondary education is more evenly marriage,81 increased household responsibilities, and an distributed across households at different income levels. inadequate supply of tertiary education. Additionally, 1.5 However, households in the top wealth quintile account for million girls are at risk for child marriage and teenage 28 percent of enrollment among the 19-24 age group. pregnancy, which increase their dropout rates and highlight the enormous challenges they face. 81 The median age for a first marriage in Chad is about 16 years old. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 109 Figure 4.2. School enrollment rates by gender and age group 11% 19-24 years 36% 33% 16-18 years 56% 48% 12-15 years 61% 45% 6-11 years 50% 7% 3-5 years 7% Source: World Bank staff calculation using data from ECOSIT 4 Figure 4.3. School enrollment rates by age group and wealth quintile 80 70 66 60 54 56 53 51 53 50 46 47 48 45 46 40 40 39 41 30 28 20 19 18 21 20 17 10 5 4 7 7 0 3-5 years 6-11 12-15 16-18 19-24 years years years years Lowest Second Middle Fourth Highest Source: World Bank staff calculation using data from ECOSIT 4 110 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Stark differences in primary school enrollment rates are Lack of schools and long distances to school were the apparent across regions, with urban centers outperforming leading reasons for never having attended a formal rural areas. In 2016, urban enrollment rates were about 30 school. About 44 percent of boys ages 7 to 24 reported percentage points higher than rural rates across all age never attending school because there was no school in groups. For the 16-18 age group, the urban enrollment rate their area or it was too far away (Figure 4.5). This share was 68 percent, while the rural rate was 33 percent. Similar was slightly lower among girls at 34 percent, though girls urban/rural disparities were also evident at the primary face unique challenges that prevent them from attending level (Figure 4.4). Across regions, Mayo-Kebbi Ouest and school. A significant share of girls reports not having Moyen-Chari had the highest primary enrollment rates at attended school because of their gender, and families about 80 percent, while Sila, Kanem, and Ouaddaï had the appear to refuse to send girls to school at a higher rate lowest rates at 14-20 percent. These gaps highlight the need than boys. Among households that had access to a school, to expand primary school access across rural areas and affordability appears to be a significant obstacle, especially underserved regions. for boys, and similar shares of boys and girls were kept out of formal schools because of their involvement in paid or unpaid work. Only a small fraction of boys and girls between the ages of 7 and 24 reported not having attended school because it was not suitable or useful. Figure 4.4. Primary Enrollment Rates among Children Ages 7–12 by Region 100 81 81 78 76 76 76 80 72 68 58 60 46 43 39 36 40 31 33 31 25 27 22 19 20 20 14 0 ha ri-B ou i jer- a is em Oc ac one tal al Ma o-Keb ul Keb Est Mo uest Oua ri at Vill Wa ilé e N ira h-E a edi l st Enn Sila Est Sal ï irm Enn -Gaza dda Had ér Bar amen a Lam ent Oue y o am one L dj Cha Bork Bat Log ciden -Ch e d di F Kan Ma and u yo- bi edi agu Tan bi O G Ori yen l 'dj M Log Source: World Bank staff calculation using data from ECOSIT 4 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 111 Children enrolled in school are less likely to participate As they transition into adulthood, both boys and girls in paid or unpaid work, though gender differences are assume greater responsibilities that result in increased significant. Older girls typically spend twice as much participation in work outside the home. In 2016, girls ages more time on domestic work than boys, and the domestic 19-24 who were not enrolled in school spent the most workload is greatest among girls ages 16-24 who are time on domestic work, while boys in the same age group not enrolled in school. (Figure 4.6). In 2016, the average who were not in school were the most likely to report unenrolled girl between the ages of 19 and 24 spent 38 participating in farm work, self-employment, wage labor, hours a week taking care of a child, collecting wood and or other income-generating activities (Figure 4.6). Across water, and performing other household tasks, compared all age groups, enrollment in school was associated with to an average of 12 hours for her male counterpart. While a lower likelihood of involvement in work outside of the school enrollment slightly reduced the number of hours home. This finding highlights the significant financial girls spent on domestic chores, their level of involvement in implications of the decision to keep children in school, as domestic activities remained far greater than that of their many household need both the income and extra labor male peers. provided by their school-aged children. Figure 4.5. Distribution of Reasons for Having Never Attended a Formal School among Respondents Ages 7–24 No school / Too far away Family refuse Boys 44 12 12 18 8 6 Paid / Unpaid work Too Expensive / Not enough money Girls 14 34 19 11 14 5 3 School not suitable / useful Others 0 10 20 30 40 50 60 70 80 90 100 Being a girl Source: World Bank staff calculation using data from ECOSIT 4 112 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure 4.6 Hours of domestic work and work outside the home in the past week, disaggregated by gender, age group, and school enrollment (mean values) Time spent on domestic work (mean hours) 19-24 years 9.6 12.3 16-18 years 8.9 9.6 9.9 13-15 years 10.5 Boys 7.5 7-12 years 8.3 3-6 years 3.6 1.9 19-24 years 24.0 38.4 23.1 16-18 years 31.4 13-15 years 21.3 24.7 Girls 14.3 7-12 years 14.5 3-6 years 6.3 2.9 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 Yes No Source: World Bank staff calculation using data from ECOSIT 4 Share of children working outside the home (%) 19-24 years 32 65 16-18 years 24 40 13-15 years 23 28 Boys 7-12 years 14 16 3-6 years 4 2 20 19-24 years 44 19 16-18 years 31 13-15 years 21 23 Girls 12 7-12 years 11 3-6 years 7 1 0 10 20 30 40 50 60 70 Yes No INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 113 Community schools had the largest number of reported Lack of teachers and teacher absenteeism were also problems. Over 80 percent of community-school parents frequently reported problems in public and community reported a lack of teaching supplies and equipment as schools. In addition to challenges with physical well as dilapidated classrooms (Figure 4.7). By comparison, infrastructure, Chad’s education system lacks adequate just 47 percent of public-school parents and 23 percent of human capital (Figure 4.7). Among parents of students private-school parents reported dilapidated classrooms. enrolled in public schools, about 70 percent reported that Private-school parents were the least likely to report teacher absenteeism was a major problem, while 51 percent problems at school, though 42 percent identified a lack cited a serious lack of teachers. Among parents of students of school supplies as a serious issue. A recent public at community schools, these figures were 60 percent and 72 expenditure analysis for Chad corroborates these findings. percent, respectively. At both community and public schools, The analysis estimates that about one-third of classrooms almost 45 percent of parents reported that poor education are temporary structures and 62 percent are in poor quality hindered their children’s ability to learn. Among condition, with community schools faring the worst.82 parents of students at private schools, only 14 percent Despite increased investments in school modernization identified poor education quality as a serious problem. and improvement since 2006, public education spending represented about 13 percent of total public spending in 2017, well below the target of 20 percent set by the Global Partnership for Education. Figure 4.7 Problems reported at schools, disaggregated by school type (%) 100 89 82 80 80 71 70 72 76 65 67 60 58 60 51 52 47 43 42 41 40 33 28 30 27 26 20 23 20 14 15 12 0 Public Private Community schools Lack of supplies Lack of equipment Teacher absence Poor education Overcrowded classrooms Lack of teachers Lack of bathroom Frequency of tuition fees Deteriorated classrooms Source: World Bank staff calculation using data from ECOSIT 4 82 World Bank (2019) “Public Expenditure Analysis: Chad.” Washington DC: The World Bank 114 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Being a girl and residing in a rural area are negatively Having siblings in school, having at least one parent with at correlated with school enrollment, while being of primary- least some primary school education, residing in a wealthier school age is positively correlated with enrollment. The household, and residing near a school are also positively results of a multivariate analysis (logistic model) reveal that correlated with enrollment. Surprisingly, a high rate of girls and rural children are the least likely to be enrolled reported problems at school does not negatively affect in school, while children ages 7-12 are more likely to be enrollment rates, though this may reflect a lack of alternatives enrolled in school than are children in any other age group. among parents who are unsatisfied with their local school. Figure 4.8. Correlates of the Likelihood of Being Enrolled in School (Logit marginal effects with 95% confidence intervals) Female Child of household head 7-12 years old 13-15 years old 16-18 years old Father primary school Mother primary school Other children in school Second wealth quintile Middle wealth quintile Fourth wealth quintile Highest wealth quintile School in village/neighborhood Many school problems Rural -.1 0 .1 .2 .3 .4 Source: World Bank staff calculation using data from ECOSIT 4 Notes: The base group for age is 3–6 years old. The regression is limited to children ages 3-18 because the survey captured local school presence (an explanatory variable) only up to high school. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 115 Men of all ages exhibit higher literacy rates, and the 13-15 percentage points among respondents ages 55 and gender gap is wider among younger age groups (Figure above. The narrowing gender gap in literacy rates among 4.9). Among respondents ages 15-25, about 62 percent of older cohorts is largely due to the smaller share of older men report being able to read or write in any language, men who are literate in any language, while literacy rates but this rate falls to just 34 percent among women—a gap among women remain relatively stable across age groups. of 27 percentage points. The disparity in literacy rates is This finding reveals how the gains in school enrollment consistent with the disparity in school enrollment rates. rates in recent years have disproportionately benefited men The gender gap in literacy rates remains broadly constant and highlights the need for targeted policies to encourage at 23-27 percentage points up to age 55, then narrows to school enrollment among girls. Figure 4.9. Literacy Rates by Gender and Age Group 70 62 60 50 45 40 35 36 34 30 27 20 20 16 13 12 10 10 3 0 15-25 25-35 35-45 45-55 55-65 >65 Male Female Source: World Bank staff calculation using data from ECOSIT 4 116 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 4.3 HEALTH Decade of civil unrest have prevented Chad from building rates, as pregnant adolescents face an elevated risk of adequate health infrastructure and system capacity, complications during pregnancy and delivery. (Figure 4.14). resulting in limited access to care and low service quality. In Chad 47 percent, of women ages 20-24 report having Following the Abuja Declaration, the Chadian government given birth before the age of 18, and the short average committed to increase public health spending, but the duration between pregnancies likely contributes to adverse sector continues to face enormous deficits in physical and maternal and perinatal outcomes.88 The low coverage of human capital.83 Unskilled workers make up almost half of reproductive and maternal health services such as family the national healthcare labor force. Professional nurses planning and antenatal care partly explain the prevalence and midwives represent slightly more than one-third of all of early and frequent pregnancies and hinder the health healthcare workers, and the country has just 15.6 nurses system’s capacity to identify high-risk pregnancies in a per 100,000 people.84 Fourteen regions experience a high timely manner. prevalence of chronic malnutrition, with rates ranging from 40.1 to 63.9 percent. Moreover, the share of children Household wealth and geographic location also have suffering from stunted growth increased from 38.7 percent a significant impact on maternal health in Chad. In 2017, in 2010 to 40 in 2015.85 These figures indicate that ensuring urban residents and wealthier households reported higher the health, nutritional, and social development of young rates of contraception use and lower rates of early marriage children remains a major challenge in Chad. (Figure 4.14). The share of women who reported completing at least four antennal care visits was 44 percent in urban Improvements in maternal mortality have been slow and areas and just 17 percent in rural areas, and the share of uneven. The national maternal mortality rate declined deliveries attended by a skilled healthcare worker was 59 from 1,420 deaths per 100,000 live births in 2000 to 1,140 percent in urban areas versus and 16 percent in rural areas. in 2015, but it remains more than twice the SSA average of Even greater disparities are observed across household 534 deaths per 100,000 live births in 2017.86 Limited access wealth quintiles: the gap between the richest and poorest to health facilities and skilled healthcare professionals is households for the share of pregnant women completing at a major driver of maternal mortality rates: in 2017, just 20 least four antenatal care visits was 34 percent, and the gap percent of Chadian women reported giving birth in a health in the share of deliveries attended by a skilled healthcare facility, and between 2000 and 2017 the share of births worker was 49 percent.89 attended by a skilled healthcare worker rose only slightly, from 15 percent to 20 percent.87 The prevalence of early Chad’s urban centers outperform rural areas in multiple marriage and adolescent pregnancy, particularly in rural child health outcomes. Children born in urban areas more regions, is likely a major contributor to maternal mortality likely to survive past the age of five, more likely to receive all 83 Under the 2001 Abuja Declaration, African Union member states pledged to allocate at least 15 percent of their annual budgets to the health sector. 84 Mills et al., 2010 85 Government of Chad, 2017 86 WHO, 2019; World Bank, 2020b 87 Obiang Obounou et al., 2020; WHO, 2019 88 UNICEF, 2016 89 Ibid. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 117 eight basic vaccines, and less likely to be stunted (Figure 4.15 Under-five mortality rates tended to be higher among ). At the national level, the under-five mortality rate dropped the poorest and the northern regions had the lowest from 185.1 deaths per 1,000 live births in 2000 to 113.8 in 2019 vaccination coverage. In 2016, only 22 percent of children but remains higher than the rates of neighboring countries, ages 12 to 23 months were fully vaccinated, with coverage far above the SSA average of 76 deaths per 1,000 live births, for rates for tetanus/diphtheria at 50 percent in 2017 and and over four times the SDG target of 25 deaths per 1,000 measles coverage at 41 percent in 2019.93 Across regions, live births.90 The major causes of under-five mortality in vaccine coverage rates ranged from 56 percent in the far Chad are birth asphyxia (32.2 percent), prematurity (27.9 south to 19 percent in the north (Figure 4.11). Household percent), and sepsis (15.8 percent).91 It is estimated that over socioeconomic indicators, such as maternal education 80 percent of all newborn deaths could be prevented by levels or salaried heads of household, were associated access to a skilled healthcare provider, but Chad continues with higher immunization rates.94 Rural areas had both to face a shortage of health workers with a mere 0.4 doctors higher stunting rates and lower vaccination rates than and 5.6 other medical personnel per 10,000 people, far urban centers. Immunization coverage was lowest among below the WHO standard of one doctor per 10,000 people. nomadic pastoralists populations, who have limited access Close to half of Chad’s doctors are in N’Djamena, and very to health services, and in some cases vaccination coverage few doctors are in rural areas.92 rates for livestock exceeded the rates for children.95 Figure 4.10 Maternal Health Indicators across Subpopulations Median age of women at first marriage : 15-49 Married women currently using any method of contraception 17 16.1 16.4 16 16.1 16.2 16 15.9 16.4 14 11.2 11.7 16 12 15 10 14 8 5.7 6 4.3 4.5 4.3 5 13 3.5 12 4 11 2 10 0 Urban Rural Lowest Second Fourth Highest Urban Rural Lowest Second Fourth Highest Middle Middle Total Residence Wealth quintile Total Residence Wealth quintile Total fertility rate for women aged between 15-49 years olds 8 6.8 77 6.8 7 6.4 6.2 6 5.4 5.3 5 4 3 2 1 0 Urban Rural Lowest Second Fourth Highest Middle 90 World Bank, 2019; World Bank, 2020b 91 UNICEF, 2016 92 IMF, 2019 Total Residence Wealth quintile 93 World Bank, 2020b; World Bank, 2020b 94 Gavi, 2019 95 Abakar et al., 2018 Source: DHIS 2014 118 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure 4.11 Child Health Indicators across Subpopulations Under five mortality rate per 10,000 Children stunted 200 161 164 50 39.9 44.7 148 141 149 142 138 41.6 41.2 39.8 40.4 150 133 40 32.4 31.5 100 30 20 50 10 0 0 Urban Rural Lowest Second Fourth Highest Urban Rural Lowest Second Fourth Highest Middle Middle Total Residence Wealth quintile Total Residence Wealth quintile Received all eight basic vaccinations 40 34.5 30.2 30 25.3 24.2 24.7 25.5 23.1 20.8 20 10 0 Urban Rural Lowest Second Fourth Highest Middle Source: DHIS 2014 Total Residence Wealth quintile Stunting is a critical dimension of the human capital cognitive development during the early years of life index (HCI). The stunting rate is a key indicator under accounts for the strong association between stunting and the HCI’s health component, and it directly affects other lifetime productivity.97 components through its impact on child survival and on learning outcomes. Stunting is a manifestation of chronic According to Chad’s 2015 Demographic and Health Survey, undernutrition, which results from inadequate quantity 40 percent of children under five are stunted and suffer and quality of food intake and/or repeated episodes of from chronic malnutrition. Chad’s stunting rate is high by infection. Stunting occurs early in life, with the highest risk the standards of countries with a similar level of GDP per during the first two years, after which its effects are largely capita (Figure 4.12), and two out of five children reach school irreversible.96 Stunting is associated with weaker immune age at a potential cognitive disadvantage compared to their responses, and stunted children have elevated morbidity non-stunted peers. The incidence of stunting has fell by an and mortality risks. Stunting is also effectively a marker average of about 1 percentage point each year from 2004 to of the inadequacy of the early childhood environment, 2015 (Figure 4.13), significantly slower that then global rate of and the interdependence between physical growth and stunting reduction over the same period (-2.1 percent). 96 Black et al., 2013 97 Currie and Vogl, 2013; Galasso and Wagstaff, 2019 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 119 Figure 4.12 Stunting Rates and GDP per Capita Source: WDI 2020; DHS 2015. Figure 4.13. Average Annual Change in Stunting Rates Source: WDI 2020; DHS 2015; MICS 2004. 120 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Despite the gradual improvement in stunting rates, large to 1.5 times in 2004, then narrowed to 1.3 times in 2010 socioeconomic and geographic disparities persist. The and remained stable until 2014 (Figure 4.14). Stunting is large gap in stunting rates between the wealthiest and especially common in the Saharan zone, with rates ranging poorest quintiles has remained broadly consistent in recent between 45 and 60 percent in Tibesti Est, Tibesti Ouest, Am- decades. In 1996, the stunting rate among households in Djarass, Kanem Nord, Kanem, Bourkou, Borkou Yala, and the poorest quintile was roughly 1.4 times higher than that Fada (Figure 4.15). of households in the wealthiest quintile; the gap widened Figure 4.14: Stunting Trends and Inequality 55 50 45 Stunting rate (%) 40 35 30 25 8 2 0 4 6 0 8 2 4 6 6 7 9 1 5 9 3 5 3 7 1 5 20 1 2 00 199 2 00 20 1 2 00 20 1 20 1 20 1 20 1 20 1 2 00 199 2 00 19 9 199 2 00 2 00 2 00 2 00 2 00 199 Source: HEFPI (2020) Population Q1 (20% Poorest) Q5 (20% Richest) Figure 4.15: Stunting Rates at the Prefecture Level Source: IHME 2017, UN Humanitarian Data Exchange INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 121 Like other social services in Chad, access to healthcare In addition to inadequate facilities and human resources, varies widely across regions. In 2017, the average distance the quality of care is a major challenge in Chad’s health to a health facility was three kilometers in urban areas and sector. The problems patients noted most when visiting five kilometers in rural areas. Men and women reported both public and private health facility were long waiting having been ill at similar rates, but men appeared to visit times, inefficient treatments, and high costs. Public health facilities more often. Even though women over 45 health facilities were generally perceived to have more experienced illness at a greater rate than men, they visited problems than private ones, with long waiting times being health facilities less often than their male counterparts an especially frequent complaint (Figure 4.18). The dearth (Figure 4.16). Across all wealth quintiles, distance from the of healthcare providers is exacerbated by their uneven nearest health facility was the main reason for not seeking distribution: of an estimated 4.3 physicians per 100,000 in care (Figure 4.17). The second most frequent reason for 2017,98 over half were based in N’Djamena, along with 88 not seeking care across all wealth quintiles was reliance percent of pharmacists, 100 percent of dentists, and 55.5 on self-treatment, though self-treatment was slightly more percent of midwives.99 common among the poorest households. Figure 4.16 Reported Incidence of Illness and Visits to a Health Facility by Gender and Age Group Experienced illness in past 30 days (%) Visited health facility (among those with illness) (%) 80 50 72.7 46.2 44.5 70 42.3 42.1 65.6 41.1 40.5 40.4 41.6 41.1 60.8 58.5 40 38.1 37.7 37.6 37.9 60 54.9 53.6 33.5 32.0 49.2 49.5 29.9 50 47.5 48.7 30.1 29.2 42.7 42.2 48.7 30 40 35.8 35.1 34.9 31.5 31.9 30 20 20 10 10 0 0 0-5 5-10 10-15 15-25 25-35 35-45 45-55 55-65 >65 0-5 5-10 10-15 15-25 25-35 35-45 45-55 55-65 >65 years years years years Male Female Male Female Source: World Bank staff calculation using data from ECOSIT 4 98 World Bank, 2020b 99 (GoC, 2010) 122 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure 4.17 Reasons for Not Visiting a Health Facility when Ill by Wealth Quintile Highest 71 20 3 3 3 Too far Fourth 68 23 4 32 Self-medication Middle 69 21 4 4 2 Not necessary Second 67 26 3 31 No money/ Too expensive Lowest 61 32 3 22 Other 0 20 40 60 80 100 Source: World Bank staff calculation using data from ECOSIT 4 Figure 4.18. Problems Reported at Public and Private Health Facilities 70 63 60 50 46 48 40 37 30 26 25 27 25 20 18 20 22 20 16 15 9 10 10 0 Facility not Long waiting Too No medicine Bad clean time qualified expensive treatment reception Public Private Source: World Bank staff calculation using data from ECOSIT 4 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 123 4.4 NUTRITION AND FOOD SECURITY Chronic food insecurity is widespread in Chad. modest share of cereals, roots and tubers in the Chadian Approximately 66.2 percent of Chad’s population is diet indicates somewhat greater dietary diversity compared estimated to be severely food insecure, the highest share in to Mali and Burkina Faso, which may be due to higher the region.100 Chad ranks among the bottom ten countries consumption of protein. in the Global Hunger Index, and most households are heavily dependent on agriculture and pastoralism for both nadequate access to basic drinking water and the food and income. High poverty rates, recurrent droughts, high variability of food consumption aggravate food desertification and environmental degradation, and insecurity. In 2017, approximately 39 percent of Chadians increasing insecurity, particularly in the Lake Chad basin, had access to basic drinking water, 15 percentage points drive chronic food insecurity. Conflict-related displacement below the average for nearby comparator countries (Table has pushed an estimated 4.5 million people into a state of 4.4). Inadequate access to drinking water increases the food insecurity, and the number of children at risk of severe incidence of diarrheal diseases, especially in small children, acute malnutrition has more than doubled in recent years which can compound the effects of malnutrition. Moreover, from an estimated 200,294 in 2017 to 461,186 in 2020.101 Chad a comparison of per capita food-supply variability across ranks lowest in the region in terms of average dietary energy the region suggests that Chad’s food systems are the least supply adequacy, indicating that Chadian households are resilient to shocks. Consequently, Chadian households are less able to meet their average dietary needs than their less able than their regional peers to cope with disruptions peers in other Sahelian countries. However, the relatively in the food supply chain. Table 4.4 Food-Security Indicators across Benchmark Countries Average dietary Share of dietary energy GDP per capita (in Per capita food- Share of population Share of female energy supply supply derived from purchasing-power- supply variability with access to basic secondary students adequacy cereals, roots, and tubers parity terms) drinking water (%) (%) Burkina Faso 122 64 2,190 30 48 48.4 Chad 95 62 1,580 31 39 31.3 Mali 135 68 2,327 36 78 44.1 Niger 121 62 1,219 40 50 41.9 Source: FAO (2020) Note: These figures refer to values recorded in 2017, 2018, and 2019, except the share of dietary energy supply derived from cereals, roots, and tubers, which is for the 2015-17 period. Per capita food supply variability is the standard deviation of per capita food supply (in dietary energy) over the previous five years. 100 WFP, 2020a 101 OCHA, 2019; US Department of State, 2018 124 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure 4.19: Food Insecurity among Rural and Urban Households Food Insecurity among Rural and Urban Households Rural 7 6 11 76 Food secure Mildly food insecure Urban 14 10 14 63 Moderately food insecure Total 8 7 12 73 Severely food insecure 0 20 40 60 80 100 Source: World Bank staff calculation using data from ECOSIT 4 Most Chadian households suffer from severe food Households in the northern and central regions tend to insecurity, especially in rural areas. The incidence of severe be more food secure than households in the south. The food insecurity among rural households is 76 percent, largely agro-pastoral population of northern and central compared with 63 percent or urban households. However, Chad raises livestock for both food and cash income, both mild and moderate food insecurity are more prevalent and some of the cattle produced in Chad is exported to in urban areas. Nationwide, just 8 percent of the population neighboring countries like Nigeria. Households in the south is food secure, but this share ranges from 7 percent in rural primarily produce crops, including staple foods like millet areas to 14 percent in urban centers (Figure 4.19). and rice and cash crops like cotton. The high levels of food insecurity in southern Chad underscore the precariousness of crop production. Even households in urban N’Djamena tend to be less food secure than those in northern Chad, which likely reflects disparities in average productivity. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 125 Figure 4.20. Average Household Food Security across Regions (0 = insecure and 8 = secure) 6.00 5.1 5.00 4.7 5.1 4.2 4.2 4.00 3.3 3.5 3.6 3.00 2.5 2.7 2.9 1.7 1.9 2.0 2.0 2.00 1.2 1.4 1.1 1.1 1.00 0.6 0.8 0.9 0.00 l one Lac Ma tal yen l ari jer- a is yo- éra st Sila ri-B jilé Kan i em h e at jam el Ma Wa ity Keb ira Oua t Enn ou edi st st Bor ï nta Mo ndou irm dda ues Had Bath Lam N'D l Ghaz bi E Enn edi E Oue Bar Salam c k en -Ch yo- di F Cha Tand Ma Gu ide agu ena bi O Ori Keb Occ one Log Log Source: World Bank staff calculation using data from ECOSIT 4 Figure 4.21. Correlates of Being Food Secure (logit marginal effects with 95% confidence intervals) Second wealth quintile Middle wealth quintile Fourth wealth quintile Highestwealth quintile Female headed household Dependency ratio Head - agriculture Head - primary school Head - secondary school Head - handicap Population (ln) Permanent market Asphalt road Rural -.1 0 .1 .2 Source: World Bank staff calculation using data from ECOSIT 4 Notes: ‘Dependency ratio’ is the share of household members under age 15 years or over age 64; ‘Population (ln)’ is the logged population size of the village or neighborhood; ‘Permanent market’ and ‘Asphalt road’ are indicators of the presence of a market or road in the village or neighborhood. 126 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 4.5 EDUCATION, HEALTH, AND FOOD SECURITY DURING THE COVID-19 PANDEMIC Wealthier urban households are the most likely to be INSEED to conduct a nationally representative high- food security, while rural households and female-headed frequency phone survey targeting a subsample of 2,833 households are the most likely to be food insecure. households previously included in the 2018/19 EHCVM. The Households headed by an individual who is physically or first round of the survey took place between May and June mentally handicapped are also especially likely to be food 2020, and the second round followed between July and insecure (Figure 4.21). Rural/urban disparities are also August 2020. significant, but the largest gaps are between wealthier and poorer households. School closures at the height of the COVID-19 pandemic left many students without alternative learning options at Containment measures adopted in response to the home. In May 2020, close to two months after schools were COVID-19 pandemic, including school closures, have had closured, almost 90 percent of students were at home and a deeply negative impact on household welfare.102 In an reported no engagement in educational activities (Figure effort to stop the spread of COVID-19, the government 4.22). This share ranged from 92 percent of students in rural implemented school closures on March 20th, 2020. Schools areas to 72 percent in urban centers. Among those students often constitute both a learning environment and safe who were able to continue their education during the school place for students, and the suspension of school feeding closure, the main forms of learning included home lessons programs threatens the food security of many children. As with parents or private teachers or classes delivered on of August 2020, an estimated 132,357 children—40 percent television or radio. All these methods were most common of whom were girls—were no longer receiving meals at among urban students, though rates of online learning school.103 An e-learning platform for secondary school were low even in urban areas. Very few students received students was set up in April, along with televised broadcasts homework to keep them engaged during school closures. in French and Arabic.104 On June 25th schools were allowed Overall, school closures are expected to result in higher to reopen under the conditions that they respect COVID-19 dropout rates and increased risks of early pregnancy, as prevention measures.105 To capture some of the emerging has been the case during previous emergencies.106 effects of the crisis, the World Bank collaborated with 102 https://www.presidence.td/fr-synth-1111-Vendredi_le_20_mars_2020.html 103 World Food Program, 2020b 104 UNESCO, 2020 105 https://reliefweb.int/report/chad/chad-emergency-update-external-23-june-2020 106 Hallgarten, 2020 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 127 Figure 4.22 Educational Activities during School Closures as of May 2020 92 No activities 73 6 Rural Home lessons 18 Urban 1 Television or radio learning 10 1 Homework 2 0 Virtual learning 2 0 20 40 60 80 100 Source: HFPS May 2020 128 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Box 3: Projected Effects of School Closures Due to COVID-19 The COVID-19 pandemic has caused extensive school closures, both worldwide and in Sahelian countries such as Burkina Faso, Chad, Mali, and Niger. While closing schools may have been effective in combating the spread of COVID-19, millions of students have been out of school for extended periods, and shocks to national education systems will likely have adverse effects on learning as well as lifetime productivity and income. The following figure presents the results of simulations based on two scenarios: the intermediate scenario assumes that schools will be closed for 40 percent of the 2020/21 academic year, while in the pessimistic scenario this share rises to 70 percent. The simulations indicate that school closures due to COVID-19 could result in a loss of between 0.3 and 0.6 learning-adjusted years of schooling in Sahelian countries, as well as a loss of future annual income of US$123.1 to US$212.1. In Chad, the projected loss in learning adjusted school years ranges from 2.2 to 2.4, and the projected loss of annual income ranges from US122 to US$212. Figure 4.23: Projected Effects of School Closures due to COVID-19 in the Sahel Impact on Learning-Adjusted Years of School 5.0 4.0 3.0 2.0 1.0 0.0 Burkina Chad Mali Niger Baseline Intermediate Pessimistic Impact on Annual Income (US$) 3,937 Niger 4,002 4,098 3,899 Mali 3,964 4,057 5,238 Chad 5,328 5,450 6,143 Burkina 6,278 6,460 Baseline Intermediate Pessimistic Source: Azevedo et al. (2020); HCI (2017). INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 129 The COVID-19 crisis is also likely to exacerbate stunting intergenerational consequences for child growth and rates, as well as indicators of acute undernutrition development and negative long-term repercussions for (i.e., wasting), threatening to reverse the gains recently education outcomes, chronic disease risks, and overall achieved in recent decades and undermine progress on human-capital formation. human-capital formation. Estimates from the International Food Policy Research Institute and the World Bank suggest The pandemic has underscored the precarious food- that the pandemic will push between 140 and 150 million security status of Chadian households, but no significant people into extreme poverty by 2021. According to the World increase in food insecurity has yet been observed. The Food Programme, the number of people in lower-middle- high-frequency phone survey included a food-insecurity income countries facing acute food insecurity will nearly module in each round, which captured households’ double to 265 million by the end of 2020. Sharp declines are experiences over the previous month.108 These responses expected in access to child health and nutrition services, were used to categorize households as being food secure similar to those experienced during the 2014–16 outbreak or mildly, moderately, or severely food insecure. In May of Ebola virus disease in West Africa.107 A consortium of 2020, 88 percent of respondents reported being moderately nutrition, economics, food, and health-systems researchers or severely food insecure, but by July this share had fallen recently projected a pandemic-driven increase in acute to 82 percent (Figure 4.24). These results are broadly malnutrition of over 14 percent worldwide, with over consistent with the findings of the 2018/19 ECOSIT 4 survey, 22 percent of the new malnourished children being in which found that about 85 percent of households were Africa. Without adequate action, the profound impact of severely or moderately food insecure. the COVID-19 pandemic on early life nutrition could have Figure 4.24. Household Food Security during the Covid-19 Pandemic Food secure May 2020 4 9 22 66 Mildly food insecure Moderately food insecure July 2020 5 13 16 66 Severely food insecure 0 20 40 60 80 100 Source: HFPS 2020 107 Headey et al. (2020) Impacts of COVID-19 on childhood malnutrition and nutrition-related mortality. The Lancet (Vol 396). 108 While the recall period differs between ECOSIT 4 (last 12 months) and the HFPS (last 30 days), a comparison nevertheless provides some useful insight into the immediate impact of the pandemic. 130 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION In the months following the onset of the pandemic, poorer while the share of food-insecure households in top two households were more likely to attribute food insecurity wealth quintiles that attributed their food insecurity to to the crisis. Two months after the implementation of COVID-19 declined at a marginally faster rate. containment measures, almost 95 percent of food-insecure households in the bottom wealth quintile felt that their The pandemic affected households’ ability to access food insecurity was due to the crisis, versus just 60 percent medical treatment through its impact on incomes. Across of households in the top quintile (Figure 4.25). By July both rounds of the high frequency survey, 80-85 percent 2020, 90 percent of the poorest food-insecure households of households reported not being able to access medical continued to attribute their food insecurity to COVID-19, treatment due to lack of money. COVID-19 itself was not Figure 4.25: Share of Food-Insecure Households that Figure 4.26: Share of Households that Attribute their Attribute their Food Insecurity to COVID-19 Inability to Meet Basic Health Needs to COVID-19 Food insecurity attributed to Covid-19 (%) - Inability to meet basic health needs attributed to Covid-19 (%) 100 94 100 89 87 85 85 84 81 80 82 78 76 78 80 76 80 74 73 72 71 65 68 66 60 60 40 40 20 20 0 0 Lowest Second Middle Fourth Highest Lowest Second Middle Fourth Highest May 2020 July 2020 May 2020 July 2020 Source: HFPS 2020 Source: HFPS May 2020 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 131 a major deterrent to accessing medical treatment. In Households that experienced a reduction in income were May 2020, about 8 percent of households did not access significantly more likely to face challenges meeting their medical treatment because they feared COVID-19, and in basic needs due to the pandemic. The surveys indicate that July no households reported that COVID-19 had negatively a recent loss of income is positively associated with the impacted their access to medical treatment. Households inability to meet basic needs due to COVID-19 (Figure 4.27). in the lower wealth quintiles were more likely to attribute By contrast, residing in a rural area significantly decreases their inability to meet their basic health needs to COVID-19 the likelihood of being unable to meet basic needs due to (Figure 4.26). Households in the higher wealth quintiles COVID-19. The negative impact of COVID-19 on households’ also reported that COVID-19 negatively affected their ability ability to meet their basic needs fell between the surveys, to attend to their basic health needs, but not to the same but the relationship is not statistically significant. extent as their poorer counterparts. Figure 4.27 Correlates of Being Unable to Meet Basic Needs Due to COVID-19 (logit marginal effects with 95% confidence intervals) Female-headed household Dependency ratio Head - agriculture Head primary Head secondary Head - handicap Permanent market Rural Mildly food insecure 2018/19 Moderately food insecure 2018/19 Severely food insecure 2018/19 Income drop July-August 2020 -.2 -.1 0 .1 .2 Source: HFPS 2020 Notes: ‘Dependency ratio’ is the share of household members under age 15 or over age 64; ‘Permanent market’ is an indicator of the presence of a market in the village or neighborhood; ‘Mildly’, ‘Moderately’, or ‘Severely food insecure 2018/19’ refers to the household status as captured in EHCVM 2018/19, with ‘Food secure’ serving as the base group; ‘Income drop’ refers to self-reported lost income since the start of the COVID-19 crisis; and ‘July-August 2020’ is an indicator of the second wave of the HFPS. 132 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 4.6 CONCLUSION AND POLICY RECOMMENDATIONS Chad faces enormous challenges, but a successful post- Ensuring equitable access to school will require addressing pandemic recovery could lay the foundation for robust deep gender disparities in enrollment and completion and sustainable poverty reduction over the long term. rates. Although school attendance is associated with fewer Low levels of human capital are both a cause and a hours spent on domestic work, girls continue to spend twice consequence of poverty, as weak education, health, and as much more time on household chores compared to boys. nutritional outcomes in one generation lead to chronic In addition, many girls denied access to education by their household vulnerability and underinvestment in the families due solely to their gender. An estimated 1.5 million welfare of the next generation, limiting their lifetime girls in Chad are at risk of child marriage and adolescent productivity and hindering their ability to escape poverty. pregnancy, and girls drop out of school at a higher rate While these effects manifest at the household level, than boys. Consequently, more than 85 percent of all Chad’s extremely low HCI scores reflect decades of conflict, Chadian girls between the ages of 10 and 19 are unlikely instability, and weak institutional capacity at the national to reach their full potential. To change social attitudes level, and coordinated interventions by the government towards the role of girls in households, communities, the and its development partners will be necessary to break labor force, and the education system, the government and the cycles of intergenerational poverty that prevent so its development partners should reach out to community much of Chad’s population from reaching its full potential. leaders and engage in public awareness campaigns that To protect and develop Chad’s human capital, policymakers stress the numerous benefits of female education. must simultaneously expand the supply and improve the quality of education and health services. Investments in early childhood development are essential to give children a chance to realize their full potential. As inadequate access is the most binding constraint Research has shown that school attendance among children on education in Chad, the government should consider ages 3-6 provides a critical foundation for both cognitive increasing education spending to at least 20 percent of and social development, and preschool attendance is total spending, the minimum level recommended by the associated with higher chances of future enrollment in Global Partnership for Education. Additional investment primary school as well as better problem-solving abilities, in the sector should focus on building new schools and socio-emotional outcomes, and lifetime earnings.109 Chad’s repairing dilapidated facilities to reduce the average low rates of preschool enrollment suggest that many distance to school. Other forms of support, such as children are unable to receive these benefits, and the subsidized school fees or conditional cash transfers to government’s objectives for the education system should households in the bottom wealth quintiles, could boost include expanded access to early childhood education. enrollment rates among poor and vulnerable groups. Subsidies and transfers could be especially effective in To mitigate the damage to education outcomes caused by rural areas, where the high opportunity cost of forgone the COVID-19 pandemic, the authorities should link social family labor prevents many households from sending their support programs to the reenrollment of school-aged children to school. children. Extended school closures threaten to undo recent 109 Martinez et al. 2012; Yoshikawa and Kabay 2015 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 133 gains in school enrollment rates. To minimize lost education and address the low vaccination coverage rates in the north opportunities, the design of cash transfer programs should of the country could further improve health outcomes. incentivize the continued education of out-of-school students, especially girls. The experience of previous health The outbreak of the novel coronavirus is an opportunity to emergencies, such as the West African Ebola virus disease build back better health systems and more resilient food outbreak, suggest that teachers must be prepared to assist systems. Currently, households bear the majority of health students who have fallen behind academically during the expenditures through direct out-of-pocket payments, and period of school closures. cost is a significant deterrent to accessing care. In this context, it is worth investigating the possibility of offering Recognizing the interdependence of education and health universal health coverage, starting with poor households. outcomes, the World Bank’s Education Strategy 2020 prioritizes early investment in childhood education and Food security is a critical cross-cutting issue that healthcare. Health interventions in Chad should reflect the adversely impacts a wide range of education and health recommendations of the Human Capital Index 2020 Update, outcomes. The severe food insecurity experienced by many which focus on expanding health coverage and improving Chadian households underscores the critical importance of the quality of care, particularly among marginalized increasing agricultural productivity. Despite its vast arable communities, and bolstering nutrition and access to land, Chad does not produce enough food to meet the needs sanitation. Improving the utilization of contraceptives will of its population, and it imports the difference. Expanded be essential to reduce infant and maternal mortality rates access to agricultural inputs and new technologies could while easing pressure on the health system and enabling boost domestic production. Special attention should the country to reap the benefits of a demographic dividend. be paid to female-headed agricultural households and households headed by persons with disabilities, which are Following the Abuja Declaration, the Government of Chad among the most likely to be food insecure. The resumption committed to increase public health spending, but actual of school feeding programs will also provide crucial expenditures on the health sector remain far below the nutritional support to children, especially in rural areas. Abuja target. To maximize the impact of limited resources, the authorities must carefully calibrate personnel spending and investment in health facilities while improving the deployment of health workers. Rural areas are in dire need of health staff and infrastructure, and increased access to skilled health workers could greatly reduce rates of infant and maternal mortality in rural Chad. Investments in health infrastructure and medical supplies could also build trust in the health system and reduce the prevalence of self- treatment. Targeted policies to discourage early marriage 134 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION DATA AND KNOWLEDGE GAPS Effective policies to reduce poverty and boost shared years after the previous national household survey, and in prosperity must be based on credible information the interim the government and its development partners regarding the nature and extent of poverty and inequality, relied on projections based on outdated data to measure as well as their root causes and social consequences. the progress in poverty reduction and shared prosperity. Producing this information requires data and tools to Demographic and households surveys are also infrequent, measure living standards and economic indicators at with gaps far longer than the recommended periodicity of the household, regional, and national levels, ascertain five years. The government has implemented a multiple- whether poverty is largely chronic of transitory, and assess indicator cluster survey to fill the gap, but these data are not the impact of policy interventions on various subsets of yet available. Eleven years after the last census, the census the population.110 Managing-for-Results (M4R) is a public- data are still not publicly available, and the government sector management approach that uses information on has not yet started preparatory work for the next census. performance and results to improve decision making. The Household survey data must be combined with census data approach involves adopting a results-oriented reporting to estimate poverty at sufficiently disaggregated levels to and assessment framework and strengthening the links improve the targeting of policy interventions. Census data between development strategies and budget processes. are also needed to scale survey-based estimates up to the In this context, the ability to produce and disseminate national level and make accurate inferences about poverty timely and reliable information about the economy and for the entire population. the wellbeing of the population is regarded as a crucial component of good policies and institutions. The production of sectoral and administrative data must be strengthened. Despite the economic importance of Chad’s National Five-Year Development Plan (2016-2020) agriculture and livestock, Chad has not conducted an acknowledges the importance of relevant, timely, and agricultural census since 1974. Moreover, annual projections high-quality statistics for monitoring and evaluating of agricultural production require a permanent agricultural national policies. A situational analysis conducted for the survey, but the latest livestock survey was done in 2014 and Chad Statistical Development Project (P159434) uncovered is now obsolete. Similar data gaps affect the economically several important weaknesses in the national statistical critical mining and oil sectors. In addition, the production system. For example, ECOSIT 4 was implemented seven of education and health statistics necessary to monitor 110 World Bank Group. 2015. A Measured Approach to Ending Poverty and Boosting Prosperity: Concepts, Data and the Twin Goals. Washington, DC: The World Bank Group INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 135 human capital accumulation is hindered by a lack of The Chad Statistical Development Project seeks to funding and insufficient qualified staff. strengthen the capacity of INSEED to collect, process and disseminate high-quality data, frequently and in Diversifying data production is necessary to cover new a timely manner. The three basic components of the areas such as enterprises, electricity, digital technologies, project are: (i) strengthening institutional capacity while fragility and violence, and infrastructure. Chad lacks a developing human resources; (ii) improving the statistical permanent enterprises survey to examine the evolution infrastructure for a timely production of high-quality data; of employment and the growth of the private formal and and (iii) improving information technology infrastructure. informal sectors. This information gap must be filled to The results of a midterm evaluation suggest that the ensure that policies support the structural transformation implementation of the project has been satisfactory of the economy. Electricity access is a key development thus far. Lessons from evaluations of similar projects constraint in Chad, while the adoption of digital reveal that statistical capacity-building operations that technologies presents vital opportunities, yet serious seek to transform a national statistical system tend to deficiencies in sector-level data prevent policymakers from have high transaction costs and are slow to produce the fully understanding the dynamics surrounding electricity intended results. However, the authorities should not lose demand and the uptake of digital technologies. sight of critical role that data quality plays in effective policymaking.111 Data constraints also hinder disaster preparedness and response. Each year, Chad faces shocks such as malaria, floods, and seasonal food insecurity that affect large numbers of households, but the lack of early warning systems to inform and guide decision-making does not allow the government to anticipate and proactively react to these situations. The COVID-19 pandemic and its deeply negative impact on household welfare and economic activity has clearly highlighted the importance of such systems. 111 World Bank Group. 2015. 136 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION REFERENCES Abakar, M. F., Seli, D., Lechthaler, F., Schelling, E., Tran, Government of Chad, (2010). Plan stratégique de N., Zinsstag, J., & Muñoz, D. C. (2018). Vaccine hesitancy développement des ressources humaines pour la santé au among mobile pastoralists in Chad: a qualitative study. Tchad (2011-2020). International journal for equity in health, 17(1), 167. Government of Chad, (2019). Rapport sur l’état d’avancement Asfaw, S., Davis, B., Dewbre, J., Handa, S., & Winters, P. (2014), de la mise en œuvre du Plan intérimaire de l’éducation au “Cash transfer programme, productive activities and labour Tchad (PIET) au cours de l’année 2018. supply: evidence from a randomised experiment in Kenya”, The Journal of Development Studies, 50(8), 1172-1196. Greatrex, H., Hansen, J., Garvin, S., Diro, R., Le Guen, M., Blakeley, S., Le Guen M, Rao KN, & Osgood, D. (2015). Baulch, B. and Hoddinott, J., (2000), “Economic mobility and Scaling up index insurance for smallholder farmers: Recent poverty dynamics in developing countries”, The Journal of evidence and insights. CCAFS Report No. 14. Development Studies, 36(6), pp.1-24. Günther, I., Harttgen, K., (2009) “Estimating households’ Calvo, C. and S. Dercon (2013) “Vulnerability to individual vulnerability to idiosyncratic and covariate shocks: A novel and aggregate poverty”, Social Choice and Welfare, 41(4), p. method applied in Madagascar”, World Development, 37(7), 721-740. 1222-1234. Carter, M. R., & Barrett, C. B. (2006), “The economics of Hallgarten, J. (2020). Evidence on efforts to mitigate the poverty traps and persistent poverty: An asset-based negative educational impact of past disease outbreaks. K4D approach”, The Journal of Development Studies, 42(2), 178- Helpdesk Report 793. Reading, UK: Education Development 199. Trust. Available at https://www.gov.uk/research-for- development-outputs/evidence-on-efforts-to-mitigate- Demographics and Health Survey (DHS). n.d. Stat Compiler the-negative-educational-impact-of-past-disease- website. Accessed September 13, 2020 at www.statcompiler. outbreaks. com. Hansen, J., Hellin, J., Rosenstock, T., Fisher, E., Cairns, J., Stirling, Gavi (The Vaccine Alliance), (2019). Joint appraisal report. C., ... & Campbell, B. (2019), “Climate risk management and Available at https://www.gavi.org/news/document-library/ rural poverty reduction”, Agricultural Systems, 172, 28-46. joint-appraisal-chad-2019. Hoddinott, J., Quisumbing, R., (2003) “Methods for Gehrke, E., & Hartwig, R. (2018), “Productive effects of public Microeconometric Risk and Vulnerability Assessments”, works programs: What do we know? What should we know?”. Social Protection discussion Paper Series, No. 0324, World World development, 107, 111-124. Bank. Government of Chad (GoC), (2017). Plan National de Hoddinott, J., (2006), “Shocks and their consequences across Développement (PND 2017-2021) – République du Tchad. and within households in rural Zimbabwe”, The Journal of Development Studies, 42(2), 301-321. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 137 International Monetary Fund (IMF), (2010). Chad: poverty Stoeffler, Q., Carter, M., Guirkinger, C., & Gelade, W. (2020). reduction strategy paper. IMF Country Report No. 10/230. “The spillover impact of index insurance on agricultural Available at https://www.imf.org/en/Publications/CR/ investment by cotton farmers in Burkina Faso.” NBER Issues/2016/12/31/Chad-Poverty-Reduction-Strategy- Working Paper w27564. Paper-24073 UNESCO, (2020). National learning platforms and Jensen, N. D. & C. B. Barrett (2017). “Agricultural Index tools. Available at https://en.unesco.org/covid19/ Insurance for Development.” Applied Economic Perspectives educationresponse/nationalresponses and Policy, 39.2, pp. 199-219. UNHCR, (2020). Chad emergency update – external. Available Martinez, S., S. Naudeau, and V. Pereira. 2012. The promise at https://data2.unhcr.org/en/documents/download/77305 of preschool in Africa: A randomized impact Evaluation of early childhood development in rural Mozambique. World United Nations Children’s Fund (UNICEF), (2016). Maternal Bank: Washington, D.C. and newborn health disparities – Chad. Available at https:// data.unicef.org/wp-content/uploads/country_profiles/ Mills, A., Kurowski, C., Wyss, K., & Abdulla, S. (2010). Human Chad/country%20profile_TCD.pdf Resources for Health: Requirements and Availability in the Context of Scaling-Up Priority Interventions in Low-Income US Department of State, (2018). Lake Chad Basin: Ongoing Countries (No. id: 2386). food security and displacement crisis. Available at https:// hiu.state.gov/#AF,U1780 Obiang-Obounou, B. W., & Fuh, M. E. (2020). The State of Maternal and Infant Health and Mortality in Chad. African World Bank, (2018). The Human Capital Project. World Bank: Journal of Reproductive Health, 24(1), 26-34. Washington, D.C. Available at https://www.worldbank.org/ en/publication/human-capital. OCHA, (2019). Chad humanitarian situation overview. Available at https://www.humanitarianresponse.info/en/ World Bank, (n.d.) Tchad - Francophone Africa Results operations/chad/infographic/tchad-aper%C3%A7u-de-la- Monitor – Basic education. situation-humanitaire-novembre-2019 World Bank, 2019. Chad public expenditure analysis – Fiscal PASEC (Progamme d’Analyse des systèmes éducatifs de space for productive social sectors expenditure. la confemen), (2014). Education system performance in Francophone sub-Saharan Africa. Available at http:// World Bank, (2020a). The Human Capital Index 2020 update: www.pasec.confemen.org/wp-content/uploads/2015/12/ Human capital in the time of COVID-19. World Bank: Rapport_Pasec2014_GB_webv2.pdf Washington, D.C. Available at https://elibrary.worldbank. org/doi/abs/10.1596/34432. Sahel Adaptive Social Protection Program: Annual Report 2019. Available at http://documents.worldbank.org/ World Bank, (2020b). World Bank Data Bank. Accessed 25 curated/en/680361585895594749/Sahel-Adaptive-Social- September 2020 at https://data.worldbank.org/indicator Protection-Program-Annual-Report-2019 138 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION World Food Programme (WFP), (2020a). WFP Chad country brief – August 2020. Available at https://reliefweb.int/ report/chad/wfp-chad-country-brief-august-2020 World Food Programme (WFP), (2020b). Global monitoring of school means during COVID-19 school closures. Available at https://cdn.wfp.org/2020/school-feeding-map/?_ ga=2.23544890.1608590319.1587311560-1934460547.1538727870 World Health Organization (WHO), (2019). Available at https://www.who.int/gho/maternal_health/countries/tcd. pdf?ua=1 Yoshikawa, H., and S. Kabay. 2014. The Evidence Base on Early Childhood Care and Education in Global Contexts. Background paper commissioned for the EFA Global Monitoring Report 2015, Education for All 2000-2015: Achievements and Challenges. UNESCO: Paris, France. Zimmermann, L. (2014), “Public works programs in developing countries have the potential to reduce poverty” IZA World of Labor INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 139 ANNEX A: TECHNICAL NOTE ON POVERTY MEASUREMENT BASED ON ECOSIT 4 (2018/19) DATA The main objective of this Harmonized Living Conditions January to April 2019. Each wave collected half the sample. Household Survey (EHCVM) is to build capacity in the A break of five months has been observed between the two design, implementation, processing and analysis of survey periods. The two-wave approach was chosen in order to data for poverty assessment. The households survey was take into account of the seasonality of consumption (both conducted in a sample of 7,500 households in the 23 in terms of habits and levels of consumption). administrative regions of Chad. The sampling method used for the data collection is a two-stage stratified survey. At the Poverty measurement is the process of generating poverty first stage, 625 enumeration areas were randomly selected indicators from survey data. Poverty measurement based on their size. An enumeration was then made in each involves three steps: (i) constructing an indicator for enumeration area before randomly selecting 12 households measuring welfare; (ii) constructing a poverty line; and from the list of enumerated households. In some (iii) aggregating the data to produce poverty indicators. enumeration areas, household replacements have been This note explains the methodological choices made for made to compensate for refusals of response or absences the measurement of poverty. The first section explains the from households. A maximum of two replacements was approach used to construct the consumption aggregate allowed per enumeration area. since the welfare indicator used is normalized annual per capita household consumption. The second section In order to obtain information on the community in which explains the methodological approach used to construct the households live and the prices that are applied in the poverty line. The third section analyzes the transition their markets, a Community questionnaire and three price from the consumption aggregate to the welfare indicator by records of the main products were carried out in each applying different deflators. Lastly, initial results are briefly enumeration area. Ecosit 4 (2018/19) data were collected in presented in the fourth section. two waves, the first from May to July 2018, the second from 140 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure A1 summarizes the components of poverty measurement. • Food purchases • Food from home production Food consumption • Food received from others • Food consumed away from home • Household expenses Non-food • Personal expenses consumption • Health expenses • Education expenses Consumption aggregates • Purchase value Value of usage of • Replacement value durable assets • Depreciation rate • Interest rate Poverty Measures • Lodging characteristics Imputed rent • Location • Food basket Food poverty lines • Calories requirement Poverty lines • Food poverty line Total poverty line • Non-food poverty line • Consumer Price Index Time deflator • Non-food poverty line • Adjusting costs of Spatial deflator living across space INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 141 1. CONSUMPTION AGGREGATE The consumption aggregate represents annual household possible to derive the unit value of acquisition. If the consumption. It is calculated by aggregating food product was purchased more than 30 days before the consumption, non-food consumption in non-durable interviewer’s visit, the value of the purchase is not provided goods and services, the use value of durable goods, and the and therefore no unit value can be obtained. Obviously if imputed rent of owner-occupied and rent-free households. consumption of the product within the household comes exclusively from auto-consumption and gifts, no unit value Food consumption is measured over the last seven days is available either. (the reference period) preceding the enumerator’s visit. It is the sum of household food consumption at home Moreover, since consumption is reported in non-standard (purchases made and actually consumed, self-consumption units (NSU) during data collection, we must also find a of the household’s own production, and gifts received way of converting these NSUs into standard units (SU) and actually consumed) and meals taken away from before applying prices. Obviously, if the data on NSUs is of home. In this survey, food consumption in the household average or poor quality, this also affects the quality of the is measured in quantities and meals taken outside the consumption data obtained after valuation. household are reported as values. Food consumption within the household is annualized by multiplying the quantities consumed by 365/7. THREE SCENARIOS HAVE BEEN TESTED TO VALUE FOOD CONSUMPTION WITHIN THE HOUSEHOLD The trickiest question therefore concerns the valuation of household food consumption (purchases, auto- Scenario A: Unit values combined with market prices. For consumption and gifts). The survey is designed to use two a given product, when a household purchased the product price vectors: the unit values of the products purchased, and during the last 30 days prior to the enumerator’s visit, the the prices recorded in the markets of the localities where valuation of consumption (including auto-consumption the sampled households live. The first set of information and, if applicable, gifts) is calculated using the unit value of (unit values) is available if the product was purchased by acquisition. In this case, there is generally no need to convert the household within 30 days prior to data collection. When quantities into SUs provided the unit of acquisition and the a product was purchased during this period, in addition to unit of consumption are the same, which happened here providing information on consumption broken down into in two thirds of cases. If the units are different, conversion purchases, auto-consumption, and gifts, the questionnaire to SUs is necessary before consumption can be valued.112 also provides information on the last purchase (quantity For households that did not purchase the product within purchased and corresponding value), which makes it the last 30 days prior to data collection, valuation is based 112 The data collected has ¬¬the following characteristics: for two-thirds of all consumption observations, the unit of consumption and the unit of acquisition are the same (it is therefore not useful to use NSUs); for about 6 percent of all observations, the two units differ, and for 26 percent of all observations, only the unit of consumption is available as there have been no acquisitions in the last 30 days. 142 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION on market prices. This is done sequentially. Consumption Scenario C: Unit values only. This scenario consists of is first valued using the average price calculated by using unit values exclusively instead of prices. For a given regional and residential area; if the price is available at product, when a household has acquired the product by this geographical level for this product, the calculation purchase in the last 30 days before the survey agent’s visit, ends. If the information is missing at the previous level, the the valuation of consumption (including auto-consumption average price calculated at the level of the Agro-Ecological and, if applicable, gifts) is calculated using the unit value Zone (AEZ) and place of residence is used.113 If information of acquisition, as in Scenario A. For households that have is still missing at this level, the price calculated by place of not purchased the product in the last 30 days prior to residence (urban or rural) at the national level is used. If collection, valuation is calculated using the unit values of the information is missing at the previous level, the average households that have purchased the product. A vector of price for the region is used followed by the average price unit values is constructed using all possible combinations for the AEZ, and finally the national price. It is important to of product and unit of acquisition. The sequence of the note that all prices are calculated by wave. In other words, quantity valuation process is the same as in the previous Wave 1 prices are not used to value the quantities of a scenario, the only difference being that market prices are household surveyed in Wave 2 and vice-versa. replaced by unit values. As information on the product or unit combination is used, conversion factors for converting Scenario B: Market prices only. In this case, consumption NSUs to SUs are not needed. of a given product is first valued using the average price calculated for the region/residential area; if the price is On the basis of these two assessment criteria, it follows available at this geographical level for this product, the that Scenario B should not be selected. Meanwhile, further calculation is ends. If the information is missing at the work is being done using Scenarios A and C. Since the previous level, the average price calculated at the level of quantities valued are the same for all three scenarios, the the Agro-Ecological Zone (AEZ)/ residential area is used. If weakness of Scenario B lies in the quality of the price data, information is still missing at this level, the price calculated which may not take sufficient account of differences in by place of residence (urban or rural) at the national level product quality. The other possible difficulty with Scenario is used. If the information is missing at the previous level, B is the use of NSUs, another data source that necessarily the average price for the region is used followed by the introduces noise. This will play a role in the choice between average price of the AEZ, and finally the national price. As Scenarios A and C. Since the latter has the advantage of not before, it should be noted that all prices are calculated by using NSUs, this is the scenario chosen here. wave. It is also important to note that if the choice is made to value quantities by market price, the conversion factors from NSUs to SUs must always be used. 113 An AEZ consists of a grouping of regions. This variable was created to be used in the construction of poverty lines and has proved useful in that the use of a single region does not always give robust results because the number of observations may be too small. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 143 Meals taken away from home. Consumption of meals on festivities and ceremonies as well as pilgrimages, which taken away from home is given in terms of value for are considered exceptional expenditure, are also excluded. the last seven days before the interviewer’s visit to the The only case in which holiday expenditure is used is for household. It is provided for each individual (for meals expenditure on clothing and footwear for religious holidays taken individually) and for the household as a whole for such as Christmas, New Year, Easter, end of Ramadan, meals taken collectively by several household’s members. Tabaski, etc. The reason for this choice lies in the fact The total value declared by the household is annualized by that the clothing acquired during these holidays is real multiplying it by 365/7. household consumption and not prestige or conspicuous expenditure, which amounts to a transfer to other households. It is also important to stress the classic debate At the end of the above valuation process, consumption over whether expenditure on education (school fees, taken within the household is added to the meals taken costs of supplies, etc.) and health (consultations, medical away from home to give the total food consumption of the examinations, medication, hospitalization) constitutes household. investment in human capital or consumption. The choice was made to include them, as has long been the practice Non-food consumption. Non-food consumption of non- in WAEMU member countries. Nevertheless, expenditure durable goods and services (including education and health) on therapeutic medical devices (crutches, wheelchairs, is measured in value terms over a reference period of 7 days, dentures, prescription glasses, etc.) was excluded from 30 days, 3 months, 6 months, and 12 months depending the consumption aggregate. Even if these items were to be on the anticipated frequency of consumption of each type included, they would be treated like durable goods. of good. The value reported during the reference period is multiplied by a factor taking into account the frequency, or Use value of durable goods. Durable goods are those that 365/7, 12, 4, 2, and 1, respectively. The important point is to render services to the household over a long period of time define durable goods as well as exceptional expenditure after their acquisition. For these goods, the use that is made in order to exclude them in the aggregation of non-food of them is considered consumption by the household. consumption. Durable goods are defined as means of It is therefore necessary to estimate this consumption, transportation (car, motorcycle, bicycle, etc.), household which is called “use value.” All goods regarded as durable appliances (television, refrigerator, freezer, oven, washing goods have been defined above. In addition, real estate machine, dishwasher, air conditioner, music system, radio, (land, buildings) and goods mainly intended for economic fans, etc.), large pieces of furniture (sofa and armchair set, production (dugout canoes and outboards, hunting rifles, dining table and chairs, bookcase, other cupboards, etc.), etc.) were ignored. and electronic appliances and other goods (computer, telephone, mobile phone, cameras, musical instruments such as guitar or piano, motorized gardening equipment, For goods regarded as durable, the use value is a function valuable jewelry and watches, carpets, etc.). These goods of the acquisition value, the age of the goods, the inflation are excluded from the calculation of food consumption rate, the real interest rate, and economic depreciation. The and will be calculated by use value instead. Expenditure acquisition value and the age of the good were provided 144 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION during the survey, an annual inflation rate of 1% and a real years (less than 3% of observations), age was limited to 20 interest rate of 2% were used for all durable goods, and years; (iii) when the number of goods was not reported and the only unknown parameter was the depreciation rate. other information was present, the number of observations For each good and each household owning it, if vrempla of the good was imputed by the mode value; and (iv) the is the value of the asset replacement cost, vacqui is the acquisition value of outliers was adjusted before proceeding acquisition value, and age the age of the asset in whole with the calculations (see adjustment of outliers below). years, the formula for calculating the depreciation rate (depret) is as follows: Imputed rent of owner-occupied households. The final component of the consumption aggregate is the imputed rent of owner-occupied and rent-free households. For households, housing is an investment good; when a The median depreciation rate (mdpret) of the asset for household has built a dwelling, it consumes it by occupying all households is then calculated. Finally, if s12q03 is the it. The general approach for estimating imputed rent is the number of goods of a given type owned by the household econometric approach. In some cases, where the number and s12q08 is the acquisition price of such goods, the use of observations was too small, an alternative approach was value of a given good (depan) is obtained by applying the chosen, which is explained below. following formula: The econometric approach is based on the following principle: since some households are renters, a hedonic housing function is estimated for these households, and this function is used to impute a notional rent to owner- The sum of this variable (depan) for all assets owned by a occupied and rent-free households. The explained variable household provides the aggregate of the use value of the of the model is the logarithm of the rent, the explanatory household’s durable goods. variables typically being: type of dwelling, number of rooms, type of walls, type of roof, type of floor, type of It is important to note that adjustments are made to the toilet, presence of electricity in the dwelling, presence of data before calculation: (i) for goods less than one year running water in the dwelling, mode of garbage disposal, old, age was assumed to be 0.5; (ii) for goods older than 20 mode of sewage disposal, and other community variables INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 145 such as the existence of a paved road in the locality, the abnormally low values, and abnormally high values. Values most common mode of transportation in the locality, etc. that are too low are defined as zero food or zero non-food The model is estimated using the stepwise procedure, consumption. Household consumption was calculated which consists of gradually introducing the variables into according to four main consumption functions: (i) food the model and retaining only those that are significant. consumption, including meals taken away from home; (ii) non-food consumption without use value of durable To account for differences in the housing market, the model goods and imputed rent; (iii) use value of durable goods; is estimated separately for the country’s capital city, other and (iv) imputed rent. Households with a zero i or a zero urban areas, and rural areas. In the capital and other ii component were removed from the databases. The logic urban areas, the econometric approach is systematically is simple: it is unlikely that a household has zero food implemented. In rural areas, the housing market is tight, consumption; rather, this household did not complete as shown by the low number of renter households in the the interview (in cases of a one-person household, samples. Thus, the econometric approach cannot produce where the householder is often absent) or refused to satisfactory results in rural areas. An alternative approach complete it. Similarly, a household cannot have zero is used. This consists of calculating the median rent of annual non-food consumption, whatever its standard of tenants according to the number of rooms, and this rent living. It is necessary to buy even basic goods for everyday is imputed to owner-occupied households occupying a consumption (household soap, matches, etc.). As a second dwelling with the same number of rooms. Here, given the step, an adjustment was made for abnormally large values. small number of tenant households, the number of rooms Contrary to the previous case, this adjustment is made variable is recoded into three modalities, for example (1 per consumption item. The logarithm of the consumption room, 2 rooms, and 3 or more rooms) and the interquartile range are then calculated. A value was considered abnormally large if it is greater than the median of the logarithm of consumption plus 2.5 times Outlier adjustment. It is always difficult to distinguish the interquartile range (this value is called the “maximum between what is an outlier (abnormally high or abnormally allowed”). These values are adjusted by replacing the value low value) and what is simply an atypical value. Improper with the maximum allowed (or “trimming”). This choice is outlier adjustment can reduce real inequalities in the made to minimize the impact on inequality. population. For this reason, adjustments should be made with caution. Here, adjustments were made in two stages: 146 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 2. POVERTY LINE The poverty line is the value of the welfare indicator that Having adopted a reference population, a national basket allows individuals to satisfy their minimum vital needs. The was constructed covering 85% of the most consumed food approach used to construct the poverty line is that of the products in this reference population, excluding meals cost of basic needs.114 A poverty line is constructed in two away from home.115 The basket was constructed based steps: (a) calculating the food poverty line; (b) deriving a not on nominal expenditure but on annual expenditure total poverty line by applying to the food line a share of for each product adjusted by the spatial deflator (see the non-food expenditure. calculation of the spatial deflator in Section 3). This ensures that differences in prices do not affect the procedure for With regard to the food poverty line, a basket of food constructing the poverty line. Before finalizing the basket, items providing each individual with 2,300 kilocalories it was verified that it represented at least 70% of the food (which is within the range of the internationally accepted consumption in each region or Agro-Ecological Zone (AEZ). standard for food consumption) is determined. The To obtain the food line, the basket was valued using unit valuation of this basket provides the food poverty line. values from the consumption records, the same unit Three factors are important in carrying out this task: (i) values that were used to value food consumption. These the reference population for determining the basket; (ii) unit values are filled in during the survey as different non- how the basket is constructed; and (iii) the price vector standard consumption units (bottle, basin, plate, heap, used to value the basket. etc.). The unit values collected in NSUs are then converted to SUs using the conversion factors from the NSU survey that took place before the main data collection. On the first point, the reference population must be households around the poverty line. The objective is to have a reference population that has, as much as possible, the consumption habits of households that are neither too poor nor too well-off. Given that the poverty lines are around 40% in the subregion, the interval from the second to third and the seventh to eighth decile is an acceptable range. Deciles 3 to 8 were used for all countries. 114 Ravallion, Martin. 1998. Poverty lines in theory and practice (English). Living standards measurement study (LSMS) working paper ; no. LSM 133. Washington, D.C. : The World Bank. 115 Meals taken outside of the household cannot be used in the construction of the poverty line. The reason is that the process requires a correspondence between quantities consumed and calorie intake, and this information is not available for meals taken outside. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 147 Two non-food poverty lines were calculated, and these led For the second of the two thresholds (zref_max), households to two overall poverty lines. However, the non-food poverty around the food poverty line are defined as those with lines were not calculated directly; instead the total poverty food consumption within plus or minus 10% of the food line is calculated using the share of food consumption of poverty line; as before, if there are no households in this households around the food poverty line. The first option range, households falling within plus or minus 20% of the consists of determining the non-food component of the food poverty line are used. If we call alpha_max the share poverty line as the share of non-food consumption of of household food consumption whose per capita food households whose total consumption is equal to the consumption is just equal to the food poverty line as defined poverty line. The second is to determine the non-food above, the maximum non-food poverty line is given by: component of the food poverty line as the share of non- food consumption of households whose food consumption is equal to the food poverty line. The second solution clearly gives a higher value than the first. For the first of the two poverty lines (zref_min), households around the food poverty line are defined as those with total consumption within plus or minus 10% of the food poverty line; if there are no households in this range, households within plus or minus 20% of the food poverty line are used. If we call zali the previously calculated food poverty lines, and alpha_min the share of household food consumption whose total per capita consumption is just equal to the food poverty line as defined above, the minimum poverty line is given by: 148 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 3. HOUSEHOLD COMPOSITION, TEMPORAL AND SPATIAL DEFLATORS, AND INDICATORS OF WELFARE The consumption aggregate is not an indicator of welfare place in the period well away from the harvest. During the because it does not allow for a fair comparison between collection period, consumer prices changed. Consumption households. Households are of different sizes and was normalized using a time index. To do this, the national compositions and face different prices depending on when household final consumer price index is an effective tool. the data was collected and where household members live. Chad has regional indices that could have been used as The welfare indicator must therefore take all these factors part of this process, but their coverage is limited as they into account. tend to focus on Ndjamena, with secondary cities and rural areas less well represented. Household composition and size. The first element to consider is the size and composition of households. Here, To calculate time deflators, if we call IPCi the consumer household composition was ignored and only size was price index at month i, i=1, ..., n the period of n collection taken into account. Household composition should be months, we can calculate IPC as the average index during reflected by an equivalence scale, and there is no consensus the collection period by: regarding the best approach to deriving an equivalence scale. Moreover, virtually all countries concerned have adopted the practice of only taking household size into account. Thus the consumption aggregate is divided by The time deflator for each collection month is given by: household size to yield annual per capita consumption. Nevertheless, for the purpose of carrying out sensitivity tests such as ranking regions in terms of poverty levels, two equivalence scales were calculated since it is easy to produce poverty figures using either one. It was pointed out above that the time deflator is applied to the annual consumption of each product before the Time deflator. The second element to consider is the construction of the poverty line. Thus, for a household k time at which the data is collected in the household. Nine surveyed in month i, the annual expenditure of product m months elapsed between the start of data collection and (depan) is normalized by the following formula: completion of this process. Data collection for the first wave took place in several of the countries immediately following the harvest, and data collection for the second wave took INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 149 Spatial deflator. It is advisable to also apply a spatial poverty line, given that the average of the minimum and deflator so as to take into account disparities in the cost maximum poverty lines was used as the non-food poverty of living between different regions and localities in the line, the same approach was followed. If we call zzaej the country. A natural candidate is the regional price index, or poverty line of the AEZ and areas j, def_spa the spatial at least the prices underlying these calculations. However, deflator of the AEZ and areas j is the ratio of the threshold as noted above, prices collected at regional level show low of AEZ or area j to the national threshold: coverage of small urban centers and rural areas. A test was conducted in order to use these as a deflator. As poverty rates of over 70% were obtained in some countries, the idea was abandoned. The poverty lines constructed by AEZ Finally, for a household k surveyed in month i and belonging and area of residence were used as spatial deflators. The to AEZ/area of residence j, if we call dtotk the total annual approach to constructing poverty lines by AEZ and area consumption of the household and hhsizek the household of residence was the same as that for constructing the size, the welfare measure indicator is: national poverty line and the same national basket was used. This basket was valued using the average unit values of the AEZ and areas for the food poverty line. The non-food poverty line was also constructed by AEZ and area using the same approach as above. In other words, for the non-food 150 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 4. SYNTHESIS OF RESULTS The indicator of poverty easiest to calculate is the incidence The ECOSIT 4 survey is important for Chad. The survey was of poverty, which is the percentage of people living below designed not only to produce poverty indicators but, more the poverty line. The incidence of poverty depends on importantly, to generate data for in-depth analytical work the chosen poverty line. Each country should have a designed to assess poverty in its many dimensions. On national poverty line. This line is important for monitoring the basis of the production of the poverty figures, one of and evaluating public policies to combat poverty in the the objectives of the project has been achieved; it is now country. The line depends on national standards, including a matter for the countries concerned to add value to the consumption preferences and the cost of living. The data. Nevertheless, it is important to stress that survey data incidence of poverty according to the national poverty line is never perfect, especially in African countries, where the is that used for poverty diagnoses in national documents level of literacy remains average. In this project, the price such as development plans or poverty reduction strategies. data is of average quality. An assessment of the survey will However, this poverty incidence is not directly comparable allow for lessons to be learned from this first round in the with that of any other country because it depends on the work in order to improve the next one. national poverty line, which takes into account specific norms and preferences. For international comparisons and the monitoring of the Sustainable Development Goals (SDG), international poverty lines are more appropriate. The extreme poverty line is US$1.90 per person per day at 2011 purchasing power parity (PPP). Here, this line was converted in FCFA taking into account the increase in the cost of living as measured by inflation between 2011 and 2018. It is important to note that the first SDG target (Eradicating Extreme Poverty by 2030) uses the above extreme poverty line. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 151 ANNEX B: SURVEY-TO-SURVEY IMPUTATION METHODOLOGY The survey to survey imputation follows the methodology With log yij the log per capita consumption and Xij a set proposed by Elbers, Lanjouw, and Lanjouw (2003). It is a of characteristics of household i in cluster j. The error two stages methodology. In the first stage, a model of log term is composed of two independent components, that is per capita household consumption is estimated based ϵij=μj+εij. μj is a cluster specific error, and εij is a household on the 2018 EHCVM data, the survey for which the welfare specific or idiosyncratic error term. The regression results indicator is available. The welfare indicator is regressed on are reported in Table B.2 below. a set of explanatory variables including demographic, labor, and housing characteristics. These explanatory variables In the second stage, the estimated parameters of the first are selected such that they are available and comparable stage model are used to impute household consumption over the two surveys, the EHCVM and the DHS (used at the in the 2014 DHS survey. This is possible since the same second stage). The list of the explanatory variables included explanatory variables used for the estimation are also in the regression is provided in Table 1 below. These available in the DHS survey. This step accounts for the explanatory variables are selected from a larger list given stochastic nature of the parameters, especially the error the significance of their parameter in the regression model. component. Indeed, an imputed value of consumption for each household is the sum of predicted consumption and The first stage model accounts for the design of the error term. The latter is randomly drawn from the empirical household survey. It allows the clustering of households distribution of the stochastic error. This imputation is and their dependence on one another within a cluster, repeated 100 times, and at each time, poverty indicators the primary sampling unit of the two surveys. As a result, are estimated. The mean and the standard deviation over the stochastic error component of the model can be the 100 simulations are treated as the point estimate of decomposed into a cluster-specific effect and a household poverty and its standard error. specific-effect. The parameters of the model are estimated through Generalized Least Square (GLS) to account for heteroscedasticity. The model is given by 152 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Table B.1 : Summary statistics for households and communities Variables Year Mean Standard error Share of dependent 2014 0.572 0.0028 0.870 0.875 0.884 Household has electricity 2014 0.089 0.0100 2018 0.094 0.0075 Household size 2014 7.879 0.0967 2018 6.883 0.0815 Roof is in straw 2014 0.695 0.0175 2018 0.643 0.0158 Household has a car 2014 0.027 0.0033 2018 0.020 0.0025 Household has a radio 2014 0.439 0.0116 2018 0.228 0.0090 Squared household size 2014 80.533 2.7574 2018 59.467 1.9730 Household has no toilet 2014 0.690 0.0182 2018 0.645 0.0169 Number of people per room 2014 3.407 0.0329 2018 3.112 0.0475 Household has a well 2014 0.717 0.0139 2018 0.747 0.0145 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 153 Variables Year Mean Standard error Household has a TV 2014 0.092 0.0099 2018 0.063 0.0057 Household head is female 2014 0.172 0.0062 2018 0.180 0.0068 Household head has a primary education level 2014 0.227 0.0101 2018 0.149 0.0086 Household members aged 15-16 2014 3.201 0.0498 2018 2.837 0.0430 Household has flush toilet 2014 0.015 0.0024 2018 0.016 0.0028 Household head has a secondary education level 2014 0.160 0.0074 2018 0.135 0.0083 Banco wall 2014 0.114 0.0084 2018 0.656 0.0153 Ground floor 2014 0.869 0.0103 2018 0.914 0.0073 Has refrigerator 2014 0.023 0.0040 2018 0.024 0.0030 Household members aged 65+ 2014 0.161 0.0062 2018 0.128 0.0062 Age of household head 2014 44.187 0.2130 2018 43.808 0.2358 Wall is in concrete 2014 0.215 0.0133 2018 0.027 0.0032 154 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Table B.2 : GLS and OLS regression results of per capita consumption Variables bGLS bOLS Age of household head -0.002 -0.001 Household head is female -0.100 -0.102 Household head has a primary education level -0.037 -0.091 Household has a well -0.086 -0.092 Household has electricity 0.169 0.195 Number of people per room -0.041 -0.028 Share of household members aged 15 to 64 -0.046 -0.053 Share of household members aged 65+ 0.057 0.063 Roof made in mud -0.121 -0.070 Household has a TV 0.137 0.152 Household has flush toilet 0.161 0.209 Household has no toilet -0.067 -0.079 Roof is in straw -0.168 -0.185 Household has a car 0.549 0.590 Household has a radio 0.167 0.172 Household has refrigerator 0.192 0.137 Share of dependent -0.679 -0.724 Household size -0.081 -0.076 Squared household size 0.002 0.002 zone -0.091 -0.072 _cons 14.162 14.044 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 155 ANNEX C: MULTIDIMENSIONAL POVERTY IN CHAD Multidimensional poverty is an expansive measure of are equally weighted within their respective dimensions. wellbeing that extends beyond income per capita. The The MPI score is the product of the incidence of deprivation multidimensional poverty index (MPI) described below (H) and the average intensity of deprivation (A). More tracks deprivation along six dimensions: (i) education, specifically, H represents the share of the population that is (ii) health, (iii) childhood and youth, (iv) access to basic deprived under at least one-third of the six dimensions of services, (v) housing conditions, and (vi) assets (Table C.1). the MPI, and A is the extent of the deprivation experienced The six dimensions are equally weighted, and all indicators by deprived households. Table C.1 : MPI : Dimensions, Indicators, Deprivation, and Weights Dimension Indicator Definition of Deprivation Weight Household members over age 15 have an average of less than nine years of Years of schooling 1/12 schooling Education Any household member over age 15 years cannot read and write 1/12 School attendance Household has at least one child between age 6 and 16 not attending school 1/18 Household has at least one child between age 7 and 17 with fewer completed Children and youth Child labor 1/18 years of schooling than the national average Household has at least one child between age 12 and 17 who works 1/18 Improved cooking fuel Household uses solid fuels and biomass (e.g., charcoal, wood, etc.) for cooking 1/24 Household lacks access to an improved drinking-water source (piped water, Improved water source 1/24 borehole, public taps/standpipes, protected well) Access to basic services Improved sanitation Household’s sanitation facilities do not include flush toilets or latrines 1/24 Electricity Household lacks access to an electricity source (grid, generator, solar) 1/24 156 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Dimension Indicator Definition of Deprivation Weight Floor made of natural or low-quality materials, including mud, wood, straw, Improved floor 1/24 metal sheet, sand, dung, etc. Improved wall Walls made of natural or low-quality materials 1/24 Housing conditions Improved roof Roof made of natural or low-quality materials 1/24 Critical overcrowding Four or more individual per room 1/24 At least one household member required healthcare in the past 30 days but did Healthcare access 1/18 not seek care from a physician, specialist, or any health institution At least one household member experienced an illness in the last 30 days and Health Health specialist 1/18 did not seek specialized services Household disposes of domestic waste by natural or rudimentary means, Waste management 1/18 including incineration or open dumping Household does not own more than one modern asset (radio, TV, telephone, Modern asset Assets Source: World Bank staff calculation using data from ECOSIT 4 bike, motorbike, refrigerator, computer, or animal cart) and does not own a car 1/6 ownership Note: HH=household or truck INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 157 C.1 INCIDENCE AND PROFILE OF MULTIDIMENSIONAL POVERTY Multidimensional poverty is widespread in Chad. About 89 Educational deprivation is the largest contributor to percent of the population is deprived in at least one-third Chad’s high MPI scores, followed by inadequate access of the MPI’s six dimensions of wellbeing. Multidimensional to basic services. Educational deprivation contributes 21 poverty rates higher for women than men, and poor percent to overall multidimensional poverty in Chad, and women suffer more intense deprivation than their male inadequate access to basic services contributes almost 20 counterparts (Table C.2). Multidimensional poverty rates percent. Greater investment in education infrastructure and vary by location, ranging from 39 percent in N’Djamena to basic services could substantially reduce multidimensional 78 percent in other urban centers and reaching 97 percent poverty, especially in rural areas. in rural areas. Moreover, the rural poor suffer more intense deprivation than their counterparts in N’Djamena and in other urban areas. Table C.2: Multidimensional Poverty in Chad, 2018 MPI Headcount Rate Average Intensity Chad 58.7 88.8 66.1 Female 62.8 91.5 68.7 Gender Male 57.7 88.2 65.5 N’Djamena 18.2 38.6 47.2 Location Other Urban 43.3 77.9 55.6 Rural 66.8 97.3 68.7 158 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Households headed by administrative or service workers Educational attainment is strongly associated with are less likely to experience multidimensional poverty multidimensional poverty. MPI scores range from 65.8 than are households headed by agricultural workers. Most percent among households headed by a person with no agricultural workers live in rural areas with limited access education to 11.8 percent among households headed by to infrastructure and services, education, healthcare, and a person with university education. Moreover, low levels quality housing. Moreover, many smallholder farmers and of educational attainment are also correlated with more rural laborers are also poor in monetary terms, which intense deprivation: MPI intensity ranges from 69 percent prevents them from acquiring modern assets. As a result, among multidimensionally poor households headed by households headed by agricultural workers face intense a person with no education to 42 percent among those deprivation and are at high risk of multidimensional poverty. headed by a person with a university education. Households headed by manufacturing, construction, or transportation workers also have high multidimensional poverty rates, as these sectors often employ unskilled informal workers at low wages in relatively costly urban areas (Figure C.1b). Figure C.1: Multidimensional Poverty in Chad (a) MPI Decomposition 18.6 20.6 Education Childhood and youth 10.9 Health 16.8 Access to basic services Housing conditions 13.5 Assets ownership 19.6 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 159 (b) MPI by Economic Sector of Household Head Agriculture 66.7 Manufacturing 54.6 Construction 49.1 Transport 46.9 Mining 45.7 Trade 45.5 Hotel 41.0 Services 31.2 Education/Health 27.0 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 (c) MPI by Employment Status of Household Head home worker/trainee 68.9 Self employed 62.6 Unqualified worker 48.8 Qualified worker 29.6 15.5 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 (d) MPI by Education Level of Households Head No education 65.8 Primary incompleted 59.2 Primary completed 50.8 Lower secondary 40.1 Hupper secondary 33.0 University 11.8 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 Source: World Bank staff calculation using data from ECOSIT 4 160 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION The incidence and intensity of multidimensional poverty and least intense in N’Djamena. Education is the main vary across regions. The MPI headcount poverty rate is contributor to multidimensional poverty in all regions lowest in N’Djamena, while the rate in the Soudanian zone except Logone, Mandoul, Mayo-Kebbi, Ouaddaï, and Tandjilé, is below the national average, and the rate in the Saharan which suggests that the government’s strategy for reducing zone is above it. The Lac Region has the highest MPI poverty multidimensional poverty should prioritize investments rate, followed by the Sila Region. MPI poverty rates appear in education. In addition, the authorities should invest in to be correlated with the intensity of multidimensional basic services, as inadequate service access contributes poverty: deprivation is most intense in the Lac Region substantially to multidimensional poverty in all regions. Figure C.2: Multidimensional Poverty Rates by Region 80 74.2 67.9 69.6 69.6 70.1 70.9 71.0 71.0 70 64.5 65.1 65.4 65.9 57.1 57.2 57.8 57.8 58.8 58.9 60.1 60 51.7 50 40 30 20 18.2 10 0 Keb na Mo Ouest Log Ta ri Occ jilé yo- tal ri-B Est Ma i one oul al edi u Oua t Sal a h-E t l em Had Guéra is ha Sila Lac Wa aï irm aza s Bar ama ir Enn orko a Lam ent Oue dd yo- e Bat Ma iden -Ch di F one nd Kan Log nd Cha ebbi Ma N’djam agu l-G Ori B jer- yen bi K e ed Vill MPI - 2018 National MPI Source: World Bank staff calculation using data from ECOSIT 4 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 161 C.2 OVERLAPPING DIMENSIONS OF POVERTY Almost all Chadians in poor households experience and lack access to basic services would address multiple deprivation in basic education and health. Just under 93 dimensions of poverty as measured by both human and percent of the poor population is considered deprived physical capital accumulation. These households represent in indicators of both education and health. Members of almost one-third of Chad’s total population. poor households who suffer from educational and health deprivation represent about 40 percent of Chad’s total Educational deprivation is closely correlated with population. The large overlap between poverty status and deprivation in the childhood and youth dimension of deprivation in education and health indicators highlights poverty. Two-thirds of people from poor households, or the pro-poor impact of investment in social services. 29.5 percent of the total Chadian population, are deprived Maximizing the impact of limited fiscal resources will require in these two dimensions of poverty. Almost one-quarter integrated interventions designed to simultaneously of people from poor, education-deprived households, or improve the supply of both education and health services. about 10 percent of the population, are not deprived in the childhood and youth dimension of poverty. Nevertheless, About 78 percent of the poor population is deprived in the substantial overlaps between poverty, education, and asset ownership and access to essential services. The health and between poverty, education, and childhood remaining 22 percent also lack access to basic services but and youth development suggest that policies designed to live in households that own several modern assets such as improve human capital should systematically integrate mobile phones or vehicles. Due to the near-total overlap multiple dimensions of deprivation, especially childhood between poverty and deprivation in health and education, and youth development. interventions targeting poor households that own few assets Figure C.3 : Overlapping Dimensions of Poverty (a) Education, health, and monetary poverty National Education Health 47.5 1.3 8.2 Education deprived Health deprived 39.4 Monetary poor 2.7 0.0 Monetary poor 162 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION (b) Access to basic services, asset ownership, and monetary poverty National Asset Access to basic services 32.5 0.3 21.2 Asset deprived Access deprived 33.1 Monetary poor 8.9 0.2 (c) Education, monetary poverty, and childhood and youth National Childhood and youth Education 27.5 3.3 21.3 Childhood deprived Education deprived 29.5 Monetary poor 1.3 10.1 1.4 Monetary poor INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 163 C.3 TRENDS IN MULTIDIMENSIONAL POVERTY While the incidence of multidimensional poverty declined The decline in MPI scores reflected significant progress between 2003 and 2018, much of the country continues to in improving housing conditions and asset ownership face severe deprivation. The national MPI score fell from among the country’s poorest households. The material 70.3 percent to 58.7 percent over the period, while the MPI quality of housing and the rate of asset ownership are headcount poverty rate declined by 8.7 percentage points the two MPI indicators that experienced the largest gains (Table C.3). Meanwhile, the intensity of multidimensional over the last 15 years. Between 2003 and 2018, the share poverty fell by 6 percentage points, from 72.1 percent of multidimensionally poor households with low-quality in 2003 to 66.1 percent in 2018, indicating that living walls and roofs declined by 20 percentage points and 13 conditions improved among households experiencing percentage points, respectively, and the share that owned multidimensional poverty. Multidimensional poverty at least some modern assets increased by 17 percentage reduction accelerated over the period, and the national MPI points (Figure C.5). score fell faster between 2011 and 2018 than between 2003 and 2011. Moreover, the decline in MPI poverty rates was significant at all poverty thresholds (Figure C.4). Table C.3 : Multidimensional Poverty in Chad in 2003, 2011 and 2018 MPI H A 2003 58.7 97.5 72.1 2011 62.8 95.2 69.6 2018 57.7 88.8 66.1 Change 2003-2011 (in percentage points) 18.2 -2.2 -2.5 Change 2011-2018 (in percentage points) 43.3 -6.4 -3.6 Source: World Bank staff calculation using data from ECOSIT 4 164 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure C.4: MPI Headcount Poverty Rates at Different Poverty Thresholds 80.0 60.0 40.0 20.0 0.0 0.07 0.13 0.20 0.27 0.33 0.40 0.47 0.53 0.60 0.67 0.73 0.80 0.87 0.93 1.00 2003 2011 2018 Source: World Bank staff calculation using data from ECOSIT 4 Figure C.5: Deprivation by Indicator among Multidimensionally Poor Households, 2003, 2011, and 2018 74.1 37.0 63.7 Nutrition 59.6 Asset 91.9 49.7 28.9 49.6 room 32.7 29.0 34.3 Health specialist 43.2 71.8 Housing roof 85.3 31.5 85.0 51.7 68.2 36.0 Housing Wall 95.7 88.9 21.5 96.2 Child labour 25.2 Housing Floor 95.7 12.3 95.0 96.4 62.0 Electricity 98.5 Child grade 66.4 97.3 73.8 71.3 47.3 Sanitation 74.4 School attendance 51.5 68.4 55.9 42.6 Water 62.8 88.7 52.0 Literacy 95.6 97.7 92.3 Cooking fuel 99.3 75.7 89.2 75.2 94.5 years of schooling Waste 98.4 77.6 95.0 2018 2011 2003 2018 2011 2003 Source: World Bank staff calculation using data from ECOSIT 4 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 165 Pro-poor nutrition and education policies also contributed and cooking fuel increased. Finally, the deprivation rate to the decline in multidimensional poverty. The share for electricity declined by a mere 1 percentage point, of multidimensionally poor households suffering from highlighting the persistent structural constraints on challenges related to nutrition, school attendance, and grade electricity access in Chad. repetition fell significantly between 2003 and 2018. While education indicators improved among multidimensionally Gains in multidimensional poverty have been concentrated poor households, Chad’s poor population continued to face in N’Djamena. Between 2003 and 2018, the capital’s MPI high levels of deprivation in terms of years of schooling headcount poverty rate fell by half, while the average (75.7 percent) and literacy rates (88.7 percent) in 2018. intensity of multidimensional poverty dropped from 53.6 percent to 47.2 percent (Table C.4). While MPI poverty rates Progress in improving access to basic services among also fell in other urban areas during the period, poor urban multidimensionally poor households has been mixed. The households outside the capital suffered more intense deprivation rate for healthcare fell by 4.5 percentage points deprivation in 2018 than in 2003. MPI headcount poverty between 2003 2018, but this improvement was more than rates and the average intensity of multidimensional poverty offset by a drop in access to health specialists, resulting in have improved more slowly in rural areas, which remain the an overall deterioration in the health component of the MPI. poorest parts of the country. While the deprivation rate for access to improved water fell by 9.3 percentage points, the deprivation rate for sanitation Table C.4 : Multidimensional Poverty in Urban and Rural Areas 2003 2011 2018 Absolute change 2003-2018 MPI H A MPI H A MPI H A MPI H A N’Djamena 41.4 77.2 53.6 32.6 65.7 49.6 18.2 38.6 47.2 -23.2 -38.6 -6.4 Other Urban 46.5 86.7 53.6 49.2 85.0 57.8 43.3 77.9 55.6 -3.1 -8.8 2.1 Rural 73.5 99.5 73.9 71.8 99.5 72.2 66.8 97.3 68.7 -6.7 -2.2 -5.2 Source: World Bank staff calculation using data from ECOSIT 4 166 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Both the MPI headcount poverty rate and the intensity of are in Logone Oriental; and 5 percent are in Moyen- multidimensional poverty have declined across all regions Chari. The Batha Region is home to just 2 percent of of Chad over the last 15 years. Regions that experienced multidimensionally poor households, but it has the second the highest rates of multidimensional poverty reduction, highest MPI headcount poverty rate in the country. Batha including N’Djamena, Logone Oriental, Mayo-Kebbi, and experienced the smallest decline in the MPI poverty rate Moyen-Chari, also experienced a significant decline in the observed between 2003 and 2018, and the intensity of intensity of deprivation (Figure C.6 and Table C.5). These multidimensional poverty remained virtually unchanged. regions are home to most of the country’s poor population: Chad’s poorest region, Lac, is also one of the most conflict- 12 percent of multidimensionally poor households are affected areas of the country, and it has experienced no in Mayo-Kebbi; 9.6 percent are in N’Djamena; 7 percent significant improvements in its MPI indicators since 2011. Figure C.6: Multidimensional Headcount Poverty Rates by Region 72.6 73.1 77.3 80.0 72.9 75.3 74.6 71.4 71.2 77.1 72.9 73.8 71.8 68.2 71.0 69.7 71.5 70.9 70.0 67.7 69.2 66.3 67.5 68.7 69.8 67.3 67.4 65.5 64.8 65.4 57.8 60.1 58.0 60.0 55.2 57.2 50.0 41.4 40.0 32.6 30.0 18.2 20.0 10.0 0.0 ha Fira i t c tal l bi ari é ena î irm nta dda ma /La djil Keb Bat den -Ch jam di- a orie agu Tan em Oua Sal yo- yen cci a kan N'D /w ri B ra/ one Ma O Mo ine Cha ne Gue Log lt o Log /Bi BET 2003 2011 2018 Source: World Bank staff calculation using data from ECOSIT 4 Table C.5 : Multidimensional Poverty Indicators by Region 2003 2011 2018 Absolute change 2003-2018 Pop. Pop. Pop. MPI H A MPI H A MPI H A MPI H A share share share Batha 6.5 72.6 99.5 73.0 3.7 72.9 98.8 73.8 4.2 71.0 97.5 72.9 -1.6 -2.0 -0.2 BET/Biltine/Wadi-Fira 4.3 75.3 99.2 76.0 5.6 69.7 98.7 70.7 7.2 65.5 96.5 67.9 -9.8 -2.7 -8.0 Chari Baguirmi 10.0 74.6 100.0 74.6 11.8 67.7 99.8 67.9 10.2 64.8 95.5 67.8 -9.8 -4.5 -6.8 Guéra/Salamat 6.7 73.1 99.2 73.7 7.8 72.9 99.1 73.6 6.7 69.2 97.6 70.9 -3.9 -1.6 -2.8 Kanem/Lac 8.7 77.3 100.0 77.3 8.4 71.4 99.0 72.1 9.3 71.5 99.4 72.0 -5.8 -0.6 -5.3 Logone Occidental 7.0 66.3 97.0 68.3 6.8 67.5 96.3 70.1 6.3 57.8 90.5 63.9 -8.5 -6.4 -4.5 Logone Oriental 7.9 73.8 99.9 73.9 8.1 68.7 98.5 69.8 7.1 60.1 93.9 64.0 -13.7 -6.0 -9.9 Mayo-Kebbi 12.1 69.8 99.5 70.1 13.8 65.4 95.4 68.5 12.2 55.2 91.1 60.6 -14.5 -8.4 -9.5 Moyen-Chari 11.8 71.2 98.1 72.6 11.3 70.9 97.6 72.6 11.1 58.0 90.4 64.2 -13.2 -7.7 -8.4 Ouaddaï 10.1 77.1 99.8 77.3 9.1 71.8 96.5 74.3 10.0 67.3 93.2 72.2 -9.8 -6.5 -5.1 Tandjilé 7.3 68.2 97.8 69.7 5.5 67.4 98.6 68.4 6.1 57.2 93.3 61.3 -11.1 -4.5 -8.5 N'Djamena 7.6 41.4 77.2 53.6 8.2 32.6 65.7 49.6 9.6 18.2 38.6 47.2 -23.2 -38.6 -6.4 Source: World Bank staff calculation using data from ECOSIT 4 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 167 168 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION ANNEX D: DRIVERS OF POVERTY REDUCTION To explore the basic factors behind the decline in poverty, changes in household consumption have been decomposed into (1) improvements in household characteristics or where is the θth unconditional quantile of log real endowments, such as more education of the head of the — per adult monthly household consumption, X the vector household, ownership of assets, and access to employment of characteristics averages, and the estimate of the opportunities and basic services; and (2) changes in the unconditional quantile partial effect. Superscripts i, i’ and rewards or returns that they get for those characteristics like * designate respectively the final year (2018 or 2012), initial returns to education, assets productivity, and return or profit year (2012 or 2007), and counterfactual values. to business. The two components have themselves been decomposed to identify specific attributes that contribute is the counterfactual quantile of the unconditional to changes in consumption, and the decomposition has counterfactual distribution; it represents the distribution been applied to each decile of the consumption distribution of welfare that would have prevailed if the relationship to understand differences in the patterns of change for between endowments and consumption had remained different income groups. constant over time. It is used to determine which changes in endowments could have helped to reduce poverty, and The approach is based on the Recentered Influence how poverty reduction could have changed as a result Function (RIF) and unconditional quantile regression of a changing relationship between consumption and method proposed by Firpo, Fortin, and Lemieux (2009), endowments. Changes in return to endowments represents in which traditional Oaxaca-Blinder decompositions are the variation of the conditional correlation between a given applied to the consumption distribution by percentile. This endowment and consumption over time. The decomposition makes it possible to assess the amount of poverty reduction works as follows: attributable to changes in the endowments of households and the amount due to changes in the Chadian economy and economic returns to people’s endowments: Table D.1 : Detrminants of change in consumption 2011-2018 at national and Rural level (Endowment and Returns effects) Extreme poor Poor Middle class Richest Chad Rural Chad Rural Chad Rural Chad Rural Total 0.808*** 0.697*** 0.480*** 0.494*** 0.408*** 0.434*** 0.271*** 0.218*** Endowments 0.117 -0.018 0.269** 0.298* 0.230** 0.210 0.196 0.032 Head employed 0.012* 0.016* 0.007 0.002 0.009* 0.004 0.007 0.003 Head second job 0.022*** 0.028*** 0.017*** 0.026*** 0.006 0.008 0.008 0.030*** Head self employed 0.032** 0.034 0.015 0.036 0.004 0.003 0.009 -0.051 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 169 Extreme poor Poor Middle class Richest Head farm employment -0.039*** -0.044** -0.026*** -0.041*** -0.027*** -0.028** -0.026** -0.026 Livestock 0.008*** 0.007*** 0.007*** 0.010*** 0.005*** 0.007*** 0.007*** 0.010*** Head nonfarm employment -0.004 0.002* 0.003 0.001 0.004* 0.002 0.003 0.003 Head education 0.001 0.001 0.003* 0.004 -0.003 0.001 0.005*** 0.001 Own mobile phone 0.143*** 0.110*** 0.135*** 0.106*** 0.096*** 0.089*** 0.102*** 0.010 Own car/motorcycle 0.000 -0.001 0.001 -0.001 0.001* -0.001 0.007* -0.001 Access to electricity 0.001 0.000 0.005** -0.001 0.006*** -0.000 0.023*** 0.007*** Access to improved water 0.002* 0.000 0.003*** 0.002* 0.004* 0.002** 0.006*** 0.004** Share of workers in the HH 0.007*** 0.008*** 0.008*** 0.005** 0.004** 0.004** 0.010*** 0.006* Returns 0.692*** 0.715*** 0.212* 0.195 0.178 0.224 0.075 0.185 Head employed -0.039 0.037 0.016 -0.064 0.001 -0.045 -0.089 -0.443*** Head second job 0.002 0.007 0.028** 0.027 -0.012 -0.001 -0.007 -0.001 Head self employed 0.028 0.032 -0.023 0.013 -0.024 -0.018 -0.035 -0.146** Head farm employment -0.025 -0.083 -0.025 0.008 -0.018 0.035 0.046 0.451*** Livestock 0.001 -0.001 0.003* 0.005 -0.000 0.002 0.005*** 0.005 Head nonfarm employment -0.037** -0.012* -0.016 -0.003 -0.001 0.003 0.054*** 0.045*** Head education -0.086*** -0.047** -0.040*** -0.067*** -0.012 -0.018 0.005* -0.016 Own mobile phone 0.018*** 0.014*** 0.012*** 0.007* 0.001 0.002 -0.002 -0.053 Own car/motorcycle -0.001 0.005 0.003 0.003 0.011** 0.004 0.024*** 0.023** Access to electricity 0.006 0.004* 0.003 0.002 0.002 0.001 0.004 0.004* Access to improved water 0.009 0.005 0.002 0.006 0.001 -0.001 0.023*** 0.015* Share of workers in the HH 0.042* 0.008 0.058*** 0.052 0.053*** 0.063** 0.070*** 0.080* Source: ECOSIT3 and ECOSIT4. Note: Extreme poor are population groups in the bottom 10 -20 percent of the distribution; the poor are in the third-fourth decile; middle class are in the fifth decile, and the richest are in the top decile. * Significant at the 10 percent level; ** significant at the 5 percent level; *** significant at the 1 percent level. Numbers in parentheses are bootstrap standard deviations based on 100 replications. 170 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Table D.2 : Detrminants of change in consumption 2011-2018 at Ndjamena and other urban level (Endowment and Returns effects) Extreme poor Poor Middle class Richest Other Other Other Other Ndjamena Ndjamena Ndjamena Ndjamena Urban Urban Urban Urban Total 0.537*** 0.311*** 0.444*** 0.236*** 0.403*** 0.211*** 0.266*** 0.304*** Endowments 0.264*** 0.563* 0.137*** 0.373* 0.150*** 0.262 0.034 -0.078 Head employed -0.010 0.003 0.005 0.002 0.012 0.012 -0.024 -0.011 Head second job 0.022* 0.000 0.017* -0.000 0.005 -0.001 -0.015 -0.005 Head self employed 0.042* -0.003 0.037** -0.002 0.018 -0.000 -0.005 0.001 Head farm employment -0.045*** 0.006* -0.052*** 0.003 -0.025*** 0.003 0.017 -0.002 Livestock 0.008** 0.001 0.007** 0.001 0.005** 0.000 0.002 -0.000 Head nonfarm employment 0.014* 0.004 0.003* 0.002 0.010** -0.001 0.018** 0.004 Head education 0.002 0.002 -0.001 0.003 -0.003 0.004 0.003 0.005 Own mobile phone 0.297*** 0.287** 0.167*** 0.092 0.143*** -0.022 0.053 0.163 Own car/motorcycle 0.001 0.010*** 0.001 0.013*** 0.000 0.022*** -0.009 0.040*** Access to electricity 0.003 0.044** 0.002 0.021* 0.003 0.038*** 0.003 0.052** Access to improved water -0.006* 0.028*** -0.006*** 0.032*** -0.003 0.019*** -0.000 0.021** Share of workers in the HH 0.016** 0.010** 0.014*** 0.007** 0.010** 0.008** 0.038*** 0.019*** Returns 0.274*** -0.252 0.307*** -0.137 0.253*** -0.051 0.232*** 0.382 Head employed 0.078 0.089 -0.100 -0.022 -0.028 0.034 -0.129 0.204* Head second job -0.009 -0.018* -0.023* -0.004 -0.023** -0.012** -0.016 -0.012 Head self employed 0.054 -0.026 0.068*** 0.013 0.033 0.016 -0.014 -0.012 Head farm employment -0.022 -0.031*** -0.028 -0.021*** -0.009 -0.011* 0.042 -0.013 Livestock -0.002 0.001 -0.001 0.001 0.000 -0.001 -0.001 -0.003 Head nonfarm employment 0.019 0.047* -0.022*** 0.021 -0.000 0.003 0.041* -0.018 Head education -0.068** 0.025 -0.037* -0.037 -0.024 0.070*** -0.035 0.060* Own mobile phone 0.047*** 0.005 0.024*** -0.021 0.021*** -0.040** -0.002 0.022 Own car/motorcycle -0.023* 0.032*** -0.003 0.018** 0.003 0.030*** 0.026*** 0.058*** Access to electricity 0.007 0.017 -0.003 -0.020 0.004 0.004 -0.006 0.044 Access to improved water -0.001 0.017 -0.004 -0.031 -0.003 -0.034* -0.014 0.006 Share of workers in the HH 0.038 0.023 0.052** -0.018 0.040* -0.001 0.114*** 0.026 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 171 ANNEX E: VULNERABILITY AND SHOCKS I. ESTIMATING VULNERABILITY TO POVERTY Vulnerability estimation follows the methodology in With logcij the log per capita consumption and Xij a set of Gunther and Harttgen (2009), which is an extension of that characteristics of household i in community j. eij reflects the proposed by Chaudhuri (2002). Chaudhuri’s methodology unexplained variance in consumption across households, is based on two main assumptions. The first is that the while the two error terms, u0j and u1j , are level two residuals unexplained part (error term) of a consumption model capturing the unexplained variance in consumption across captures the impact of household-specific and community- communities. Zj is a set of community characteristics, and specific shocks on consumption. And the second hypothesis Zj Xij is a set of interaction terms for which β11 is significant. is that the error term is correlated with observable The final list of explanatory household and community household and community characteristics. The extension variables included in the model with their descriptive consists of incorporating multilevel analysis to Chaudhuri’s statistics is given in Table 1 below. methodology. The multilevel analysis has two advantages. It allows decomposing the shocks to consumption into a The model described in the first step allows estimating two household component and a community component. Also, types of error terms, a household level error, eij , capturing it accounts for the hierarchical structure of the data and the idiosyncratic shocks and community level errors, u0j thus corrects for inefficient estimators. and u1j , capturing the covariate shocks. In the second step, the squared of the error terms are regressed on a set of The extended methodology consists of a two steps household and community characteristics. That is estimation procedure. In the first step, a multilevel model of log per capita household consumption is estimated on a set of community and household characteristics. Besides, the model includes interactions between household and community characteristics as explanatory variables. But, only the interaction terms for which the coefficients are significant (at the 10% level) are kept in the final multilevel This two steps procedure allows estimating the expected model which is given by mean as well as the idiosyncratic, covariate, and total variances of households’ consumption. The regression results are reported in Table B.1 below. 172 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION The estimates of the multilevel model are then used to A household is considered vulnerable when the above measure vulnerability. More specifically, the probability of probability is higher or equal to 20%. Moreover, the current a household falling below the poverty line is computed framework allows decomposing vulnerability into two given the assumption that consumption is log-normally sources, poverty and risk. The former, poverty induced ˆ is computed as: distributed. That is the probability P vulnerability, is defined as the category of households for which the expected mean of consumption is already below the poverty line. The latter, risk induced vulnerability, captures the group of households having their consumption above the poverty line but a probability of falling below the poverty line higher than the threshold of 20%. With z the poverty line, ĉij the estimated expected mean consumption and σ ̂ij2 the idiosyncratic consumption variance. Table E.1 : Summary statistics for households and communities Individual level characteristics National Urban Rural Area of residence 0.8 Household size 6.9 7.4 6.7 Living in capital city 0.1 0.4 Squared age of household head 2099.4 2215.4 2061.9 Proportion of household members aged between 0 and 14 0.5 0.5 0.6 Household headed by a female 0.2 0.2 0.2 Household head works in agriculture 0.7 0.2 0.8 Household head is unemployed 0.1 0.1 0.1 Household head has a higher education level 0.0 0.1 0.0 Assets score 0.0 1.2 -0.4 Total cattle owned by household 8.0 4.2 9.3 Household head has a bank account 0.1 0.2 0.0 hhi (diversity of food item) 0.1 0.1 0.2 Log of food consumption expenditure 11.9 12.4 11.8 Community level characteristics Proportion of household head working in agriculture 0.7 0.2 0.8 Proportion of firm owner household head 0.8 0.6 0.9 Proportion of household head having a bank account 0.1 0.2 0.0 infrastricture score 0.0 0.7 -0.2 Availability of electricity network 0.1 0.5 0.0 Availability of drinking water network 0.2 0.5 0.1 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 173 Table E.2 : Regression results of per capita consumption (two-level model) Variables (log) Per capita expenditure Household size -0.010*** (0.001) Residence in rural areas -0.026 (0.021) capital -0.273*** (0.015) Square age of household head (years) -0.000*** (0.000) Number of children under 14 -0.220*** (0.012) Female-headed household -0.031*** (0.006) Works in agriculture/livestock/fishing (HH head) -0.027*** (0.007) Unemployed (HH head) 0.007 (0.008) head_tertedu 0.036** (0.011) Asset index 0.050*** (0.006) Number of cattle 0.001*** (0.000) HH has a bank account 0.074*** (0.012) HHI (diversity of food items) -0.512** (0.161) (log) Annual per capita food consumption 0.772*** (0.005) % working in agriculture -0.115*** (0.027) % enterprise owner -0.032 (0.035) % bank account -0.005 (0.062) Infrastructure index 0.001 (0.001) Electricity in a community 0.011 (0.012) Drinking water in a community 0.028** (0.010) % working in primary sector * asset score 0.133*** (0.013) 174 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Variables (log) Per capita expenditure % working in primary sector * rural area 0.131*** (0.031) % bank owner * food diversity 1.406*** (0.352) % bank owner * household size -0.013* (0.006) % electricity in the community * asset score -0.014* (0.006) % enterprise owner * food diversity 0.442* (0.177) Var(e) 0.031 R^2(0) 0.889 #Obs(HH) 7265.000 Var(u) 0.007 R^2(1) 0.960 #Obs(Community) 606.000 Standard errors in parentheses * p<0.05 ; ** p<0.01; *** p<0.001 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 175 II. VULNERABILITY TO POVERTY ACROSS REGIONS There is high regional disparity in vulnerability to poverty rates also displayed poverty rates hovering around 60% in Chad with vulnerability rate varying from 70 percent to while in N’Djamena less than 15% of household were below 10 percent across regions. The highest rates of vulnerability the poverty line. The difference of more than 10% between to poverty, above 60%, were registered in Tandjilé, Mandoul, the poverty and vulnerability rates in Logone Occidental, Mayo-Kebbi Ouest, Guéra and Mayo-Kebbi Est (Figure A-1). and Kanem indicates that many of the currently non-poor Households in Ennedi (11%) and N’Djamena (10%) exhibited households are likely to fall into poverty in the event of the lowest vulnerability rates in the country. The trend a shock. Across all the regions, covariate shocks induced across regions was largely similar when considering poverty vulnerability rates are moderately greater than idiosyncratic rates. In effect, the five regions with the highest vulnerability shocks induced vulnerability (Figure E.1). Poverty rate Figure E.1. Vulnerability rates vary considerably across regions Vulnerability rate (total) 60 40 Percentage (%) 20 0 Barh-El-Gazal N'djamena Sila Moyen-Chari Wadi Fira Logone Oriental Mandoul Kanem Logone Occidental Chari-Baguirmi Guéra Batha Borkou Tandjilé Hadjer-Lamis Lac Ennedi Ouest Salamat Mayo-Kebbi Ouest Mayo-Kebbi Est Ouaddaï 176 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Figure E.2. Covariate shocks induced vulnerability is prevalent across regions 60 40 Percentage (%) 20 0 Barh-El-Gazal Sila Moyen-Chari N'djamena Wadi Fira Logone Occidental Logone Oriental Mandoul Kanem Chari-Baguirmi Batha Borkou Guéra Tandjilé Hadjer-Lamis Lac Ennedi Ouest Salamat Mayo-Kebbi Ouest Mayo-Kebbi Est Ouaddaï Risk-induced vulnerability was more pronounced across Vulnerability rate regions in Chad. The share of households affected by risk (covariate) induced vulnerability ranged from 8% to 60% while poverty induced vulnerability rates fluctuated between 3% and 10%. Vulnerability rate This indicates that for many households with sufficient (idiosyncratic) human and capital endowments, future exposure to shocks increases their likelihood of falling into poverty. Households in Logone Oriental and Mayo-Kebbi Est more often faced risk induced vulnerability than poverty induced vulnerability. On the other hand, their peers in Ennedi Ouest, Borkou and N’Djamena were more exposed to poverty induced vulnerability compared to risk induced vulnerability. INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 177 III. EXPOSURE TO SHOCKS ACROSS REGIONS Regions across Chad were relatively similar in terms of their exposure to shocks. More than 80% of households in 17 out of the 21 regions were exposed to shocks (Figure E.3). Among the three regions the most impacted by the occurrence of shocks, namely Hadjer-Lamis (99.5%), Mandoul (99%), and Guéra (98%), the incidence of covariate shocks was greater than idiosyncratic shocks. This was not however a generalized pattern among the regions exhibiting a high rate of exposure to shocks. Indeed, between 95% and 97% of households in Logone Occidental, Logone Oriental and Moyen-Chari who experienced shocks were more vulnerable to idiosyncratic shocks compared to covariate ones. Figure E.3. There were no major differences across regions Any shock in terms of their exposure to shocks Covariate shock Idiosyncratic shock 100 80 60 Percentage (%) 40 20 0 Barh-El-Gazal Sila N'djamena Moyen-Chari Wadi Fira Logone Oriental Mandoul Kanem Logone Occidental Chari-Baguirmi Batha Guéra Borkou Tandjilé Hadjer-Lamis Lac Ennedi Ouest Salamat Mayo-Kebbi Ouest Mayo-Kebbi Est Ouaddaï 178 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION ANNEX F: ADDITIONAL DATA ON VULNERABILITY IN CHAD Figure F.1 Risk induced vulnerability across regions 60 40 Percentage (%) 20 0 Barh-El-Gazal N'djamena Sila Moyen-Chari Wadi Fira Logone Oriental Mandoul Kanem Logone Occidental Chari-Baguirmi Guéra Batha Borkou Tandjilé Hadjer-Lamis Lac Ennedi Ouest Salamat Mayo-Kebbi Ouest Mayo-Kebbi Est Ouaddaï Poverty induced vulnerability Risk induced vulnerability Source: EHCVM 2018/19 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 179 Table F.1 Proportion of households exposed to shocks Shocks National Urban Rural Severe illness or injury of a HH member 39% 38% 40% Drought / irregular rainfall 24% 11% 27% Death of a HH member 23% 26% 22% High food prices 20% 25% 19% Theft of money, assets, production or other goods 13% 18% 11% farmer-pastoralist conflict 11% 5% 13% Flooding 10% 11% 10% High rate of crop disease 9% 4% 11% High rate of animal disease 8% 3% 9% Locust attacks or other pests 6% 2% 7% Divorce/separation 5% 7% 4% Other shocks 4% 4% 4% Fire 4% 3% 4% High input prices 3% 1% 4% Important output price drop 2% 1% 3% Bankruptcy of nonfarm enterprise 2% 4% 2% Important loss of nonfarm income 2% 2% 2% Loss of wage employment 2% 5% 1% Armed conflict/violence/insecurity 1% 1% 1% Important loss of salary incomes 1% 3% 0% End of regular transfers from other HHs 1% 1% 1% Landslide 0% 0% 0% 180 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Table F.2 Proportion of female and male headed households exposed to shocks Shocks Male headed Female headed Severe illness or injury of a HH member 39% 41% Death of a HH member 21% 28% Drought / irregular rainfall 24% 24% High food prices 21% 18% Divorce/separation 3% 11% farmer-pastoralist conflict 12% 10% Theft of money, assets, production or other goods 13% 10% Flooding 11% 9% High rate of crop disease 10% 6% High rate of animal disease 9% 5% Locust attacks or other pests 6% 4% Other shocks 4% 3% Bankruptcy of nonfarm enterprise 2% 3% Fire 4% 3% High input prices 4% 2% End of regular transfers from other HHs 1% 2% Important output price drop 3% 1% Important loss of nonfarm income 2% 1% Loss of wage employment 2% 1% Armed conflict/violence/insecurity 1% 1% Important loss of salary incomes 1% 1% Landslide 0% 0% INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 181 Table F.3 Proportions of households exposed to shocks following the outbreak of the coronavirus by residence zones National Urban Rural High food prices 69% 50% 75% Illness of an income earner in the household 18% 14% 19% Bankruptcy of nonfarm enterprise 14% 18% 13% Loss of wage employment 8% 18% 5% High input prices 8% 5% 8% Theft of money, assets, production or other goods 6% 6% 5% Death of an individual who sends money to the household 5% 5% 5% Bad harvest owing to lack of labor 4% 1% 5% Death or disability of an active adult household member 4% 5% 4% Other 3% 4% 2% Important output price drop 2% 2% 2% Loss of an important acquaintance 1% 2% 1% Locust attacks or other pests 0% 0% 0% 182 INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION Table F.4 Proportions of households exposed to shocks following the outbreak of the coronavirus by poverty status and gender of the head of household Non poor households Poor Female headed Male headed High food prices 66% 73% 65% 69% Illness of an income earner in the household 17% 21% 18% 18% Bankruptcy of nonfarm enterprise 14% 14% 17% 14% Loss of wage employment 8% 7% 8% 8% High input prices 6% 11% 8% 8% Death of an individual who sends money to the household 5% 6% 6% 5% Theft of money, assets, production or other goods 6% 4% 5% 6% Death or disability of an active adult household member 4% 6% 5% 4% Important output price drop 2% 2% 2% 2% Bad harvest owing to lack of labor 5% 4% 2% 5% Other 3% 2% 1% 3% Loss of an important acquaintance 1% 1% 1% 1% Locust attacks or other pests 0% 0% 0% 0% INVESTING IN RURAL INCOME GROWTH, HUMAN CAPITAL, AND RESILIENCE TO SUPPORT SUSTAINABLE POVERTY REDUCTION 183 Table F.5 Coping strategies following the outbreak of the coronavirus Female Male National Urban Rural Non poor Poor headed headed Reducing consumption 27% 18% 30% 26% 29% 18% 29% Using savings 27% 43% 22% 29% 23% 26% 27% Help from family or friends 14% 16% 13% 15% 13% 32% 11% Selling livestock 9% 2% 11% 7% 11% 5% 9% Buying cheap food 8% 7% 8% 7% 8% 7% 8% Did nothing 4% 7% 4% 4% 5% 4% 5% Selling food stock 3% 1% 3% 3% 3% 1% 3% Renting/pawning land 2% 1% 2% 2% 3% 2% 2% Selling durable goods 2% 2% 2% 1% 3% 2% 2% Selling agricultural assets 1% 0% 2% 1% 1% 1% 1% Reducing education/health expenses 1% 0% 1% 1% 0% 0% 1% Taking a loan 1% 2% 1% 1% 1% 0% 1% Practicing dry season agriculture 0% 0% 1% 0% 0% 0% 1% Help from government 0% 0% 0% 0% 0% 1% 0% Seeking supplementary income 0% 0% 0% 0% 0% 1% 0% Others 0% 0% 0% 0% 0% 0% 0% Increase in fishing 0% 0% 0% 0% 0% 0% 0% Selling assets 0% 0% 0% 0% 0% 0% 0% Help from NGOs 0% 0% 0% 0% 0% 0% 0% Engaging in spiritual activities 0% 0% 0% 0% 0% 0% 0% Children below 15 started working 0% 0% 0% 0% 0% 0% 0% Unemployed household members started working 0% 0% 0% 0% 0% 0% 0% World Bank Poverty and Equity Global Practice, Africa Region