Niger: Investing in Human Capital, Agricultural Productivity, and Social Protection for faster Poverty Reduction A POVERTY ASSESSMENT 2 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER 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 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 3 ACKNOWLEDGMENTS This report was produced by the World Bank Poverty and Equity Global Practice, Africa 2 Region, with collaboration from the National Institute of Statistics of Niger. The Task Team Leaders were Nga Thi Viet Nguyen (Senior Economist EECPV) and Clarence Tsimpo Nkengne (Senior Economist EAWPV). The team was comprised of Prospere Backiny-Yetna (Senior Economist EAWPV), Aly Sanoh (Senior Economist EAWPV), Aboudrahyme Savadogo (Economist EAWPV), Tomomi Tanaka (Consultant EAWPV), and Tao Tao (Consultant EAWPV). Emmanuel Skoufias (Lead Economist EAEPV), Arthur Alik Lagrange (Senior Economist EPVGE), and Franck Adoho (Senior Economist EAWPV) were peer reviewers for the report. Helpful comments and inputs were received from Stephanie Brunelin (Senior Economist HAWS2), Snjezana Plevko (Senior Economist HAWS2), Mahamane Maliki Amadou (Senior Social Protection Specialist HAWS2), Felix Lung (Social Protection Specialist HAWS2), and Mahaman Achirou Yahaya Arde (E T Consultant HAWS2). The team gratefully acknowledges guidance from Johan Mistiaen (Practice Manager, EAWPV), Andrew Dabalen (Practice Manager ESAPV and former Practice Manager, EWWPV), Soukeyna Kane (Country Director, AWCW3), Kofi Nouve (Manager, operations, AWCW3), Abebe Adugna (Regional Director, EAWDR), Joelle Beatrice Dehasse (Country Manager, AWMNE), Jean- Pierre Christophe Chauffour (Program Leader, EAWDR), and Christophe Rockmore (Program Leader EAFH2). Santosh Kumar Sahoo (Program Assistant, EAWPV), Zeinabou Bizo Hassane (Program Assistant, AWMNE), Senait Yifru (Operations Analyst, EAWPV), and Arlette Sourou (Senior Program Assistant, EAWPV) provided assistance and various supports in the preparation of the report. Finally, the team would like to thank Mr Idrissa Alichina Kourgueni, the General Director of the Institut National de la Statistique (INS), Zakari Oumarou, the Director of Households survey of INS, and everyone at INS for the important support and critical feedback provided throughout the implementation of the survey and the preparation of the report. 4 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION TABLE OF CONTENTS Acknowledgments 3 Abbreviations and Acronyms 10 Executive Summary 11 Introduction 17 1. Niger’s progress in reducing poverty 18 1.1. Recent development in poverty, food security, and inequality 19 1.1.1. Poverty continues to be a rural phenomenon 19 1.1.2. There are striking disparities in regional poverty 21 1.1.3. Social assistance does not seem to reach regions in need 23 1.1.4. Diversification of Nigerien food basket is low 24 1.2. Trend in poverty 26 1.2.1. Poverty declined significantly in rural areas and increased in urban areas. 26 1.2.2. Consumption growth favors the poorest, and welfare gap between the poorest and richest regions has narrowed. 27 1.2.3. But Nigerien’s perception of economic growth has been negative 29 1.3. Non-monetary dimensions of welfare 30 1.3.1. The poor have accumulated durable assets and gained slightly more access to financial services 30 1.3.2. But progress on access to basic services and achievement in human development remain limited among the poor 32 1.4. Drivers of poverty reduction 36 1.4.1. Growth and redistribution explain recent progress in poverty reduction; however, redistribution favors rural households but not urban residents 36 1.4.2. Higher returns to changes in households’ location, returns to changes in incidence of shocks, returns to access to services, and returns to asset ownership accounts for a vast majority of consumption growth. 37 1.4.3. For the poorest, returns to households’ location, education and labor of household members has increased significantly, but returns to household composition and has declined sharply. 38 1.5. The covid-19 pandemic is likely to wipeout part of the achievement in recent years 39 1.6. Conclusion 40 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 5 TABLE OF CONTENTS 2. Rural Income growth 41 2.1. Rural income, assets, market 42 2.1.1. Farming and livestock activities generate an important source of income, but opportunities for non-farm are limited 42 2.1.2. Human capital asset and productivity are low 46 2.1.3. Access to markets and financial services is limited 47 2.2. Opportunities for Nigerien rural income growth 49 2.2.1. There is a potential for increasing productivity of existing crops 49 2.2.2. Greater commercialization can grow incomes and incentivize investments 50 2.2.3. Commercialization goes hand in hand with diversification 53 2.2.4. Growing livestock income holds promise 53 2.2.5. Rural non-agricultural income growth will primarily come from mining and migration 55 2.3. Conclusion 56 3. Compound effects of shocks on welfare 57 3.1. Main Shocks to Welfare 59 3.1.1. Drought is the most frequently reported shock 59 3.1.2. Security environment is worsen 60 3.1.3. Health shock affects the poor population the most 62 3.1.4. Price is another important source of shocks 62 3.2. Household coping strategy 65 3.3. Effects of shocks on welfare 68 3.4. Conclusion 71 4. Human capital 72 4.1. Achievements in human capital index 73 4.2. Education 76 4.3. Health 85 4.4. Food security 89 4.5. Conclusion 92 References 93 Annex 1: Technical note on poverty measurement based on EHCVM 2018/19 data 96 Annex 2: Reconstruction of poverty estimates in 2014 to obtain comparable poverty trend between 2014 and 2018 106 Annex 3: Construction of Multi-Deprivation Index 108 Annex 4. Decomposition results 112 Annex 5. Impacts of shocks on households’ welfare 121 6 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION LIST OF TABLES AND BOXES Table 1. Poverty Measures for Niger in 2019 20 Table 2. Regional poverty headcount, poverty mass, and poverty density 21 Table 3. Poverty trend 2014-2019 26 Table 4. GDP annual growth 27 Table 5. Poverty declined signficantly across all regions, particularly Dosso, Zinder, and Maradi. 28 Table 6. Indicators of Educational Investment and Attainment 33 Table 7. Indicators of Health Investment and Attainment 35 Table 8. Productivity trend 49 Table 9. Projection of demand for different livestock-derived food types in Niger 54 Table 10. Type of shocks by region 58 Table 11. Human Capital Index and its components across benchmark countries 73 Table 12. Human Capital Index (HCI) and its components for the year 2018 and the year 2020 74 Table 13. Education investments and outcomes across benchmark countries 76 Table 14. Food security indicators across benchmark countries 90 Box 1. What is the official household survey EHCVM 2018/19? 19 Box 2. How is poverty measured in 2019? 20 Box 3. Safety nets in Niger 63 Box 4. Evidence on the impact of safety nets 70 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 7 LIST OF FIGURES Figure 1. 95 percent of the poor reside in rural areas 20 Figure 2. Poverty incidence is the same for male and female headed households 20 Figure 3. Poverty prevalence is highest in Dosso, Zinder, and Maradi regions 22 Figure 4. The poor concentrates in Dosso and Maradi. 22 Figure 5. Niamey, Dosso, and Maradi regions have the highest number of poor per square kilometer. 22 Figure 6. Half of Nigerien poor reside in Zinder and Maradi regions. 22 Figure 7. Nearly two thirds of the budget was spent on food in Zinder, Tillaberi, Maradi and Tahoua regions. 23 Figure 8. Food assistance does not seem to reach region in need 24 Figure 9. Poverty and cash transfer 24 Figure 10. Food diversity is lowest in Tillaberi, Maradi, Tahoua region 25 Figure 11. A quarter of the food budget comes from own production. 25 Figure 12. Nearly a quarter of food budget was spent on only two food items. 25 Figure 13. On average, about 40 percent of Nigerien households are net food buyers. 25 Figure 14. Consumption growth favors the poorest. 28 Figure 15. Fewer people believe the country’s economic situation and their own living conditions were good in 2018 than in 2014. 29 Figure 16. Fewer people indicated that they never experienced food hunger, lack of portable water, medical care, or cooking fuel. 29 Figure 17. The share of Nigerien people living in multi-deprivation has declined across all regions. 30 Figure 18. Asset ownership has increased, particularly mobile phones among the poor. 31 Figure 19. Access to financial services, especially mobile money, has improved. 31 Figure 20. Capacity to save money has increase among the poorest 31 Figure 21. Access to electricity increased, but not much for the bottom 40 percent… 32 Figure 22. And not much for the lagging regions. 32 Figure 23. Access to piped water is limited in the lagging regions but progress has been significant… 33 Figure 24. Many more poor people had access to piped water in 2018 than in 2014. 33 Figure 25. Niger net primary enrollment is below expected level when compared to other countries with similar incomes. 34 Figure 26. Similar story is observed for primary completion rate. 34 Figure 27. The country’s adolescents out of school rate is significantly higher than its peers. 34 Figure 28. Adult literacy rate is lower than the expected level of its income group. 34 Figure 29. Redistribution favors poor rural households but not poor urban households 36 Figure 30. Return to households’ endowment explains most of the reduction in poverty 37 8 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION LIST OF FIGURES Figure 31. Particularly, higher return to changes in households’ location, asset ownership, access to service, and incidence of shocks accounts for most of the progress in poverty reduction. 37 Figure 32. For the bottom 40 percent of the consumption distribution, returns to households’ endowments accounts almost entirely for consumption growth. 38 Figure 33. For the poorest, returns to households’ location, education, access to services, and labor of household members has increased significantly, but returns to household composition has declined sharply. 38 Figure 34. A major source of income for rural households comes from farming and livestock activities 42 Figure 35. …and also among the bottom 40 percent. 42 Figure 36. Nigeriens allocate 90 percent of land area on 3 major crops: millet, sorghum, and cowpea. 43 Figure 37. Crops grown vary by regions. 43 Figure 38. Over 90 percent of livestock income comes from sale of live animals. 44 Figure 39. The composition of animal sold varies by region. 44 Figure 40. Trade, mining, and industry are three major sectors contributing to rural income 44 Figure 41. The poorest tend to participate in non-farm enterprises in trade and services more than the richest. 44 Figure 42. Rural poor are less likely to receive domestic remittances… 45 Figure 43. But they have a higher probability to receive international remittances… 45 Figure 44. Domestic remittances contribute to a fifth of poor recipients’ household income … 45 Figure 45. …while this ratio is a third for international remittances 45 Figure 46. Land size is equitable distributed between the bottom 40 and upper 60 percent of the consumption distrubtion. 47 Figure 47. Agadez and Diffa regions have the highest number of cattle and small ruminants while poultry is raised in other regions. 47 Figure 48. Commercialization rate is low in Niger 48 Figure 49. Only a quarter of agricultural land is currently used for cereal production. 50 Figure 50. Land productivity varies little across regions. 50 Figure 51. Except for rice, very little cereal is imported. 51 Figure 52. Self-sufficiency ratio is higher in Niger than in its peers, except Mali. 51 Figure 53. Marginal food budget shares in rural and urban areas 52 Figure 54. Model projections of production and demand in Niger 54 Figure 55. Drought is most frequently reported 59 Figure 56. Total rainfall is poor in 2017 59 Figure 57. Poor rainfall in the north makes agriculture households vulnerable 60 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 9 LIST OF FIGURES Figure 58. Security incidents roar 61 Figure 59. Fatalities concentrate in Diffa and Tillaberi regions. 61 Figure 60. Conflicts in Diffa and Tillaberi regions consist mostly of violence against civilians. 61 Figure 61. Urban households are more likely to be affected by food prices 63 Figure 62. Most households rely on support from family and friends, sale of livestock, and savings to manage risks. 65 Figure 63. Savings is underused in Niger compared to other Sahelian countries. 65 Figure 64. Rural households sell livestock 66 Figure 65. Poor are more likely to get support from family or friends 66 Figure 66. Coping strategies vary by shock categories 67 Figure 67. Conflicts and droughts are significantly associated with lower household consumption 69 Figure 68. The magnitude of correlation between shock and food consumption is even larger. 69 Figure 69. Human capital and poverty 75 Figure 70. School enrollment rates by gender and age group 77 Figure 71. School enrollment rates by age group and wealth quintile 78 Figure 72. Primary school-age (7–12 years) enrollment rate by region 79 Figure 73. Distribution of reasons for having never attended a formal school (ages 7–24) 80 Figure 74. Hours of domestic work and work outside home in the past week, disaggregated by gender, age group, and school enrollment (mean values) 81 Figure 75. Problems reported at schools, disaggregated by school type (%) 82 Figure 76. Correlates of the likelihood of being enrolled in school (logit marginal effects with 95% confidence intervals) 83 Figure 77. Literacy rates by gender and age group 84 Figure 78. Maternal health across subpopulations 85 Figure 79. Child health across subpopulations 86 Figure 80. Experiences of illness and rates of visiting a health facility, by gender and age group 86 Figure 81. Reasons for not visiting a health facility when ill, by wealth quintile 87 Figure 82. Problems reported at public and private health facilities 88 Figure 83. Distribution of food security categories 89 Figure 84. Average household food security across regions (0-8 scale, 0 = insecure and 8 = secure) 90 Figure 85. Correlates of being food secure (logit marginal effects with 95% confidence intervals) 91 10 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION ABBREVIATIONS AND ACRONYMS ACAPS The Assessment Capacities Project IPC Integrated Phase Classification ACLED Armed Conflict Location and Event Data KCAL Kilocalorie AEZ Agro-Ecological Zone KM Kilometer AQUASTAT FAO’s Global Information System on Water LDF Livestock-Derived Food and Agriculture LSMS Living Standards Measurement Study ASNP Adaptive Safety Net Project MDGs Millennium Development Goals CCA Cellule de Crise Alimentaire MPI Multidimensional Poverty Index CFS Cellule Filets Sociaux MT Metric Ton CFW Cash for Work NDVI Normalized Difference Vegetation Index CHIRPS Climate Hazards Group InfraRed Precipitation NER Niger with Station data NFE Non-Farm Enterprise CPF Country Partnership Framework NSU Non-Standard Units CPI Consumer Price Index OB Oaxaca-Blinder DHS Demographic and Health Survey OECD Organization for Economic Cooperation DNPGCA Dispositif National de Prévention et de and Development Gestion des Crises Alimentaires OLS Ordinary Least Square EHCVM Enquête Harmonisée sur les Condition de Vie des Ménages OOP Out-Of-Pocket FAO United Nations Food and PEAMU Projet Eau et Assainissement en Milieu Urbain Agriculture Organization PER Public Expenditure Review FAOSTAT Food and Agriculture Organization Corporate PMT Proxy Mean Test Statistical Database PPP Purchasing Power Parity FCFA Franc de la Communauté Financière Africaine R&D Research and Development FEWSNET Famine Early Warning Systems Network SCD Systematic Country Diagnostic FIES Food Insecurity Experience Scale SDGs Sustainable Development Goals FINDEX Global Financial Inclusion Database SSA Sub-Saharan African FLFP Female Labor Force Participation SSAF Sub-Saharan French-Speaking Africa countries GDP Gross Domestic Product SU Standard Units GPS Global Positioning System SWEDD Sahel Women’s Empowerment and GYGA Global Yield Gap Atlas Demographic Dividend HA Hectare UNECA United Nations Economic Commission for Africa HCI Human Capital Index UNESCO The United Nations Educational, HEA Household Economy Approach Scientific and Cultural Organization HFPS High Frequency Phone Survey UNICEF United Nations Children’s Fund HI Herfindahl Index USR Unified Social Registry ICR Implementation Completion Report WAEMU West African Economic and Monetary Union ICT Information and Communications Technology WDI World Development Indicators IFPRI International Food Policy Research Institute WFP World Food Program INS Institut National de la Statistique NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 11 EXECUTIVE SUMMARY This poverty assessment aims to strengthen the across years. The agricultural sector which employs most analytical foundation for poverty-reduction policies and of the poor has seen not only substantial growth during interventions in Niger. With a Human Development Index this period, but also higher growth than the industry, of 0.394 in 2019, Niger is ranked at the bottom, 189 out of 189 manufacture and services sectors. Growth in the agriculture countries and territories. The country is experiencing rapid sector averaged 7 percent per year. Growth in agriculture population growth (estimated at about 3.8 percent per was mainly driven by good agro-climatic conditions and the annum) driven by a high fertility (6.9 children per woman), expansion of cultivated land. the highest in the World. Over the past decade, economic growth averaged almost 6 percent (but only 1.6 percent in Between 2014 and 2019, there was a substantial reduction per capita terms). The country earns its foreign exchange of poverty particularly in rural areas where most mainly from uranium and gold, which has limited domestic households earn their livelihood from agricultural sector. economic linkages. Distant second, livestock export also In the meantime, there was an increase in urban poverty. provide important revenue to the country. Overall, most of The strong macroeconomic performance translated into a the labor force is employed in a low productivity and shock substantial reduction of the proportion of the population prone rainfed agricultural sector. With poverty incidence at living in poverty between 2014 and 2019. The proportion above 40 percent, almost 9 million Nigerien continue to live of the population living below the national poverty line in poverty—of which almost half live in two regions: Zinder declined by 5.4 percentage points, to 40.8 percent in 2019. and Maradi. The COVID-19 pandemic growth slowdown is However, because of the high population growth, the estimated to have pushed up to an additional 270,000 people number of poor has increased from 8.6 million in 2014 into poverty. Niger is further beset by vast infrastructure to 9 million in 2019. Changes in the intensity and severity gaps, intensifying terrorism attacks domestically, persisting of poverty1 follow similar patterns to those observed for conflict in the Sahel region, and intensifying environmental the poverty headcount. Trend in poverty reduction was pressures. Robust poverty reduction and steady gains in different across residence areas. Those living in rural areas shared prosperity will require investments to bolster human have experienced a substantial improvement, with the capital, interventions to increase agricultural productivity proportion of rural population living below the national and develop the rural economy, and mitigate vulnerability poverty line falling rapidly, by 6.9 percentage points to 46.8 to shocks both at the household and community levels. The percent in 2019. However, during the same period, urban effectiveness of these interventions will hinge on improved poverty has increased by 3.8 percentage points, to 11.8 governance, successful conflict resolution, and percent in 2019. enhanced security. Poor households earn most of their income from POVERTY IN NIGER agriculture and livestock, and consequently poverty continues to be a rural phenomenon in Niger. It is From 2015 onward, Niger experienced accelerating and estimated that the country has 9 million poor inhabitants, agricultural sector driven economic growth. Following a among which over 95 percent reside in rural areas. Not only deep contraction in 2015, Niger’s average GDP growth was do rural inhabitants have higher risk of being poor, but solid at nearly 6 percent per year and was relatively even they also fall deeper into poverty compared to their peers 1 Also called poverty gap and squared poverty gap respectively. 12 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION in urban areas. The depth of poverty, which represents the 2019, the percentage of people living in multi-deprivation average distance between the consumption per capita of has dramatically dropped from 70 percent to 60 percent. the average poor and the poverty line is 13.0 percent in rural During the same period, ownership of modern assets such areas, against only 2.4 percent in urban areas. Income from as mobile phones and motorcycle has increased, especially agriculture and livestock represents more than half of the mobile phone coverage among the poor. Alongside the total income for rural households in the bottom 40 percent expansion of mobile phone coverage, access to financial of the consumption distribution. For poor households, services, particularly mobile banking, has improved. income from non-farm activities is smaller. However, financial savings and capacity to borrow have declined dramatically for all Nigeriens. the fact that poor There are important disparities at the sub-national level: households are less capable of saving and borrowing is Dosso, Zinder, and Maradi are lagging regions. At 6.9 a worrisome sign that they might be more vulnerable to percent, the poverty rate is by far the lowest in Niamey. On short-run economic shock. the opposite, Dosso, Zinder, Maradi and Tillaberi regions have the highest poverty rates, estimated at 48.4%, 47.8%, Gender inequality remains a serious development 46.1%, and 42.3% respectively. Out of all 8 regions, Zinder constraint in Niger, despite Government’ efforts to address and Maradi houses the largest poverty mass, or the number the issue. The probability of being enrolled in a school is of the poor – nearly half of Nigerien poor live in these higher for boys compared to girls with a bigger gap at the two regions alone. This compelling concentration of the tertiary related age group. The incidence of child marriage number of poor in Zinder and Maradi is crucial for budget continued to be among the highest in Sub Saharan Africa. allocation across regions, and cost estimates of a social Estimations from the EHCVM 2018/2019 data reveals that policy targeted the poor, such as social safety net programs. more than half of women got married at age of 18 or below. To target poverty, it is important to pay attention to two Child marriage contribute to maintain fertility rate at a high lagging regions: Zinder and Maradi. level. In addition, early marriage has negative consequence on girls’ ability to achieve high level of education, and also Consumption growth favors the poorest, resulting in a limit girls’ ability to participate to the labor market. In Niger, reduction of inequalities. Between 2014 and 2019, growth decision making within household is dominate by men. in real consumption per capita was positive and higher There is still huge gap in market production and income for the bottom 70 percent of the welfare distribution. But in agriculture, livestock, household enterprises, and wage this growth was negative for those at the top of the welfare employment between women and men. distribution. Noticeably, gains were particularly pro-poor in rural areas. However, in urban areas, growth was not pro- CONSTRAINTS ON POVERTY REDUCTION poor, those at the bottom of the distribution had a zero or negative growth, and therefore did not benefitted from At the macroeconomic level, there is a strong correlation the economic growth. All this resulted in a reduction of between economic growth and poverty reduction. Niger’s inequality, the Gini index decreasing from 36.9% in 2014 to growth elasticity of poverty is estimated at -1.8, indicating 35.0% in 2019. that a 1 percentage-point increase in GDP is associated with a 1.8 percentage-point decline in the poverty rate, this very Alongside the reduction in monetary poverty, the close to the average growth elasticity of poverty in Sub- proportion of Nigerien living in multi-deprivation has Saharan Africa (SSA). In addition, a pro-poor distribution declined. Designed to complement monetary measures, of returns to growth has contributed to a reduction of the multi-deprivation indicator takes into account access inequality. In 2014-19, the average consumption growth to health and education services, access to other basic was highest and positive among bottom 70 percent of services such as electricity, water, and sanitation, as well households of the consumption distribution, resulting in a as housing conditions and asset ownership. The level of reduction of inequalities. improvement across these non-monetary dimensions indicates more favorable conditions for households to avail Given the strong poverty elasticity of growth and a pro- themselves of economic opportunities and raise the quality poor redistribution of growth, both economic growth of their living condition. Measured against this indicator, and redistribution factors are keys drivers of poverty Niger has made significant progress. Between 2014 and reduction in Niger. Between 2014 and 2019, consumption NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 13 growth reduced the poverty rate by 2.9 percentage points, firms and consumers, and slowing global economic activity and a pro-poor redistribution of growth decreased the resulted in job losses, declining labor income, rising prices, poverty rate by 2.5 percentage points, and the interaction and diminished remittance inflows. As a result, it is estimated between growth and redistribution was negligeable, close that in 2020, 270,000 Nigerien felt into poverty because of to zero. Consequently, the national poverty rate declined by the COVID-19 pandemic. And by 2022, this number could 5.4 percentage points over the period. reach one million. Meanwhile, school closures are expected to further weaken educational outcomes among the current Redistribution and changes in returns to households’ generation of school-age children. The ongoing crisis is also endowments contributes to most of the consumption increasing the intensity of poverty for many of the country’s growth between 2014 and 2019. Between 2014 and 2019, poorest households, and these losses, combined with the poverty declined by 5.4 percentage points, among which 2.9 pandemic’s direct impact on public health, are likely to be percentage points come from growth and 2.5 percentage felt for decades to come. points from redistribution. In other words, redistribution and growth effects accounted each for about 50 percent The COVID-19 crisis has had an especially devastating of all poverty reduction in the past 5 years. However, impact on food security. Most household have experienced redistribution favors rural households but not urban a reduction of their income. As a consequence, a standard residents. Poverty declined by 6.9 percentage points copping mechanism was to reduce food consumption. In in rural areas. Economic growth contributes to only 2.7 high-frequency surveys, at the start of the pandemic, a very percentage points of this reduction, while redistribution high proportion of households were food insecure. Indeed, accounts for the rest. However, during this same period, about six out of ten households was eating almost always urban poverty increased by 3.8 percentage points. This the same thing, 57 percent of households worried about not increase is explained entirely by the negative redistribution. having food and 55 percent about not having food. Food Between 2014 and 2019, most of the impact of welfare insecurity is more pronounced in rural areas compared to came from returns. In particular, the largest share of the urban and has impacted poor households more improvement in consumption growth can be attributed to than non-poor. higher returns to labor of household members, returns to change in households’ location, and returns to changes Exposure to Other Type of Shocks in asset ownership and access to basic services. Due to high population growth, demographic factors contributed Nigerien households are highly vulnerable to a wide range of negatively to poverty reduction. shocks. The dependence of the rural population on rain-fed agriculture and pastoralism exposes a majority of the labor Several factors attenuate the relationship between force to weather-related shocks and the long-term impact growth and poverty reduction in Niger. These include: (i) of climate change. Much of the country’s manufacturing widespread exposure to multiple shocks, including the and services sectors also depend on agricultural output, COVID-19 pandemic, (ii) the low productivity of the rural intensifying macroeconomic vulnerability to droughts, economy, (iii) low overall levels of human capital, and (iv) a floods, pests, crop and livestock disease, and conflict- deep and persistent gender gap. induced disruptions in the rural economy. Over 40 percent of Nigerien households reported experiencing a shock, Exposure to Multiple Shocks either covariate or idiosyncratic shocks, in the past 3 years. The most important shocks Nigerien households face The COVID-19 Pandemic are natural hazards, health shocks, high food prices, and insecurity. Increased violence in bordering regions with Chad, Mali and Nigeria, and intercommunal conflict have The COVID-19 crisis has reversed much of the progress caused an unprecedented humanitarian crisis in the Niger in poverty reduction achieved in Niger in recent years. along with higher levels of food insecurity and malnutrition. Between 2014 and 2019, the national poverty rate fell by Given limited coverage of social protection, and limited 5.4 percentage points. However, because of the population penetration of insurance/financial services, to cope with growth, the number of poor increased during the same shocks, the most commonly used strategy for Nigerien period by about 400,000 people. In 2020, a combination of household is to rely on friends and family, sale of livestock, emergency health measures, precautionary behaviors by 14 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION and savings. multivariate regressions suggest that conflicts, do affect set of keys health related indicators such as child draught and related health shocks are negatively correlated and maternal mortality, nutrition, etc. As a result of limited to household food consumption in rural areas. access and quality of social services, Niger is still among the bottom in the Human Capital Index ranking. Niger ranks The Low Productivity of the Rural Economy 155th out of 157 countries on the World Bank’s Human Capital Index. There is lot of potential to increase rural income. The vast majority of the population resides in rural areas. A Deep and Persistent Gender Gap Consequently, the agriculture sector, including farming and livestock, continues to be the largest employer. Millet Gender inequality is a major cross-cutting challenge, and sorghum are the two main staple crops in Niger. despite Government’ efforts to address the issue. Despite Livestock plays a small role in income generation for rural progress in recent years, Niger continues to experience high households although it is relatively more important in the gender disparities in primary and secondary education. Niger northeast part of the country. The fact that the country has ranks second lowest on net enrollment rate for secondary one main harvest season, results in strong seasonality in school and percentage of secondary and primary school rural labor demand, income, and prices. Agricultural income students who are female. The incidence of child marriage are constraints by low productivity due to limited access continued to be among the highest in Sub Saharan Africa. to inputs, including financial services. Commercialization The median age at first marriage for women aged 20 to 49 is also very low. Most households failed to create surplus, is estimated at 15.8. Child marriage contribute to maintain and only produce for their own consumption. Opportunities fertility rate at a high level. In addition, early marriage has to increase rural income are many: increasing productivity negative consequence on girls’ ability to achieve high level of existing crops, greater commercialization, diversification of education, and also limit girls’ ability to participate to the into animal products and new crops. labor market. In Niger, decision making within household is dominate by men. Low Overall Levels of Human Capital Gender is a strong determinant of access to productive According to the HCI, a child born in Niger today can expect resources, both in terms of land size and livestock to achieve just 32 percent of her lifetime productive holdings. There is still huge gap in market production and potential, due to a combination of inadequate education income in agriculture, livestock, household enterprises, access, poor health outcomes, and high mortality rates. and wage employment between women and men. Female- The average child can expect to receive only five years headed households have smaller land and own less of schooling. Some indicators of education attainment livestock. Enterprises owned by women show 61 percent have improved, but others are still much lower than the lower profits than those owned by men. Gaps are also large average of the SSA region. The completion rates for primary for wage earners. The gender gap in earnings is estimated and lower secondary school in 2018 were 20 percent and at 29 percent when comparing similar male and female 19 percent respectively signaling that Niger will still need workers. These gender-based differences in access to significant effort and investment to catch up with the productive resources are reflected in the composition of regional average. Between 2014 and 2017, mortality rate for income, as the share of agricultural income in total income children under 5 declined significantly, from 99 children per for male-headed households is larger than that of female- 1,000 live births to 84 although it was still higher than the headed households. SSA average. Meanwhile, a decline in maternal mortality from 573 deaths per 100,000 births down to 509 signals an POLICY PRIORITIES improvement in maternal health between 2014 and 2018. However, a reduction in contraceptive prevalence poses a Going forward, for poverty reduction to happen in challenge to the government efforts to control the already- a sustainable way, the following actions should be high fertility rate. This reverse trend can adversely affect considered to overcome some of the many challenges. women’s health as well as their opportunity to participate This report has documented that Niger has experienced in the labor market. Niger has the highest fertility rates in a substantial poverty reduction in recent years. However, the world, this together which social norms and poverty, poverty reduction was concentrated in rural areas. Based NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 15 on the set of constraints to poverty reduction identified in Improving Productivity and Commercialization in the this report, and in order to end poverty and boost shared Rural Economy prosperity in a sustainable manners, policymakers and their partners may want of design and implement effective A pro-poor growth strategy must focus on expanding interventions to: (i) counter the impact of the COVID-19 crisis, livelihood opportunities in areas where the poor live and (ii) develop human capital, (iii) improve productivity and work. In Niger, poverty is overwhelmingly concentrated in commercialization in the rural economy, (iv) build resilience rural areas, and accelerating rural income growth will require to shocks by expanding coverage and improve targeting of enhancing the marginal productivity of the rural workforce. formal social safety nets programs, and (v) address gender Policymakers can boost agricultural productivity by inequality. Ensuring the success of these interventions will facilitating the use of fertilizer and other inputs, leveraging require steadily improving the quality of governance while the country’s water resources through expanded irrigation achieving lasting peace and security. systems, promoting improved soil-conservation practices, and introducing new techniques to support integrated crop Countering the Impact of the COVID-19 Crisis and livestock production. Commercialization is also limited as most farmers produce for their own consumption. The COVID-19 pandemic is exacerbating household- and Improvement of productivity, and a market-oriented community-level vulnerability, and its combined economic agriculture will help boost rural income given available and public-health shock poses unique challenges for domestic and international markets. This will also require a policymakers. The crisis is putting upward pressure on shift to high value crop, further investment in research and food and input prices, while illness, disrupted value chains, development, and policies to encourage farmers to adopt and the loss of wage employment are reducing income new technologies. Insecurity is one of the most binding levels. Cash and in-kind transfers to affected households constraint to agricultural productivity. To enable farmers could provide immediate relief, but Niger’s current safety- and entrepreneurs to invest in improving their long-term net programs are small and poorly targeted. In this productivity, the government must consolidate the rule of context, policymakers must leverage the ongoing efforts of law nationwide while building public institutions capable development partners to support the implementation of of protecting property rights and resolving disputes over the national social protection strategy, extend the coverage access to land and other resources. of social protection mechanisms, and create the necessary fiscal space to finance cash and in-kind transfers, while also Building Resilience to Shocks laying the foundation for the safe resumption of normal economic activity. Nigerien households typically have few resources to draw on in the event of a shock, and they often deploy Investing in Human Capital coping strategies that deplete their productive assets and human capital, diminishing their long-term productivity. To accelerate human capital development, policymakers Increasing household resilience will require developing must simultaneously increase the quantity and quality of adaptive social protection systems that can: (i) provide education and health services. The country needs to close income support and information to build and protect education and health gaps in order to improve human human capital in the face of shocks and crises; (ii) facilitate capital. A population that is well educated and in good access to critical education, health, and nutrition services; health will be more productive. Higher level of productivity (iii) promote equality of opportunity; and (iv) support the linked to higher income level and better living conditions. development of workforce skills to increase and diversify Effective learning requires prepared students, effective labor income. Creating new delivery platforms could enable teachers, adequate inputs, skilled management at the the authorities to implement a multisectoral approach to school level, and good governance across the educational social protection. system. Increasing access to preschool and expanding school feeding programs could yield especially large gains in child development at a relatively modest fiscal cost. 16 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Households across Niger face uninsured risks that prevent Addressing Gender Disparities them from investing more in more profitable agricultural activities. Offering agricultural index insurance or index- Women and girls play a crucial role in breaking cycles based livestock insurance could enhance resilience among of intergenerational poverty, and the success of the smallholder farmers and pastoralists. Agricultural index government’s poverty-reduction efforts will hinge on insurance links payouts to an average estimate of crop providing equitable access to education, health, and losses based on an index of factors, such as an area’s average productive opportunities. Policies designed to address rainfall or vegetation growth rate, and similar indexes gender disparities must reflect the unique constraints that underpin index-based livestock insurance. Establishing women face and the influence of social norms on women’s agricultural index insurance in the cash-crop-and-cereal voice and agency within their households and in the society livelihood zone and index-based livestock insurance in the as a whole. While gender parity is a cross-cutting issue that pastoralism-and-transhumance zone could prevent farmers should be mainstreamed into all poverty-reduction policies, and pastoralists from adopting costly coping mechanisms dedicated interventions should target gaps in human such as selling livestock and other assets, reducing food capital investment, customary practices related to land and intake, or withdrawing children from school in the event of asset ownership, early marriage and childbearing, decision- a shock. Moreover, mitigating exposure to crop-related risks making practices within households and communities, the would enable farmers to invest in higher-value cash crops division of labor for household tasks, and the critical issue such as cotton, rather than relying on low-value cereal of gender-based violence. grains such as millet and sorghum. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 17 INTRODUCTION This poverty assessment analyzes trends in monetary and 2018/19 EHCVM surveys. These two surveys are nationally nonmonetary aspects of poverty in Niger, mainly based representative cross sections and it is from this series that on two nationally representative household expenditure the official monetary poverty and other MDGs estimates are surveys conducted by the INS in 2014 and 2018/19. The study derived. But poverty estimates from the two EHCVM are not determines the drivers of poverty reduction by looking at directly comparable, adjustments were made in order to location, demography, labor, assets ownership dimensions. come-up with robust overtime poverty estimates. Data from The report also discusses cross-cutting issues relevant for the WDI are also used to assess the country’s performance poverty reduction, such as service delivery, human capital, compared to its peers. gender, and shocks. This study aims to provide policy makers with the knowledge needed to improve the effectiveness The various analysis conducted under this poverty of their programs to reduce and finally eradicate extreme assessment has proven to be critical in understanding poverty in Niger. poverty in Niger. It has served as a learning experience for World Bank staff as well as government officials, and This report examines Niger’s progress in reducing poverty builds the ground for future collaboration in planning and over five years, with a specific focus on the period 2014 to analysis of the upcoming 2021 EHCVM survey. 2019. The focus on this period is due to data availability. Niger’s progress in reducing poverty from 2014 and 2019 This poverty assessment consists of four chapters. The first is substantial. A reduction of 5.4 percentage points was chapter presents levels and main trends in monetary and mainly driven by rural areas. On the opposite, poverty has nonmonetary poverty. In addition, the chapter examines increased in urban areas. The observed reduction is in line the incidence of consumption growth and uses regressions with the macroeconomic performance observed during the and decompositions to examine the drivers of consumption same period where annual GDP growth average 5.6 percent. growth and poverty reduction, focusing in particular on four important factors that are derived from the literature: At seven births per woman, Niger has the highest fertility location, demographics, education, assets ownership, and rate in the world, resulting in high dependency ratio. labor. The second chapter explores the main income sources Moreover, the majority of the population is still living in for rural households. In particular, the chapter highlights rural areas, relying mainly on agriculture for livelihood. The the importance of agriculture for rural households. A special dispersion and low density of the population, especially attention is given to constraints faced by farmers, including in rural areas makes it very difficult to provide public issues related to access to land, access to agricultural inputs, service to the population. As a consequence, the country is access to productive assets, access to financial services performing poorly in terms of access to basic services such including insurance, low productivity, gender gaps, and as electricity and piped water. difficulties to access markets/commercialization. Chapter 3 relies on descriptive and inferential statistics to provide Unfortunately, data collection effort has been inconsistent information on incidence of shocks, households coping in Niger and therefore, the country has only two household mechanisms, and effect of shocks on welfare. Chapter 4 consumption surveys that are ready to be use in a provides an assessment of human capital endowments, robust and consistent way for this report: the 2014 and and how this relates to welfare. 18 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 1 NIGER’S PROGRESS IN REDUCING POVERTY NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 19 1.1. Recent development in poverty, food security, and inequality Poverty estimates used in this section are based on the recent official household survey, Harmonized Living Conditions Household Survey (EHCVM) 2018/19. This survey was conducted by the National Institute of Statistic and Demography of Niger (INSD), financed by the World Bank with support from the WAEMU Commission. Box 1 provides a brief description of the EHCVM survey. The methodology of poverty measurement is discussed in Box 2 and technical details are available in Annex 1. Throughout this report, the term “poverty” refers to “national poverty” unless indicated otherwise. 1.1.1. POVERTY CONTINUES TO BE A Despite the great concentration of poverty in rural areas, particular attention must be given to urban poverty as RURAL PHENOMENON well. Close to 450,000 poor lives in urban areas, and most of them are concentrated in the Tahoua, Maradi and Dosso Based on the official household data, poverty in Niger regions. It is important to note that these regions include continues to be a rural phenomenon in 2019. Among 9 some of the biggest cities in Niger. A concentration of poverty million poor inhabitants, over 95 percent reside in rural in urban areas can be source of social tension, therefore, areas (Figure 1). Close to half of rural population do not meet urban areas shouldn’t be excluded when designing pro- their basic nutritional and non-food needs, compared to 12 poor programs. percent of urban population. Not only do rural inhabitants have higher risk of being poor, but they also fall deeper into Poverty incidence is the same for male and female headed poverty compared to their peers in urban areas (Table 1). households. Poverty rate among female headed households is 40.7 percent (Figure 2), that is the exact same estimate as for male headed households (40.8%). Box 1. What is the official household survey EHCVM 2018/19? The Niger EHCVM 2018/19 is part of the joint effort between the National Statistic Offices, the World Bank and the WAEMU Commission to conduct new nationally-representative household surveys in each of the 8 WAEMU member countries. For Niger, the advantages of the newly introduced survey, EHCVM 2018/19, are threefold: (i) it follows international standards for poverty measures, (ii) it is comparable to the same household surveys conducted in other WAEMU countries, and (iii) it will be comparable to the next EHCVM survey planned in 2021/22. The EHCVM 2018/19 is representative at national level, regional level, and urban/rural level, and consists of 20 modules covering household consumption, food security, health, education, access to services, etc. While the EHCVM 2018/19 is comparable to the EHCVM 2014 in terms of survey design and implementation, the methodology to measure poverty differs significantly. To address the critical question of how to estimate changes in poverty between 2014 and 2018/19, we reconstruct household consumption aggregate in 2014 using the 2018 methodology to obtain the comparability between two years. 20 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Table 1. Poverty Measures for Niger in 2019 National poverty (%) Headcount Depth Severity National 40.8 11.2 4.3 Distribution of rural poor 11.8 2.4 0.7 Rural 46.8 13.0 5.0 95+5+A Figure 1. 95 percent of the poor reside in rural areas Figure 2. Poverty incidence is the same for male and Urban female headed households 40.0 5% 30.0 20.0 40.8 40.7 40.8 10.0 95% 0.0 Male Female Rural Gender of the household head National Source : World Bank staff calculation based on EHCVM 2018/19 Source: World Bank staff calculation based on EHCVM 2018/19 Box 2. How is poverty measured in 2019? Poverty measures are based on two building blocks: consumption aggregates and poverty lines. The consumption aggregate represents annual household consumption, and is calculated by aggregating food consumption, non-food consumption in non-durable goods and service, the use value of durable goods, and the imputed rent of owner-occupied and rent-free households. Food consumption comes from different sources: purchases in the market, households’ own production (cereals, vegetable, meat, etc.), food received from private or public transfers, and food away from home. The second component, non-food consumption covers households and personal expenses, as well as expenses on health and education. To measure the annual usage value of each durable asset, purchasing value, replacement value, depreciation rate and interest rate are taken into account. Finally, rent value is applied based on imputation from location and lodging conditions such as construction materials and access to services. The poverty line is the value of the welfare indicator that allows individuals to satisfy their minimum vital needs. Our approach follows the cost of basic needs proposed in Ravallion (1998) and is constructed in two steps. First, the food poverty line is estimated from a basket of food items providing each individual with 2,300 kilocalories, which is within the range of the internationally accepted standard. Second, a share of non-food expenditure is added to the food poverty line to constitute the national poverty line. See Annex 1 for further details. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 21 1.1.2. THERE ARE STRIKING DISPARITIES IN poor people per square kilometer respectively (Figure 5). Niamey has the highest poverty density because of its tiny REGIONAL POVERTY land area although it houses few poor people as mentioned above. This measure of poverty is particularly important for There are striking disparities in poverty at the sub-national policy makers on service delivery programs. In this context, level. As of 2019, a Nigerien inhabitant in Niamey has only services may come in many forms and include social services a 7-percent chance of being poor while this probability for such as primary education for all, economic services such a person living in Dosso region is 48 percent. Being the as irrigation systems for poor farmers, or information richest region in the country, Niamey houses very few poor services such as mobile phone coverage. The spatial people. The region has the lowest national poverty rate in distribution of poverty density and maps of current public the country, at only 7 percent. Unlike neighboring countries services coverage are critical for policy makers to decide such as Cote d’Ivoire and Togo where the capitals have the whether a new service delivery program can be offered, or biggest concentration of the poor despite relatively low an existing program can be expanded in a cost-effective poverty rates, Niamey accommodates less than 1 percent way. If so, how many locations can the programs reach, of the total poor population (Figure 3 and Figure 4). Such and where are these locations to be found? The coverage within-country disparities can present a potential source of of such programs depends heavily on the projected costs increasing tensions between poorer and richer areas, and (e.g., upfront investment such as schools, piping for water potentially affect the country’s future growth and connections, electricity lines and poles, etc.), which in turn security situation. are largely determined by the density of users and the current status of public service coverage. Dosso, Zinder, and Maradi regions are by far the most lagging regions. They have the highest poverty rate: more To target poverty, it is important to pay attention to two than 45 percent of their population are below the national lagging regions: Zinder and Maradi. As mentioned above, poverty line. Out of all 8 regions, Zinder and Maradi houses these two regions alone accommodate half of the poor the largest poverty mass, or the number of the poor – population (Figure 6). This compelling concentration of the nearly half of Nigerien poor live in these two regions alone number of poor in Zinder and Maradi is crucial for budget (Figure 6). Trailing behind Niamey, Dosso and Maradi have allocation across regions, and cost estimates of a social the highest poverty density in the country, at 37 and 49 policy targeted the poor, such as social safety net programs. Table 2. Regional poverty headcount, poverty mass, and poverty density Poverty headcount (%) Poverty mass (number of poor) Poverty density (num of poor/km2) Agadez 14.6 84,447 0.1 Diffa 34.0 246,549 1.6 Dosso 48.4 1,263,413 13.2 Maradi 46.1 2,063,754 14.4 Tahoua 35.7 1,492,671 14.9 Tillaberi 42.3 1,451,170 37.3 Zinder 47.8 2,249,417 49.4 Niamey 6.9 81,414 202.5 Source: World Bank staff calculation based on EHCVM 2018/19. 22 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Figure 3. Poverty prevalence is highest in Dosso, Zinder, Figure 4. The poor concentrates in Dosso and Maradi. and Maradi regions Figure 5. Niamey, Dosso, and Maradi regions have the Figure 6. Half of Nigerien poor reside in Zinder and 1+1+31416172325A highest number of poor per square kilometer. Maradi regions. Niamey Agadez Diffa Niamey Zinder 1% 3% Dosso Maradi 1% 25% 14% Dosso Tillaberi Tillaberi 16% Zinder 23% Tahoua 17% Diffa Maradi Tahoua Agadez 0 50 100 150 200 250 Source: World Bank staff calculation based on EHCVM 2018/19 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 23 1.1.3. SOCIAL ASSISTANCE DOES NOT SEEM TO REACH distribution. However, only one out of the four regions with the highest food insecurity prevalence benefit significantly REGIONS IN NEED from these programs (Figure 8). Dosso region has nearly 15 percent of its population living below the food poverty Moreover, Nigerien people in Maradi, Zinder, Dosso, line and food support program reached 26 percent of the and Tillaberi regions are the most vulnerable to food population. Nevertheless, the coverage of these program is insecurity.2 These regions are also the poorest in the relatively low in the other top three food insecurity regions, country where households tend to devote the largest namely Maradi, Zinder, and Tillaberi. Only 8, 10, and 14 proportion of their budget on food (Figure 7), thus are percent of the population in these regions, respectively, more vulnerable to food insecurity. Nearly one in every five benefits from food support. It is important to note that inhabitants in Maradi and Zinder regions does not have the the region with the largest food assistance coverage has a minimum 2,300 calories intake per day. Meanwhile, about relatively low level of food insecurity. In Diffa region, only 7 15 percent of the population in Dosso and Tillaberi regions percent of its inhabitant are food-insecure, yet 27 percent of face food shortage. the population receive some types of food support. Yet, food assistance does not seem to reach regions Coverage of cash transfer programs is limited across in need. Food assistance is the biggest program in the country. On average, less than 1 percent of households social protection system in terms of population coverage. benefit from either cash transfers or cash for work. Again, Nationwide, about 15 percent of Nigerien households Diffa region has the highest coverage of cash transfer, at receive food support, either through food for work, school nearly 5 percent of its population benefiting from cash feeding, or cereal distribution. In Niger, more than 80 support. (Figure 9). percent of food assistance beneficiaries receive cereal Figure 7. Nearly two thirds of the budget was spent on food in Zinder, Tillaberi, Maradi and Tahoua regions. Niamey Agadez Dosso Diffa Tahoua Maradi Tillaberi Zinder Rural Urban National 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Food Health Education Rent Housing Transportation Durable Assets Personal/household items Other Source: World Bank staff calculation based on EHCVM 2018/19. 2 Defined as the share of Nigerien population below the food poverty line. 24 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Figure 8. Food assistance does not seem to reach region Figure 9. Poverty and cash transfer in need 0.3 0.25 0.2 0.15 0.1 0.05 0 Niamey Agadez Diffa Tahoua Tillaberi Dosso Zinder Maradi Food insecurity rate Food support Source: World Bank staff calculation based on EHCVM 2018/19. 1.1.4. DIVERSIFICATION OF NIGERIEN FOOD BASKET Across most regions, up to a quarter of households’ food budget is spent on two food items only, rice and millet IS LOW (Figure 12). An average household in Zinder, Dosso, Maradi allocates about 20 percent of their food budget on millet Households in lagging regions, Tillaberi, Maradi, and alone. With a diet highly depending on a few food items, Tahoua, face another challenge: low diversification of households are more vulnerable to the volatility of food food intake. It is arguable that limited availability of food prices and production which can lead to food insecurity. options will affect individuals’ diet and nutrition that is necessary to maintain a healthy lifestyle and physical and Despite a high share of home production in households’ cognitive development, particularly for children. We use the food basket, about 40 percent of Nigerien households are Herfindahl Index (HI), also known as the Hirschman Index net food buyers (Figure 13), defined as households that or Hirschman-Herfindahl Index as an inverse measure buy more food on the market than they sell in 2018/19 of variety in food consumption.3 The HI ranges from 1/n (WFP, 2009). In the case of Niger, household’s sale revenue to 1, and reaches a maximum value of 1 if the share of includes both agricultural and livestock production. In consumption is entirely concentrated on a single food item. other words, a Nigerien household is a net food buyer when In other words, the HI measures diversity, where the higher their sale revenue from agriculture and livestock is lower the value of the index, the lower the diversity (Lee and than their food purchase from the market. This implies that Brown, 1989). Figure 10 suggests that Tillaberi, Maradi, and many Nigeriens are vulnerable to the fluctuation in food Tahoua population have a less diverse food basket than prices. It is expected that almost all inhabitants of Niamey their fellow citizens living in the rest of the country. In these are net food buyers (95 percent). However, approximately regions, about a quarter of household food consumption half of residents in Agadez and Tahoua regions have less come from their own home production (Figure 11). This can food spending than revenue from food production (e.g. partially explain the low diversification of food intake as selling of agricultural and livestock products). home production offers a less variation of food availability than the markets. 3 The HI is calculated as the sum of squared food shares: where the HI of household h is the sum of the budget shares s of each indi- vidual food item i consumed in household h. The HI ranges from 1/n to 1. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 25 Figure 10. Food diversity is lowest in Tillaberi, Maradi, Figure 11. A quarter of the food budget comes from Tahoua region own production. Figure 12. Nearly a quarter of food budget was spent on Figure 13. On average, about 40 percent of Nigerien only two food items. households are net food buyers. Source: World Bank staff calculation based on EHCVM 2018/19. 26 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 1.2. Trend in poverty Poverty trend in this section is based on the two official surveys EHCVM 2014 and EHCVM 2018/19. While the two surveys are comparable in terms of survey design and implementation, the methodology of constructing household consumption aggregates deviates significantly between the two surveys. The methodology in 2018 applies a harmonized approach among WAEMU countries which follows the international standards and take into account imputation of rent and usage of durable goods. As discussed above, Annex 1 describes the 2018 approach in details. The methodology in 2014 has major differences, including handling of outliers, spatial deflators, durables goods, imputed rent among others. To compare poverty between the two years, we apply the 2018 methodology to EHCVM 2014 to reconstruct household’s consumption aggregates in 2014. To maintain comparability, we also use the same poverty line as used in 2018 and adjust to 2014 prices based on the national CPI (Annex 2 provides additional results). 1.2.1. POVERTY DECLINED SIGNIFICANTLY IN RURAL shows that rural to urban migration is not yet substantial to affect poverty in a significant manner. Very few are AREAS AND INCREASED IN URBAN AREAS. migrating, and the few that migrate seems not to have the minimum endowment to be successful in accessing quality Over the past 5 years, poverty declined significantly, mostly jobs and better incomes urban areas. from the reduction in rural poverty. On the opposite, there was an increase in the proportion of those living under the The pace of poverty reduction in Niger is consistent with poverty line in urban areas. Between 2014 and 2018, national the pace of GDP growth in the past 5 years. Between 2014 poverty headcount declined by 5.4-percentage points to and 2018, Niger’s average GDP growth was solid at nearly 40.8 percent in 2018 (Table 3). A substantial improvement is 6 percent per year, and was relatively even across years. recorded in rural areas where the share of rural population The agricultural sector which employs most of the poor living under the national poverty line fell rapidly from 53.7 has seen not only substantial growth during this period, percent to 46.8 percent during the same period. However, but also higher growth than the industry, manufacture and urban poverty has increased from 8 percent to 11.8 percent. services sectors (Table 4). Despite a deep GDP contraction Urbanization is happening but at a very slow pace. Between in 2015, growth in the agriculture sector averaged 7 percent 2014 and 2019, the share living in urban areas increased by a per year. bit less that one percentage point. A sectoral decomposition Table 3. Poverty trend 2014-2018 2014 official 2014 reconstructed for comparability 2018 official Headcount Depth Severity Headcount Depth Severity Headcount Depth Severity National 44.1 13.1 5.4 46.2 14.7 6.3 40.8 11.2 4.3 Urban 8.6 1.4 0.4 8.0 1.4 0.4 11.8 2.4 0.7 Rural 51.0 15.4 6.3 53.7 17.2 7.5 46.8 13.0 5.0 Source: World Bank staff calculation based on EHCVM 2018/19 and EHCVM 2014. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 27 Table 4. GDP annual growth Annual % growth 2014 2015 2016 2017 2018 GDP 7.5 4.3 4.9 4.9 6.5 GDP per capita 3.4 0.4 1.0 0.9 2.5 Agriculture 8.3 1.9 11.1 5.7 7.5 Industry 0.8 -0.8 3.7 5.6 3.2 Manufacture 1.3 1.8 5.9 5.9 1.9 Services 9.5 7.3 2.7 4.4 6.9 Source: World Bank staff calculation based on World Development Indicators 2019. 1.2.2. CONSUMPTION GROWTH FAVORS THE POOREST, a negative growth, while the top 20 percent had a positive growth. This resulted in a reduction of inequality, the Gini AND WELFARE GAP BETWEEN THE POOREST AND index decreasing from 36.9% in 2014 to 35.0% in 2018. RICHEST REGIONS HAS NARROWED. Consumption gains were widely shared geographically Growth in consumption was pro- poor. As shown in growth as well, with substantial progress in the poorest regions, incidence curves (Figure 14), growth in real consumption Dosso, Maradi, and Zinders. Between 2014 and 2018, per capita between 2014 and 2018 was highest among poverty declined sharply by 17.6 percentage points, 16.6 households in the bottom of the consumption distribution percentage points, and 5.7 percentage points in these where the poorest decile experienced a 5-percent annual regions, respectively (Table 5). Although they remain the growth in consumption. Gains were particularly high in rural poorest regions in Niger, the welfare gap to the richest areas where all households in the bottom 70 percent of the region has shrunk. In fact, median consumption in Dosso, consumption distribution had positive growth. However, in Maradi, and Zinders increased significantly by 24 percent, 21 rural areas, those in the top 30 percent of the consumption percent, and 7 percent respectively compared to a national distribution have experienced a decrease of their average of 5 percent growth. consumption. In urban areas, the bottom 80 percent had 28 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Figure 14. Consumption growth favors the poorest. Table 5. Poverty declined significantly across all regions, particularly Dosso, Zinder, and Maradi. Region 2014 official 2014 reconstructed for comparability 2018 official Change Agadez 9.9 6.2 14.6 8.4 Diffa 32.9 38.8 34.0 -4.8 Dosso 57.4 66.0 48.4 -17.6 Maradi 65.0 62.7 46.1 -16.6 Tahoua 27.6 28.7 35.7 7.0 Tillabéri 40.0 46.4 42.3 -4.1 Zinder 52.7 53.5 47.8 -5.7 Niamey 5.2 7.5 6.9 -0.6 National 44.1 46.2 40.8 -5.4 Source: World Bank staff calculation based on World Development Indicators 2019. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 29 1.2.3. BUT NIGERIEN’S PERCEPTION OF ECONOMIC significantly, from 70 percent to 25 percent. Similarly, fewer people believe their own household living conditions were GROWTH HAS BEEN NEGATIVE in good position in 2018 than in 2014 (Figure 15). During the same period, the proportion of Nigerien population Interestingly, Nigerien people’s perception of progress indicating that they never experienced food hunger, lack in economic condition over the past 5 years is negative, of portable water, medical care, or cooking fuel dropped and contradicts the decline in poverty trend. Between significantly (Figure 16). 2014 and 2018, the share of population describing the country’s economic situation as good or very good declined Figure 15. Fewer people believe the country’s economic Figure 16. Fewer people indicated that they never situation and their own living conditions were good in experienced food hunger, lack of portable water, medical 2018 than in 2014. care, or cooking fuel. 70.4 57.4 41.95 49.38 42.21 65.9 50.38 24.95 26.17 26.61 47.35 29.19 Country’s Economic Condition Household’s Living Condition Never without Never without Never without Never without food water medical care cooking fuel 2015 2018 2015 2018 Source: World Bank staff calculation based on Afrobarometer 2015 and 2018 30 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 1.3. Non-monetary dimensions of welfare 1.3.1. THE POOR HAVE ACCUMULATED DURABLE (see Annex 3 for details). The level of improvement across these non-monetary dimensions indicates more favorable ASSETS AND GAINED SLIGHTLY MORE ACCESS TO conditions for households to avail themselves of economic FINANCIAL SERVICES opportunities and raise the quality of their living condition. Measured against this indicator, Niger has made significant The share of Nigerien people living in multi-deprivation progress. Between 2014 and 2018, the percentage of people has declined. Going beyond the sole focus on monetary living in multi-deprivation has dramatically dropped from measure of poverty, the multi-deprivation indicator takes 70 percent to 60 percent. The level of progress is observed into account access to health and education services, across all regions (Figure 17), with remarkable improvement access to other basic services such as electricity, water, and recorded in Agadez and Dosso regions. sanitation, as well as housing conditions and asset ownership Figure 17. The share of Nigerien people living in multi-deprivation has declined across all regions. 0.800 0.700 0.600 0.500 0.400 0.300 0.200 0.100 0.000 Agadez Diffa Dosso Maradi Tahoua Tillaberi Zinder Niamey 2014 2018 Source: World Bank staff calculation from EHCVM 2014 and EHCVM 2018/19. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 31 At the same time, the poor accumulated some durable distribution (Figure 19). However, financial savings and and financial assets. Ownership of modern assets such as capacity to borrow have declined dramatically for all mobile phones and motorcycle has increased, especially Nigeriens, particularly for the poorest (Figure 20). Between mobile phone coverage among the poor (Figure 18). For 2014 and 2017, the share of Nigerien people in the lowest the bottom 40 percent of the consumption distribution, quintile of the consumption distribution with financial the rate of mobile money coverage was merely 3 percent in savings plummets from 55 to 14 percent. During the same 2014. This number increased slightly in 2017 with significant period, the proportion of the poorest who were able to improvement among households in the second quintile borrow also declined from 59 to 44 percent. While the slight of the income distribution (Figure 19). In parallel to the improvement in financial tools like mobile banking signals expansion of mobile phone coverage, access to financial a promising future, the fact that poor households are less services, particularly mobile banking, has improved albeit capable of saving and borrowing is a worrisome sign that still with slow progress. Ownership of a bank account they might be more vulnerable to short-run increased substantially among the rich, but the progress economic shock. is much slower for those at the bottom of the income Figure 18. Asset ownership has increased, particularly Figure 19. Access to financial services, especially mobile mobile phones among the poor. money, has improved. 15 20 15 10 % change 10 5 5 0 0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Consumption quintile Bank account Mobile money account Mobile phone Motor cycle 2014 2017 Source: World Bank staff calculation based on EHCVM 2014 and EHCVM 2018/19 Source: World Bank staff calculation based on FINDEX 2014 and 2017. Figure 20. Capacity to save money has increase among the poorest 100 80 60 40 20 0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Saved money Borrowed money 2014 2017 Source: World Bank staff calculation based on FINDEX 2014 and 2017. 32 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 1.3.2. BUT PROGRESS ON ACCESS TO BASIC SERVICES Coverage of electricity is also uneven across regions, and progress has been slow. The level of access is AND ACHIEVEMENT IN HUMAN DEVELOPMENT REMAIN disproportionally high in Niamey where at least two thirds LIMITED AMONG THE POOR of its residents had connection to electricity in 2018. Meanwhile, in the poorest regions, namely Zinder, Dosso, For the poor, access to basic services is extremely limited and Maradi, only 10 percent of the households had access. and has seen little progress over time. Absolute levels of Improvement on electricity access has been stagnant in all access are still very low compared to the regional average, regions except in Tillaberi and Tahuoa regions (Figure 22). while the gap in progress between those at the top 60 percent and bottom 40 percent of the consumption distribution has Compared to electricity, the regional gap in access to widened slightly. For instance, only 16 Nigerien households piped water is narrower, and noticeable improvement is had access to electricity in 2018, compared to nearly one recorded. The level of access is still highest in the capital in every two SSA residents (World Development Indicators with more than 80 percent of its inhabitants had access to 2019). While this level of electricity coverage already reflects piped water in 2018. However, other regions are catching up, a modest improvement from 2014, the progress skews especially Diffa, Maradi and Dosso region (Figure 23). In 2018, toward the richer proportion of the population. Access to 36 percent of the residents in Diffa region had connection electricity remains extremely low with little improvement to piped water marking a sharp increase from 16 percent among the bottom 40 of the consumption distribution in 2014. Maradi region has also seen an impressive trend (Figure 21). Compared to electricity, coverage of piped with 47-percent and 32-percent coverage in 2018 and 2014 water is slightly better for the poor. In 2018, a quarter of respectively. Similarly, Dosso region increased its coverage households in the bottom 40 percent had connection to from 20 percent to 30 percent of the population during the piped water marking a significant progress since 2014. same period. However, this level is still much lower than the access level among the better-off population where nearly half of them had piped water connected to the house (Figure 24). Figure 21 : Access to electricity increased, but not much Figure 22 : And not much for the lagging regions. for the bottom 40 percent… 27 23 16 13 4 2 Top 60% Bottom 40% National 2014 2018 Source: World Bank staff calculation based on EHCVM 2014 and EHCVM 2018/19. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 33 Figure 23 : Access to piped water is limited in the lagging Figure 24 : Many more poor people had access to piped regions but progress has been significant… water in 2018 than in 2014. 48 43 24 36 31 18 Top 60% Bottom 40% National 2014 2018 Source: World Bank staff calculation based on EHCVM 2014 and EHCVM 2018/19. Although public resources allocated to education has recorded for secondary school enrollment. However, despite declined over the past years, encouraging trend is still a relatively high school enrolment rate, the completion rate observed. Between 2014 and 2018, while government is dismay reflecting the low quality of education, social expenditure on education has declined from 7 percent to 5 norms, perceived returns to education, and opportunity percent, it is still slightly higher than the regional average cost. Albeit some small progress, the completion rates (Table 6). Some indicators of education attainment have for primary and lower secondary school in 2018 were 20 improved, but others are still much lower than the average percent and 19 percent respectively signaling that Niger will of the SSA region. As of 2017, 65 percent of school-age still need significant effort and investment to catch up with children were enrolled in primary school, a 5-percentage the regional average. point increase from the figure in 2014. Progress is also Table 6 : Indicators of Educational Investment and Attainment Niger SSA 2014 2018 2014 2018 Government expenditure on education, total (% of GDP) 6.7 4.9 4.3 4.6 Government expenditure on education, total (% of government budget) 21.7 16.8 17.3 17.8 School enrollment, primary (% net) 60.1 65.1* N/A N/A Primary completion rate (% relevant age group) 57.6 71.6** 68.1 68.8 School enrollment, secondary (% net) 15.2 20.1* 34 35.6 Lower secondary completion rate (% relevant age group) 12.2 19 42.4 44.1 * data available in 2017 **data available in 2016 Source: World Development Indicators 2019 34 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION As of now, Niger is still among the bottom in the Human translated to higher education. When focusing on the out- Capital Index ranking. According to cross-country regression of-school rate (that is, the inverse of net enrolment), nearly analysis, Niger net primary enrollment and completion rates 80 percent of lower-secondary school-age children were are below expected level when compared to other countries not in school in 2018 (WDI 2019), a rate that is significantly with similar incomes (Figure 25 and Figure 26). In addition, higher than that of its income group (Figure 27). Similarly, enrollment in secondary school remains very low while net Niger’s literacy rate remains extremely low compared to enrollment in primary schools is relatively high. Only 20 other countries with the same income level although the percent of school age children were enrolled in secondary share of literate adults has increased from 29 percent to 33 school in 2017. This means the increase in primary school percent over the past 5 years (Figure 28). enrollment and completion rate has not successfully Figure 25. Niger net primary enrollment is below Figure 26. Similar story is observed for primary expected level when compared to other countries with completion rate. similar incomes. Figure 27. The country’s adolescents out of school rate is Figure 28. Adult literacy rate is lower than the expected significantly higher than its peers. level of its income group. Source : World Bank staff calculation from World Development Indicator 2019. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 35 As health expenditure increased over the past years, Niger 573 deaths per 100,000 births down to 509 signals an made impressive progress in health care for children and improvement in maternal health between 2014 and 2018. women. Between 2014 and 2017, a larger share of GDP was However, a reduction in contraceptive prevalence poses a allocated to health expenditure that brought the country’s challenge to the government efforts to control the already- level of health expenditure above the regional average high fertility rate. This reverse trend can adversely affect (Table 7). During the same period, mortality rate for children women’s health as well as their opportunity to participate under 5 declined significantly, from 99 children per 1,000 in the labor market. live births to 84 although it was still higher than the SSA average. Meanwhile, a decline in maternal mortality from Table 7. Indicators of Health Investment and Attainment Niger SSA 2014 2018 2014 2018 Current health expenditure (% of GDP) 5.9 7.7* 4.9 5.2 Mortality rate, under 5 (per 1,000 live births) 98.5 83.7 87.5 68.8 Maternal mortality ratio (per 100,000 births) 573 509 571 534.0* Contraceptive prevalence, modern methods (% of women ages 15-49) 14.4** 10.5* N/A N/A * data available in 2017 **data available in 2016 Source: World Development Indicators 2019 36 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 1.4. Drivers of poverty reduction This section is based on the reconstructed consumption aggregates for EHCVM 2014 to ensure comparability of poverty estimates between 2014 and 2018 (see section 1.2 above or Annex 2 for more information). To understand key factors behind the observed poverty trend in Niger, a number of decomposition methods, namely growth-redistribution decomposition, non-linear Oaxaca Blinder decomposition, and unconditional quantile regressions were applied in this section (see Annex 4 for detailed results). 1.4.1. GROWTH AND REDISTRIBUTION EXPLAIN RECENT Redistribution favors poor rural households but not poor urban households. Niger has seen a substantial decline PROGRESS IN POVERTY REDUCTION; HOWEVER, in rural poverty in the past five year. Poverty declined by REDISTRIBUTION FAVORS RURAL HOUSEHOLDS BUT 6.9 percentage points in rural areas. Economic growth NOT URBAN RESIDENTS contributes to only 2.7 percentage points of this reduction, while redistribution accounts for the rest (Figure 29). Overall, progress in poverty reduction in Niger comes from However, during this same period, urban poverty increased both growth and redistribution. Between 2014 and 2018, by 3.8 percentage points. This increase is explained entirely poverty declined by 5.4 percentage points, among which 2.9 by the negative redistribution. In other words, urban poor percentage points come from growth and 2.5 percentage were “taxed” more than benefits they should have received points from redistribution (Figure 29). In other words, from social assistance. In a context were coverage of social redistribution and growth effects accounted each for about protection programs is limited, the importance of the 50 percent of all poverty reduction in the past 5 years. redistributive effect here could mainly be related to how the market adjusted/reacted given strong performance of the agricultural sector (mainly benefitting the poor) vs other sectors (dominated by non-poor). Figure 29. Redistribution favors poor rural households but not poor urban households National -2.5 -2.9 Urban -1.0 4.9 Rural -4.2 -2.7 -0.8 -0.6 -0.4 -0.2 0.0 2.0 4.0 6.0 Growth Redistribution Residual Source: World Bank staff calculation based on EHCVM 2018/19 and EHCVM 2014 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 37 1.4.2. HIGHER RETURNS TO CHANGES IN HOUSEHOLDS’ More specifically, the largest share of the improvement in consumption growth can be attributed to higher LOCATION, RETURNS TO CHANGES IN INCIDENCE OF returns to returns to change in households’ location, SHOCKS, RETURNS TO ACCESS TO SERVICES, AND returns to changes in asset ownership, returns to RETURNS TO ASSET OWNERSHIP ACCOUNTS FOR A changes in incidence of shocks, and returns to access VAST MAJORITY OF CONSUMPTION GROWTH. to basic services (Figure 31). Consistent with the role of redistribution in rural poverty in section 1.4.1. above, returns to households residing in rural areas have increased Overall, return to households’ endowments accounts for significantly, and accounts for 343.1 percent of consumption most of the poverty reduction. Between 2014 and 2018, growth (see Table A4.3 in Annex 4 for regression results). overall households’ endowment in terms of where they In addition, returns to ownership of a house has risen and live, education and employment of household members, makes for 37.7 percent of consumption growth. Meanwhile, household composition, asset ownership, and their higher returns to having access to clean drinking water and experiences with shocks seems to have deteriorated a electricity is responsible for 80.3 percent and 15.3 percent of little bit, contributing negatively to consumption growth in Nigerien growth, respectively. The negative effect of change Niger (Figure 30). Changes in households’ endowment only to returns to education was driven by the number of family explains about -3 percent decrease of the overall changes members with no education, which accounts for -101.9% in consumption over the past 5 years. Meanwhile, returns to percent decrease of consumption. Driven by the high level those endowment has increased substantially and accounts of fertility, returns to demographic factors accounted for a for 109 percent of consumption growth. high reduction of consumption. Figure 30. Return to households’ endowment explains most Figure 31. Particularly, higher return to changes in of the reduction in poverty households’ location, asset ownership, access to service, and incidence of shocks accounts for most of the progress in poverty reduction. 120.0% 109.2% 300.0% 100.0% 200.0% 80.0% 100.0% 60.0% 0.0% 40.0% -100.0% 20.0% -200.0% 0.0% -3.0% -300.0% -20.0% -6.2% -400.0% Endowments Return to Interaction -500.0% Endowments Location Demographics Education Asset ownership/ Housing Access to service Shock Labor Endowments Return to Endowments Interaction Source: World Bank staff calculation based on EHCVM 2018/19 and EHCVM 2014 38 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 1.4.3. FOR THE POOREST, RETURNS TO HOUSEHOLDS’ education and employment of household members, household composition, asset ownership, and their LOCATION, EDUCATION AND LABOR OF HOUSEHOLD experiences with shocks explain almost entirely the growth MEMBERS HAS INCREASED SIGNIFICANTLY, BUT in consumption of the poorest (Figure 32). Meanwhile, RETURNS TO HOUSEHOLD COMPOSITION AND HAS changes in households’ endowment play a much less DECLINED SHARPLY. important role. Understandably, changes in endowment and returns to Particularly for the poorest, returns to households’ endowment are not uniform across the consumption location, education, access to services, and labor of distribution.That means, the poor may not acquire household members has increased significantly, but education, access to services, employment, or physical returns to household composition have declined sharply mobility at the same rate as the non-poor. In addition, the over the past 5 years. The location where a household lives returns to these changes in endowment among the poor continues to be a key factor in explaining the households’ may not be as high as those among the rich. Thus, we use consumption growth. In fact, returns to being in rural areas the Unconditional Quantile Regression method to explore has improved substantially and accounts for 216.1 percent such differences across per capita consumption quintiles. of the overall growth in consumption among the bottom We use log of consumption per capita as the dependent 40 percent of the consumption distribution. Moreover, variable and focus on the bottom 40 percent of the higher returns to the number of family members being consumption distribution. employed also makes for 147.5 percent of the consumption growth. However, returns to age of household head and number of household members without an education has Similar to other income cohorts, households in the continued to decline and contributes negatively to changes bottom 40 percent of the consumption distribution have in household’s consumption. seen significantly increasing returns to endowment. In fact, higher returns to households endowment of location, Figure 32. For the bottom 40 percent of the consumption Figure 33. For the poorest, returns to households’ distribution, returns to households’ endowments accounts location, education, access to services, and labor of almost entirely for consumption growth. household members has increased significantly, but returns to household composition has declined sharply. 120.0% 102.1% 600.0% 100.0% 400.0% 80.0% 200.0% 60.0% 0.0% 40.0% 30.1% 20.0% -200.0% 0.0% -400.0% -20.0% -600.0% -32.2% -800.0% Location Demographics Education Asset ownership/ Housing Access to service Shock Labor Endowments Return to Interaction Endowments Endowments Return to Endowments Interaction Source: World Bank staff calculation based on EHCVM 2018/19 and EHCVM 2014 Note: Here, the unconditional quantile regression is carried out for the bottom 40%; and for the change in welfare between 2014 and 2019. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 39 1.5. The covid-19 pandemic is likely to wipeout part of the achievement in recent years The World Bank has partnered with the INS to conduct a The Covid-19 pandemic and the related economic high frequency household survey, tailored to monitor the downturn is having an adverse effect on progress made impact of the Covid-19 on living conditions. It is anticipated on poverty reduction. Micro-macro simulations suggest that the Covid-19 will have negative and likely long-lasting that the international poverty rate will increase by at least effects on the population, in particular, the poor and most 1.6 percentage points in 2020 due to the Covid-19, this vulnerable. In addition to the direct health effect, social correspond to an additional 375,000 new poor due to the distancing and other containment measures have resulted Covid-19. As discussed above, one policy response from the worldwide in a reduction of the economic activity. Data Government will be to expand the existing social protection collection for the first round was conducted in September/ programs to help the poor and most vulnerable. October 2020. The survey instruments were designed to capture a wide range of information on the socio-economic Given differences in how men and women are being impact of the Covid-19, including knowledge and practices affected by the Covid-19, the response to the pandemic of recommended social behaviors, access to basic services should not be gender blind. It is important to note that (health and education), access to food and food insecurity, men and women will not be affected the same way by the labor market and income sources including remittances, Covid-19. Existing literature4 suggest that there is gender and social protection programs. differentiated transmission channels and impacts on outcomes across the three areas of (i) endowments; (ii) Findings of the high frequency survey are not yet available, economic conditions; and (iii) and agency. On endowments, but based on the literature, and results observed in it is demonstrated that due the pandemic, health resources similar countries in Africa and across the globe, it is may be shifted away from women’s health services, which anticipated that the pandemic will lead to a reduction could have critical impact on women in short and long of incomes/revenues, and that the poor will be the most term. In some places, due to reduction of income and social affected. Most employees, nonfarm businesses owners and norms could affect intra-household allocation of resources, farmers will likely experience a reduction of income due with a preference of investing in boy’s education. Due to to business closure and loss of incomes. By virtue of the school closure, drop out could be higher for girls because type of job, the poor and vulnerable are more likely to stop of prevalence of child marriage and teenage pregnancy. working during Covid-19, as they cannot work from home All these resulting in growing gender gaps. On economic because they are either in agriculture or in informal sector. conditions, women will experience more pressure on their As far as education is concerned, during Covid-19, non-poor time, for instance with school closure they will have to children are less likely to remain intellectually engaged, spend more time taking care of children, limiting the time while children from non-poor families, could afford to rely they could spend on labor market. In developing countries, on various channels, including ICT to continued learning. women are more likely to be engaged in informal sectors, To cope with income reduction, households may opt to with limited social protection, and limited possibility reduce food consumption, which will ultimately increase for home-based work. Regarding gender implication for the already challenging issue of malnutrition. agency, it is demonstrated that the Covid-19 could results in increased gender-based violence due to confinement. 4 For example, see De Paz et al (2020). 40 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 1.6. Conclusion Between 2014 and 2019, the proportion of the population of social assistance is low are poorly targeted. Monetary living below the national poverty line has declined in rural poverty is correlated with food insecurity, with those living areas where most households earn their livelihood from in poorest region struggling to meet their nutritional needs. agricultural sector. In the meantime, there was an increase Between 2014 and 2019, growth favors the poorest, resulting in urban poverty. This chapter has documented the poverty in a reduction of inequalities. trends, as well as main drivers of poverty between 2014 and 2019. Despite an improvement of survey methodology The high fertility in the country is slowing down progress in 2018/19, the EHCVM remains comparable to the previous in reducing poverty. Findings from the Gender Assessment survey. However, for robust comparison overtime, poverty Report (World Bank ,2020) point to a set of complementary estimates for 2014 were recomputed using the newly reforms that must be implemented in order to address the proposed methodology. Findings suggest that poverty high population growth, mainly empowering women and continues to be a rural phenomenon, with 95 percent of the young girls trough: enhancing women’s entrepreneurial 9 million poor living in rural areas. However, the chapters skills; promoting women’s employment and closing the highlighted the fact that poverty pockets also exist in urban earnings gap; increasing women’s productivity in agriculture; areas, and policy programs should take this into account. protecting adolescent girls through a legal framework; and There are huge spatial disparities, nearly half of Nigerien promoting universal education. poor live in these two regions: Zinder and Maradi. Coverage NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 41 2 RURAL INCOME GROWTH 42 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION With 95 out of 100 Nigerien poor living in rural areas, reducing poverty in Niger needs to address income growth in rural areas. This chapter aims to examine how those who currently live in rural areas can earn higher incomes in the future, and identify key opportunities for rural income growth. Rural households can boost their incomes by moving into non-agricultural sectors, through migration or by pursuing non-agricultural activities in rural areas. Such transitions need to be encouraged, but they will not have a large impact right away. In the near term, for the majority of rural Nigerien households, income growth will largely come from increases in agricultural income. 2.1. Rural income, assets, market 2.1.1. FARMING AND LIVESTOCK ACTIVITIES employer nationally and in rural areas. Farming and livestock activities are the main occupation for 75 percent GENERATE AN IMPORTANT SOURCE OF INCOME, BUT of Niger’s workforce. On average, about 40 percent of OPPORTUNITIES FOR NON-FARM ARE LIMITED household income is generated by agricultural activities (Figure 34). It contributes to more than half of the total The agriculture sector, including farming and livestock, income for rural households in the bottom 40 percent of continues to be the largest employer and dominates the consumption distribution (Figure 35). This is relatively Nigerien rural household income. Although agriculture’s lower than the sub-Saharan African average of 68 percent contribution to GDP has remained stagnant at about of rural income. 40 percent over the last 20 years, it is by far the largest Figure 34. A major source of income for rural households Figure 35… and also among the bottom 40 percent. comes from farming and livestock activities Share of income from different sources, by region Share of income from different sources, by consumption quintiles 1 1 .8 .8 .6 .6 .4 .4 .2 .2 0 0 Q1 Q2 Q3 Q4 Q5 Agadez Diffa Dosso Maradi Tahoua Tillaberi Zinder Salary Non-employment Salary Non-employment Livestock Farm Livestock Farm Enterprise Remittance Enterprise Remittance Source : World Bank staff calculation from EHCVM 2018/19. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 43 Figure 36. Nigeriens allocate 90 percent of land area on 3 Figure 37. Crops grown vary by regions. major crops: millet, sorghum, and cowpea. Share of cultivated area, by consumption quintiles Share of cultivated area, by region 1 1 .8 .8 .6 .6 .4 .4 .2 .2 0 0 Q1 Q2 Q3 Q4 Q5 Agadez Diffa Dosso Maradi Tahoua Tillaberi Zinder Millet Sorghum Millet Sorghum Paddy rice Cowpea Paddy rice Cowpea Peanut Onion Peanut Onion Other crops Other crops Source: World Bank staff calculation from EHCVM 2018/19 Within the agricultural sector, farmers are largely engaged production of livestock such as diary or meat is extremely in the production of food crops, particularly rainfed limited (Figure 38). The composition of animal sold varies cereals. Millet and sorghum are the two main staple significantly by region. Diffa region is the key hub for the crops in Niger and are grown by everyone across the commercialization of sheep, goat, and camels. Meanwhile, consumption distribution (Figure 36). On average, rural rural inhabitants in Agadez region only sell camels while households allocate at least 60 percent of their cultivated those in Tahoua region only sell cattle. area to the production of millet alone. This trend is consistent across all regions, except Agadez where onion After agricultural sector (farm and livestock), non-farm is the major crop and is exclusively cultivated the region enterprise presents the second major source of income (Figure 37). More than 20 percent of land in Zinder and across regions. On average, 20 percent of the rural household Tahoua regions is used for sorghum plating, higher than income comes from non-farm enterprise. Trade, service, and other zones. Meanwhile, cowpea is particularly important in electricity and water are the three main sectors of income Dosso, Maradi, and Zinder regions. generation across all wealth groups. Mining is important in the southeast as people flock to the southwestern Niger Livestock plays a small role in income generation for rural to seek fortunes on goldmining5. Niger’s mining industry households although it is relatively more important in has long been centered on uranium extraction. Although the northeast part of the country. On average, livestock since early 2000s, it has begun branching out into gold and generates about 6 percent of rural household income other lucrative subterranean resources, uranium is still the despite the fact that in rural areas 95% of the households largest component of the country’s export, accounting for own or keep animals. In Agadez and Diffa regions where over 30 percent of total export. livestock serves as the main livelihood, it comprises about 23 percent of rural household income. Over 90 percent of livestock income comes from sale of live animals. Other 5 https://theconversation.com/as-more-people-flock-to-nigers-gold-mines-economic-boon-may-become-a-new-migration-risk-75417 44 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Figure 38. Over 90 percent of livestock income comes from Figure 39. The composition of animal sold varies by sale of live animals. Other livestock production remains region. extremely limited. % of animals sold, by region Camels Cattle Goats Other Sheep 0 .2 .4 .6 .8 1 Agadez Diffa Dosso Maradi Tahoua Tillaberi Zinder Source: World Bank staff calculation based on EHCVM 2018/10. Figure 40. Trade, mining, and industry are three major Figure 41. The poorest tend to participate in non-farm sectors contributing to rural income enterprises in trade and services more than the richest. Distribution of enterprise income by industry, by region Distribution of enterprise income by industry, by consumption quintile 1 1 .8 .8 .6 .6 .4 .4 .2 .2 0 0 Q1 Q2 Q3 Q4 Q5 Agadez Diffa Dosso Maradi Tahoua Tillaberi Zinder Agriculture, livestock, fishing Trade Mining and quarrying Transport Industry, electricities, water Services Construction Other Source : World Bank staff calculation from EHCVM 2018/19. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 45 Relatively speaking, remittances represent a smaller part that international remittances contribute more significantly of income, but still significant in rural areas. In 2019, about to household income. Among poor recipients, about a third a third of poor households received either domestic or of their total income derives from money sent by members international remittances. Rural poor are less likely to living abroad (Figure 43). Meanwhile, transfers from receive domestic remittances than their counterparts in domestic migrant accounts for only a fifth of the income of urban areas (Figure 41), but they have a higher probability to poor households (Figure 44). receive international remittances (Figure 42). It is expected Figure 42. Rural poor are less likely to receive Figure 43. But they have a higher probability to receive domestic remittances… international remittances… 60 40 30 40 % households % households 20 20 10 0 0 1 2 3 4 5 1 2 3 4 5 Consumption quintile Consumption quintile Urban Rural Urban Rural Figure 44. Domestic remittances contribute to a fifth Figure 45… while this ratio is a third for of poor recipients’ household income… international remittances 40 60 50 30 40 % households % households 20 30 20 10 10 0 0 1 2 3 4 5 1 2 3 4 5 Consumption quintile Consumption quintile Urban Rural Urban Rural Source: World Bank staff calculation from EHCVM 2018/19 46 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Gender differences in income are substantial. Women are Land holdings are large, but informal. Based on data from less likely to join the labor force and work for pay. Even EHCVM 2018/19, almost all households have access to land, when they do work, they are more likely to work part-time and the average size of land owned is 3.4 hectares and or in the informal sector. Time use constraints related to relatively equitably distributed between the bottom 40 and social norms, including the burden of domestic chores, also upper 60 percent of the consumption distribution (Figure play a role in constraining women’s ability to work. All this 46). However, formal land ownership is very limited. With 80 leads to substantial gender gaps in income and productivity. percent of the cultivated land are owned, only 0.8 percent Over half of the income from female-headed households have titles. The land rental market is close to nonexistent. come from remittance. Income share from farming is According to (Deininger, Savastano, & Xia, 2017), 7 percent 24% and 42% for female- and male-headed households, of households in Niger rent land, among which 25 percent respectively. Female-headed households have smaller land being landless. Land rental allows the land poor to access and own less livestock. Enterprises owned by women show land. The fact that land holding varies little between poor 61 percent lower profits than those owned by men6. Gaps and non-poor, suggesting that the size of land holdings are also large for wage earners. The gender gap in earnings is not a driver of income inequality. Only 5 percent of the is estimated at 29 percent when comparing similar male 270,000 hectares of land with irrigation potential in Niger is and female workers. equipped to provide water to crops (AQUASTAT, 2019). The reliance on agriculture, combined with one main Households own a large number of animals, 12 on average. harvest season, results in strong seasonality in rural labor Livestock are the main asset and factor determining food demand, income, and prices (Mortar and Tristian, 2005). security and poverty reduction in agropastoral and pastoral This pattern generates highly seasonal consumption, with livelihood zones in Niger (FEWS, 2017) and 95 percent of consequences for nutrition outcomes, particularly for poor rural households own some types of animals. Ruminants households. Indeed, the proportion of households with low (cattle, sheep, goat, camels, horse, etc.) form the majority of dietary diversity declines by nearly half during the harvest livestock owned by rural households, followed by poultry season (Somes and Jones, 2018). (chicken and duck, etc.) across all wealth groups. According to FEWS NET (2017), The average national livestock herd size in Niger between 2010–2014 included 10.3 million cattle, 10.6 2.1.2. HUMAN CAPITAL ASSET AND PRODUCTIVITY million sheep, and 13.9 million goats. Almost all regions ARE LOW in Niger have substantial stocks, except the urban capital district and the desert Agadez Region. Tillabéri, Zinder, Across regions, human capital outcomes are very low, for Tahoua, and Maradi Regions are the livestock powerhouses, both bottom 40 percent and the rest of households. In accounting for close to 80 percent of the cattle and small 2018, Niger ranks 155th out of 157 countries on the World ruminant population, while Diffa and Dosso Regions have a Bank’s Human Capital Index, falling well below the low- lower but still significant share (Figure 42). Almost all of the income and regional averages. The country’s performance livestock in Agadez region are ruminants. on educational indicators is particularly weak (See chapter 4 for more details). Niger also has the highest fertility rates in the world. Average births per woman have been 7 from 2016-2018, compared to 4.8 in SSA region and 4.6 in low income countries in 2017 (World Development Indicator 2019). The average age of first marriage for female is 16.7. This early average age of marriage and first birth as well as the high fertility rate increase health risks for women, while reducing the amount of time they have to fully participate in economic opportunities. 6 Economic Impacts of Gender Inequality in Niger (2018) NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 47 Figure 46. Land size is equitable distributed between Figure 47. Agadez and Diffa regions have the highest the bottom 40 and upper 60 percent of the consumption number of cattle and small ruminants while poultry is distribution. raised in other regions. Average size of land owned (hectares), by region Number of livestock, by region 4 15 3 10 2 5 1 0 0 Agadez Diffa Dosso Maradi Tahoua Tillaberi Zinder Agadez Diffa Dosso Maradi Tahoua Tillaberi Zinder Top 60% No. of poultry Bottom 40% No. of ruminants No. of non-ruminants Source: World Bank staff calculation from EHCVM 2018/19 2.1.3. ACCESS TO MARKETS AND FINANCIAL SERVICES Input use—fertilizer, improved seeds—is very low. While it has been empirically proven in the literature that increasing IS LIMITED the use of fertilizer could foster productivity growth and increase agricultural income among poor households Most households have poor access to domestic and (see Larson 2016), the rate of input use is low in Niger. international markets. As a landlocked country, Niger must About 22 percent of households in the top quintile of the rely on land, air, or river transportation to reach its principle, consumption distribution use inorganic fertilizer, while only non-African trading partners , specifically, the European a little over 10 percent of the bottom 40 percent households Union. Niger’s nearest port, Cotonou, Benin, is located at use it. By comparison, the proportion of households using a distance of 1,050 kilometers from Niamey, which impacts fertilizer rose from 38 percent in 2009/10 to 48 percent in the country’s access to international markets. The country 2011/12 in Niger (Theriault, Smale, & Haider, 2018). Use of is also sparsely populated and ranks low on road density pesticide is higher among rich households and Agadez and and market accessibility indices relative to other low- Diffa regions. income countries. According to the Logistics Performance Index, Niger ranked 100 in 2016 and 157 in 2018, presenting Commercialization rates are also very low. On average, a deteriorated logistics environment. This makes accessing Nigerien farmers sell 17.5% of their production on the domestic markets challenging. As mentioned in Chapter 1, market: 3.8% of Millet, 3.2% of Sorghum, and 27% of Paddy electricity access is also very poor, while digital connectivity Rice. The portion is lower in poor households (Figure 48). is slightly improving. For comparison, the corresponding values are 31, 42 and 39 percent for Niger, Ugandan and Malawian farmers, respectively. 48 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Figure 48. Commercialization rate is low in Niger 0.45 0.40 % of agricultural production being sold 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 Q1 Q2 Q3 Q4 Q5 National Burkina Malawi Uganda Faso Niger vs. other AFR countries Source: World Bank staff calculation based on EHCVM 2018/19 and MAFAP 2017 Rural financial services are limited. As discussed in Strong norms of mutual assistance help households cope chapter 1, only 6 percent of the poorest 40 percent of the with shocks, but public safety nets or private insurance population have a bank account. In addition, households markets are largely absent, outside of small pilots. capacity to borrow or to save has declined significantly over Households primarily cope with shocks by helping each time (FINDEX 2014-2017). However, access to mobile banking other though remittances, gifts, and transfers. Norms of has increased slightly and signaled an encouraging trend. mutual assistance are strong. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 49 2.2. Opportunities for Nigerien rural income growth The predominance of agriculture in incomes and the slow pace of transition mean that opportunities for income growth in the immediate term are largely in agriculture, particularly if commercialization and diversification are encouraged. Growth in rural non- farm work offers some opportunities, mostly from mining sector. The need for faster structural transformation requires increasing emphasis on rural to urban migration. 2.2.1. THERE IS A POTENTIAL FOR INCREASING with a reduction in forest area. Increases in land to crop production may not be optimal if they come at the expense PRODUCTIVITY OF EXISTING CROPS of reductions in pasture or forests given the income that can be derived from livestock and forest products, and the Crop production could increase from cultivated land climate benefits from forest cover. expansion but it should be expanded through land restoration to avoid converting pasture or forest land to Land productivity varies little across regions. On average, crop production. The country is endowed with 47 million ha land and labor productivity is very low among its comparable of agriculture land in 2016, of which about 11 million hectares countries (Table 8). In 2014, the land productivity is $77/ha, are currently under cereal cultivation (World Development compared with $222/ha in Burkina Faso, $101/ha in Mali, Indicator 2019), the remaining agriculture land is non- and an average of $322/ha in sub-Saharan Africa. Within cereal cultivation and pasture. In addition, about 1 million the country, however, little regional heterogeneity is found. hectares are currently forested. The constraint to how much Productivities vary little across zones and are almost the land a household farms is often not the amount of land same across crops, except paddy rice and onion. The two that there is, but the amount of land that can be farmed most exported crops have higher land productivity than with the labor available during peak production periods. other crops (Figure 50). Adoption of mechanization could increase the area farmed. Increases in cropped area in the past have in part come Table 8. Productivity trend Land productivity (in constant 2004–2006 US$) Labor productivity (in constant 2004–2006 US$) Country/Region 1990 2000 2010 2014 1990 2000 2010 2014 Sub-Saharan Africa 198 255 321 332 1,326 1,597 1,885 2,105 Burkina Faso 111 146 196 222 297 296 388 378 Mali 44 51 95 101 822 885 1,345 1,347 Niger 35 51 74 77 500 545 793 711 Senegal 111 147 198 178 374 397 428 314 Uganda 382 433 454 434 584 588 517 421 Ethiopia 99 145 250 286 255 218 300 315 Source: World Bank (2020). 50 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Figure 49. Only a quarter of agricultural land is currently Figure 50. Land productivity varies little across regions. used for cereal production. Share of agriculture land under cereal production Land productivity (FCFA/Ha; log), by region 50 15 40 30 10 ha (million) 20 5 10 0 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Agadez Diffa Dosso Maradi Tahoua Tillaberi Zinder Land under cereal production Millet Sorghum Agriculture land Paddy rice Cowpea Peanut Onion Source : World Bank staff calculation from World Development Indicator 2019. Source: World Bank staff calculation from EHCVM 2018/19 Yield potential is low compared to its peers, but potential Investments in natural resource management and improvement with irrigation and technlogogy exists. Niger’s improved production practices can also increase yields. average cereal yield is among the lowest in the region (World Investing in improving soil quality will also increase Bank, 2016). The national average yield in Niger for rain- yields (Naab, Mahama, Yahaya, & Prasad, 2017). Increase fed pearl millet is 0.53 metric tons per hectare, with a yield household access to plough and improved technologies can potential of 1.18 MT/ha without irrigation. With irrigation, the provide room for increasing labor productivity (Mahajan, potential increase signficantly to 4.03 MT/ha. For sorghum, 2019) (Murray, Gebremedhin, Brychkova, & Spillane, 2016). the national average yield is 0.35 MT/ha, with a rain-fed The previous section also indicates that there is room to potential of 3.3 MT/ha and an irrigated potential of 5.62 MT/ increase investments in fertilizer and pesticide. ha (GYGA, 2016). With low yields, farm sizes need to be large to provide enough income. This highlights the potential for increasing labor productivity through mechanization. 2.2.2. GREATER COMMERCIALIZATION CAN GROW Technological innovation can also increase long-term yield INCOMES AND INCENTIVIZE INVESTMENTS potential, and there is a particular need to develop seeds that take into account increasing variability in the timing At the national level, Niger consumes most of what it of rain during the growing season; the labor constraints produces, and at household level, Nigeriens consume farmers face at peak production times; and low levels of most of what they produce as well. Nearly all the cereals education. Investments in research and development (R&D) produced are consumed domestically. Except for rice, very would have payoffs in the long run rather than in the next little cereal is imported (Figure 51), making Niger largely 5-10 years, as there is little availability of improved cultivars self-sufficient in cereals. In fact, Niger has one of the in sorghum and millet, which are the main staple crops in highest self-sufficiency ration in Africa (Figure 52). This Niger. Other opportunities for income growth in Agadez and self-sufficiency ratio at the national level is matched at the Dosso regions, such as the development of livestock value household level as much of the food produced by Nigerien chains (discussed below) or even rural-to-urban migration, households is consumed by the households that appear more attainable in the medium run and may also be produce it. a more cost-effective long-run development strategy. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 51 Figure 51. Except for rice, very little cereal is imported. Niger: Evolution of cereal imports - tons (1980-2017) 600000 400000 200000 0 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Flour of wheat Maize Millet Rice Sorghum Wheat Other Source: World Bank staff calculation based on FAOSTAT. Figure 52. Self-sufficiency ratio is higher in Niger than in its peers, except Mali. 2 1.6 Self-suffieciency ratios (-) 1.2 0.8 0.4 0 West East Burkina Ghana Mali Niger Nigeria Ethiopia Kenya Tanzania Uganda Africa Africa Faso Source: (van Ittersum, 2016) 52 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Without commercialization, households have little access markets. It is also possible that both need to happen incentive for increasing production within the existing simultaneously. Recent experimental evidence comes from crop mix. As shown in Chapter 1, 13 percent of Nigerien Benin and Senegal, where market incentives were increased population are food insecure and would gain from while other variables stayed equal. Results show that, when production increases. Increasing food production is contracts secure the market (Benin) or grading provides particularly important in Zinder, Maradi, Dosso, and Tillaberi incentives for quality (Senegal), farmers respond by regions where food insecurity rate is the highest. On average, increasing investment (see Arouna et al. 2019 and Bernard demand analysis shows that among rural households only et al 2018). In addition, stakeholder discussions underscored 5-15 percent of increases in crop production would be that output market access had been a key component consumed within the household (Figure 53). of successful interventions to increase production. In contrast, some projects that focused on increasing crop When households can sell their production in the market, production with no attention to marketing saw local prices their investments are higher. There is a strong correlation collapse and production remain unsold, dissuading farmer between market access and households’ investments in investments the following season. Although constraints to inputs. Input use increases with commercialization and market access and productive investments often need to access to markets more than with income (World Bank be addressed at the same time, there is some evidence that 2020). An important question is whether market access can addressing market access first can allow some households incentivize higher rates of investment or whether high rates to also invest more and increase their incomes. of investment in production allow households to better Figure 53. Marginal food budget shares in rural and urban areas Rural Urban 0.400 0.400 0.350 0.300 0.300 0.250 0.200 0.200 0.150 0.100 0.100 0.050 0.000 0.000 Rice Sorghum Wheat Maize Millet Fruits & vegetables Oils Meats Fish & seafood Dairy Beverages Rice Sorghum Wheat Maize Millet Fruits & vegetables Oils Meats Fish & seafood Dairy Beverages Source: World Bank staff calculation based on EHCVM 2018/19 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 53 2.2.3. COMMERCIALIZATION GOES HAND IN HAND Diversification also requires cereal market development; the two need to go together. Lack of cereal market WITH DIVERSIFICATION development raises the cost of consuming purchased goods in rural areas and results in substantial price volatility. Commercialization will require diversification into animal This makes growing crops for sale and purchasing food for products and new crops, as the demand for domestically consumption both costly and risky (Fafchamps 1992). produced cereals—sorghum and millet—is limited. Demand analysis undertaken for this diagnostic shows that urban households primarily consume imported cereals—rice—not 2.2.4. GROWING LIVESTOCK INCOME HOLDS PROMISE much domestically produced ones (millet and sorghum). As a result, urban growth (be it driven by population growth The growing demand for animal-based food provides an or income growth) does not result in higher demand for important opportunity for a profound transformation of domestically produced cereals (Figure 48). In contrast, the livestock sector and a strong incentive to maximize urban growth increases consumption of animal products livestock income growth potential. In the next 30 years, and fruit, and vegetables. The more households increase overall consumption of all animal products is estimated their commercial production of these goods, the more to double. Demand for beef, poultry, and pork can even the domestic demand for cereals may increase (as those be triple (Error! Reference source not found.). In 2010, the rural households may start purchasing cereals, if their food total supply of livestock-derived food (LDF) in Niger was 192 consumption preferences do not change), but diversification kcal per person per day. Dairy accounted for 48 percent of is needed first. the supply while meat was 52 percent and eggs less than 1 percent. Within meat category, beef had the highest share, at 70 percent of meat supply followed by lamb at 26 An export-oriented agricultural growth strategy is percent. Under the scenario of moderate economic growth, possible as Niger is surrounded by countries with which total supply of LDF increases to 268 kcal in 2030 and to it trades. Like other Sahel countries, Niger has a revealed 442 kcal in 2050. By 2050, the share of dairy is projected comparative advantage in groundnuts, millet, sorghum, to reduce to 35 percent while the share of meat, mostly rice, fruits, even at current yields and costs of transacting. beef, will increase to 64 percent of all LDF supply. According It is also shown that the country has a natural comparative to (Enahoro, Njiru, Thornton, & Staal, 2019), production of advantage for the production of cattle and small ruminants beef and dairy products will not be sufficient to supply for export to the coastal countries of the Gulf of Guinea and the domestic market in the near future. By 2050, dairy Central Africa (OECD 2008). Regional demand for livestock production can only meet less than half of the domestic products continues to grow. Given the small size of the demand (Table 9). domestic market, a focus on products where there is also comparative advantage may make most sense in the long run. 54 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Table 9. Projection of demand for different livestock-derived food types in Niger 2010 2030 2050 kilocalories per capita per day (% of total supply) Beef 69.57 114.63 219.81 Pork 1.14 1.93 3.88 Lamb 25.30 33.15 49.32 Poultry 3.13 5.46 11.24 Dairy 91.66 111.82 155.69 Eggs 1.31 1.73 2.59 All meat 99.13 155.17 284.25 All LDFs 192.10 268.72 442.54 Source: (Enahoro, Njiru, Thornton, & Staal, 2019) IMPACT model results for moderate economic growth, no climate change (Middle No CC) scenario. Figure 54. Model projections of production and demand in Niger National demand and production of Beef in 2010 (reference, A) and National demand and production of Dairy in 2010 (reference, A) and 2050 (various scenarios, B-P) 2050 (various scenarios, B-P) 3000 7000 2500 6000 5000 2000 4000 1500 3000 1000 2000 500 1000 0 0 A B C D E F G H I J K L M N O P A B C D E F G H I J K L M N O P Total Beef Demand, ‘000 MTs Total Dairy Demand, ‘000 MTs Total Beef Production, ‘000 MTs Total Dairy Production, ‘000 MTs Source : World Bank staff calculation from World Development Indicator 2019. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 55 In addition, Niger plays a critical role as a major livestock exporter to neighboring countries. Niger has the highest 2.2.5. RURAL NON-AGRICULTURAL INCOME GROWTH herd population in the Sahel region, with an estimated 11.4 WILL PRIMARILY COME FROM MINING AND MIGRATION million heads of cattle by 2014. The main animal breeds kept in Niger are: Cattle, goats, sheep, camels and horses. The prospects for growth in non-farm, non-tradable The majority of exported livestock goes to Nigeria, through services are weak across the country. Low population principal entry points that follow major roads and rivers, density limits demand for non-tradable rural services such even though animals are often trekked across borders to as restaurants and hair-dressers, even if spending power avoid formal customs procedures. Niger’s large wholesale were to increase through a growth in agricultural income. markets are located along the border with Nigeria, from The exception is Niamey, with higher population. where supply is transported to export markets in northern Nigeria, like Jibia and to consumption markets in southern Mining has potential to increase non-farm incomes, if Nigeria such as Lagos, Ilorin, Ibadan, Port-Harcourt. Livestock production practices improve. Mining brings direct job exports to Nigeria account for more than 90 percent of benefits and also indirect benefits, as it increases demand overall livestock exports,98 with Ghana and Cote d’Ivoire as for goods and services around mining sites. Industrial other key consumers of Nigerien livestock, especially small mining accounts for a larger share of economic output (more ruminants. Traders from Nigeria purchase animals in Niger than half of Nigerien exports in 2018), but employment in when selling cereal on local Nigerien markets. Therefore, the artisanal and small-scale mining sites is much higher, changes or shocks to this symbiotic relationship between and it brings bigger welfare benefits: areas around artisanal Nigerien and Nigerian traders can disrupt normal cereal and small-scale mining sites have seen significantly faster supply systems (FEWSNET). consumption growth than areas without mining (Bazillier and Girard 2017). As of now, over 450,000 Nigeriens working As regional demand continues to grow, Nigerien livestock in these operations, and more than 20 percent of them producers will have access to an even larger market. depends on mining as an important income source (UNECA Estimated demand for poultry in West Africa could triple 2019). However, for artisanal mining to reach its potential to reach approximately 14 million tons by 2050. Similarly, to sustainably improve the well-being of the rural poor, the demand for beef and small ruminant meat could increase mode of production currently used will need to change. up to 50 percent to nearly 4 million tons. In a normal There is also robust evidence of a negative correlation development progress of trade as today, the regional between proximity to mining sites and school enrollment, production will supply less than half of the demand for as income today is preferred over investments in human poultry and about 20 percent less for beef and ruminants. capital that will increase income-earning potential tomorrow (Stoeffler Q. and C. Guirkinger 2019). The incentive for increasing livestock production is clear, but production growth must not only come from Remittances are an important source of income, an increase in cattle stock but also from improvement particularly for male youth who are those most likely to of productivity technologies. Using the same historical migrate, but at current trends the potential for growth growth rate of the past two decades, the stocks of all in migration looks weak. Networks are a strong driver of livestock is estimated to triple in the next 30 years. As migration, and with a large regional network it is likely that mentioned above, a larger animal population is needed, regional migration will continue to be important (see De although this will increase pressure on natural resources Brauw et al 2014, Wouterse 2012, and Meghir et al 2019). and pose more challenges to the environment and human However, rates of regional migration have been declining health. Therefore, it is increasingly more important to invest over the past two decades. Migration from rural to urban in livestock productivity, which could be realized through areas is currently too low in Niger to be a pathway for livestock genetic improvement, animal nutrition, zoonotic income growth for many rural households. Faster job disease control, and supporting policies. creation in urban spaces is needed to encourage faster rural to urban migration. Growth in formal jobs in urban areas has been 3.7 percent per year over the last five years, not much higher than population growth. 56 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 2.3. Conclusion The agriculture sector, including farming and livestock, Improvement of the rural income and prospects for continues to be the largest income sources for rural a sustainable economic transformation will involve households in Niger. Building on existing data, this chapter ambitious agriculture and rural economy reforms. These provided a detailed diagnostic of rural income growth in reforms should help tackle all the major constraints, Niger. In Niger, 95 percent of the poor live in rural areas. including improving access to land, technology, markets, and Therefore, understanding rural income sources will provide finance. Farmers should also be guided to move from low important insight for the design of poverty alleviating value crops to high value crops. Currently, Niger, and most reforms and programs. It is estimated that 75 percent of Sub-Saharan African countries are struggling to meet their Niger’s workforce is employed in farming and livestock food needs, resulting in the use of substantial resources activities. Productivity is very low, due to several constraints, to import food. The country ambitious should be to reform including limited access to productive and modern inputs, the agricultural sector so as to reduce this dependance on limited mechanization, limited diversification, negatives imports, but more importantly, with improved productivity, shocks related to global warming, reliance on rainfall. Access Niger can generate important income by boosting its to market is also a big issue, with poor households having exports of food products, including in the regional market. limited access to domestic and international markets. As a The need for faster structural transformation requires consequence, commercialization is low, most households increasing emphasis on rural to urban migration. But to be just produce for their own consumption. Farmers are also successful, migrant must be endowed with minimum skill. struggling to access financial market. Land holdings are large, but informal. A land reform is needed to solve this problem, and if possible, address the important issue of landless. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 57 3 COMPOUND EFFECTS OF SHOCKS ON WELFARE 58 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION The reliance on rainfed crop and livestock production shocks such as conflict, natural hazard, crop and animal results in high levels of income risk. About 40 percent of diseases concentrate in certain regions only. For example, Nigerien households reported experiencing a shock, either conflict is felt the most in Diffa region, while draughts are covariate or idiosyncratic shocks7, in the past 3 years. particularly reported in Tillaberi and Tahoua regions with Assessing vulnerability in different households is important more than 40 percent of the population affected, and flood as it provides policymakers with guidance on how to is reported mainly in Dosso and Agadez regions. During efficiently allocate resources to prevent households from this period, crop diseases attack Zinder region, and animal falling into poverty. diseases spread across Agadez, Diffa, and Dosso regions. While health shocks and high food prices are widespread At national level, the most important shocks Nigerien across Niger, all other shocks are quite localized (Table households face are natural hazards, health shocks, high 10). On average, one in every four Nigeriens experienced food prices, and insecurity. As commercialization rate and a health shock, defined as illness or death of a household fertilizer input for agricultural production are extremely member, over the past 3 years. However, the prevalence is low (see chapter 2), it is expected that very few households highest in Tillaberi, Maradi, and Tahoua regions. Similarly, reported decline in agricultural product prices and increase households across region experienced high food prices, but in agricultural input prices as household shocks. This the situation is most acute in Tillaberi region where more chapter aims to explore these most commonly reported than half of the population are net food buyers. However, shocks in Niger and their impacts on Nigerien welfare. Table 10. Type of shocks by region % of population affected Illness/Death of a Crop Animal High food Violence/ Locus Draught Flood household member diseases diseases prices conflict attack Agadez 25% 4% 17% 2% 15% 11% 1% 0% Diffa 18% 20% 4% 2% 11% 15% 32% 10% Dosso 25% 32% 22% 3% 12% 10% 0% 6% Maradi 29% 22% 5% 8% 8% 8% 1% 3% Tahoua 29% 42% 4% 8% 4% 12% 0% 8% Tillaberi 32% 43% 6% 8% 7% 21% 1% 4% Zinder 15% 28% 4% 12% 7% 8% 0% 10% Niamey 27% 1% 5% 0% 0% 11% 0% 0% Source: World Bank staff calculation based on EHCVM 2018/19 Note: The low prevalence of violence in Tillabéri and Tahoua regions is surprising. This likely reflect the fact that administrative data (Such as ACLED) are more precise in capture conflicts vs standard households surveys. 7 Covariate shocks affect many households in one place at the same time, like drought, conflict, and price shocks., and idiosyncratic shocks affect individual households rather than the whole community like loss of job, death or illness of family member, and divorce. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 59 3.1. Main Shocks to Welfare 3.1.1. DROUGHT IS THE MOST FREQUENTLY population in high food insecurity because of the loss of crops and livestock. Seasonal flooding is another challenge. REPORTED SHOCK Flash floods in 2019, particularly in Agadez, Diffa, Maradi, and Zinder, have affected over 259,000 people and resulted Natural hazards such as drought or flash floods pose in the disruption of livelihoods and loss of livestock. major challenges to Nigerien agricultural production, including both farming and livestock rearing. Niger is one Drought is the most frequently reported shock that of the hottest countries in the world. It has three basic Nigeriens face. According to EHCVM 2018/19 data, 35 climatic zones: the Saharan desert in the north, the Sahel percent of the rural population and 4 percent of the urban to the south of the desert, and the Sudan in the southwest population reported drought as a key shock in the past 3 corner. The intense heat of the Saharan zone often causes years (Figure 55). This is consistent with the CHIRPS data, the scant rainfall to evaporate before it hits the ground. which shows poor precipitation level in 2017 (Figure 56). The Moreover, the rainy season is very short, typically during a lowest rainfall level in 2011 in Figure 56 corresponds to the single two-month period. Thus, a decline in rain level could 2011 Sahel drought, when Niger’s crops failed to mature in have serious effects on households production. In fact, the heat and resulted in famine. the severe drought in 2011 pushed nearly half of Nigerien Figure 55. Drought is most frequently reported Figure 56. Total rainfall is poor in 2017 Total rainfall (mm) 40 300 35 250 30 200 25 20 150 15 100 10 50 5 0 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Illness/accident of family Death of family Drought/erratic rains Flood High prices of food products Rural Urban All Source: World Bank staff calculation based on EHCVM 2018/19 Source: CHIRP 2019. 60 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Figure 57. Poor rainfall in the north makes agriculture households vulnerable Source: World Bank staff calculation using CHIRPS and NDVI Low precipitation level in the north makes households more vulnerable. Figure 57 shows the average monthly 3.1.2. SECURITY ENVIRONMENT IS WORSEN rainfall and vegetation index in the planting period (June- Increased violence in bordering regions with Niger, Mali September) in 2018. It ranges from 2.5mm in the north to and Nigeria, and intercommunal conflict have caused an 252mm in the south. This is consistent with the finding from unprecedented humanitarian crisis in the Niger along household survey data that of those who reported drought/ with higher levels of food insecurity and malnutrition. It erratic rains shock, over 97 percent are in the five southern currently harbors 246,000 refugees and 186,000 displaced regions, where the major food crops are planted. On the persons, primarily in Diffa and Tillabéri and more recently, other hand, greenness is higher in Agadez region. in Maradi, which is further exacerbating the country’s fragility. Security conditions have deteriorated in recent But households in the south experienced higher incidences years, particularly in the areas bordering Nigeria, Niger, of rain volatility. In Tahoua and Tillaberi regions, more than and Mali, where armed groups have established bases and 40 percent of the population experienced drought compared carry out repeated attacks against the security forces and to 4 percent in their northern peer, Agadez region. As shown civilians. If the security situation worsens over time, more in Chapter 2, this pattern could be partially explained by people may be at risk of being deprived of employment the fact that a relatively larger share of household income opportunities and access to basic services, and at risk of in these southern regions was derived from farming and falling into poverty. livestock activities (Figure 34) thus making them more vulnerable to climate shocks compared to households in Conflicts and fatalities concentrate in Diffa and Tillaberi the north. regions (Figure 59). In these two regions, security incidents consist most of violence against civilians and battles with a high number of fatalities. A third of Diffa residents confirmed that they experienced conflicts in the past three years. During the survey period of EHCVM 2018/19, there were 627 and 172 fatalities in Diffa and Tillaberi regions, respectively. At the time of writing of this report (June 30, 2020), these numbers have increased to 900 and 907 respectively. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 61 Figure 58. Security incidents roar Figure 59. Fatalities concentrate in Diffa and Tillaberi regions. 400 350 300 250 200 150 100 50 0 Jun-18 Jul-18 Aug-18 Sep-18 Oct-18 Nov-18 Dec-18 Jan-19 Feb-19 Mar-19 Apr-19 May-19 Jun-19 Jul-19 Figure 60. Conflicts in Diffa and Tillaberi regions consist mostly of violence against civilians. 180 160 140 120 100 80 60 40 20 0 Agadez Diffa Dosso Maradi Niamey Tahoua Tillaberi Zinder Battles Explosions/Remote violence Protests Riots Strategic developments Violence against civilians Source : World Bank staff calculation based on ACLED 2019. 62 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 3.1.3. HEALTH SHOCK AFFECTS THE POOR POPULATION 3.1.4. PRICE IS ANOTHER IMPORTANT SOURCE THE MOST OF SHOCKS Health shocks, including the current coronavirus outbreak, High food prices, typically accompanying a plummet in could have adverse impacts on household welfare. agricultural production due to other covariate shocks Illness of a household member can increase the risk of a such as drought, is another major shock to Nigerien household becoming destitute if there are significant out- households. As 40 percent of Nigeriens are net food buyers, of-pocket (OOP) healthcare expenditures. Although only increase in food prices are likely to have negative impacts 55.4 percent of people with illness currently seek health on households’ welfare. Particularly for the poor net food services because most of them opts for self-medication, buyers who often allocate a large budget share on food, the household to which a sick individual belongs may food prices could be one of the decisive factors on whether still forgo earnings if there are work-days lost by the they would be food insecure. However, due to the lack of sick individual or his informal caregivers. If households frequent and detailed food prices, we were not able to indeed seek health services, OOP expenditure cost could quantify the effects of food prices on Nigerien households wipe out at least 4 percent of the poor’s income which is in this section. already at a vulnerable level. (2018) finds that the poorest households, such as those in Sub-Saharan French-Speaking It is, nevertheless, expected that more urban households Africa countries (SSAF) may be the most vulnerable to are affected by food prices than rural households as they health shocks. As Figure 55 shows, illness, accident, and are more likely to be net food buyers. Across regions, a death of family members are the most commonly reported higher share of urban population experiencing high food idiosyncratic shocks, especially by urban households. A price than rural population (Figure 61). It should be noted substantial 26 percent of the urban households reported that Zinder and Tillaberi regions have very few urban illness and accident of family members as the most serious households in the data set, thus, the urban numbers shown shocks happened in the past three years. A shock such as here are indicative only. death of a family member affected 8 percent of the rural households and 12 percent of the urban households. While the prevalence of health shock does not vary by poverty status of the households, female-headed households report illness/death of household members as shock more frequently. On the other hand, as men are more involved in production activities, male-headed households report weather shocks more frequently. This may be explained by smaller household size and higher dependency ratio in female-headed households. For households involved in agriculture production, the average household size is 4 in female-headed households and 6 in male-headed households. The dependency ratio is 2.3 and 1.7 in female- and male-headed households, making female-headed households more vulnerable. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 63 Figure 61 : Urban households are more likely to be affected by food prices Urban Agadez Rural Urban Diffa Rural Urban Dosso Rural Urban Maradi Rural Urban Tahoua Rural Urban Tillaberi Rural Urban Zinder Rural Urban Niamey Rural 0 0.05 0.1 0.15 0.2 0.25 Box 3 : Safety nets in Niger In Niger, safety nets are mainly divided into food distributions households. These response mechanisms are respectively and safety nets for resilience8. Food distributions, in the managed by the Food Crisis Unit (CCA) and the Safety Net form of targeted distributions or subsidized cereal sales, are Unit (CFS), established within the National Institution for the short-term and seasonal responses to food crises, targeted to Prevention and Management of Food Crisis (DNPGCA9). transiently food insecure households. Safety nets for resilience, in the form of cash transfers with accompanying human capital In Niger, food distributions and safety nets for resilience use and productive inclusion measures, and cash for work, are long- different targeting mechanisms. In the first case, transiently term and predictable transfers targeted to poor and vulnerable food insecure households are geographically targeted based on 8 Other safety net mechanisms include school feeding programs, subsidized agricultural inputs, nutrition, and health programs. 9 Dispositif National de Prévention et de Gestion des Crises Alimentaires. The DNPGCA overall objective is to contribute to improving the resilience of vulnerable groups to climate change, crises and disasters. 64 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION the annual Integrated Phase Classification (IPC) (category 3 and regions of Niger. In 2017, 600,000 vulnerable people, including 4), while moderately and severely food insecure households are ASNP1 and WFP resilience beneficiaries, received long-term, identified through a community-based approach (Household predictable cash transfers. This represented 6.5 per cent of the Economy Approach). Programs targeted to the poor and Nigerien population living under the poverty line and 10.9 per chronically vulnerable (ASNP1, ASNP2) use a mix of geographical cent of those living in chronic poverty (PER 2020). In 2018, due targeting and proxy means test (PMT). Geographical allocations to a pipeline break and movement restrictions in some regions, are proportional to poverty rates based on the most recent the ASNP1 suffered some implementation delays (ASNP1 ICR EHCVM (EHCVM 2014 for ASPN1 and the EHCVM 2018 for ASPN2). 2020). The Second Adaptive Safety net Project (ASNP2), approved Extremely poor households are identified through PMT, whose in 2019, will cover 156,000 poor households (approx. 1,1 million efficiency to identify the poorest was evaluated at 82 per cent. people) through cash transfers and CFW for resilience by 2026. Evidence has shown the appropriateness of these targeting mechanisms, as PMT performs better to identify persistently Since 2016, the government of Niger and development partners poor households, while HEA performs better to identify have been supporting development of the foundations transiently food insecure households10. of an adaptative social protection system, including the development of a Unified Social Registry (USR) as well as So far, food distribution has been the default response ruled-based mechanisms for scaling up the safety net systems modality used by the Government of Niger to tackle food to respond to slow and rapid onset shocks. Led by the DNPGCA, insecurity. Food distribution coverage fluctuates over time, with the USR is a key component of an adaptive social protection peaks in years of crises. In the period 2013-2017, more than 2 system, as it provides a unified database of socioeconomic million food insecure people were targeted each year with information which can help to select beneficiaries of safety nets food distributions, also in years classified as normal or good, programs more accurately. Progress is ongoing to strengthen the covering almost 90 per cent of the households classified as institutional, regulatory, and operational framework for the USR. food insecure. In the last ten years, three peaks in the number A common database of around 640,000 households (approx. of beneficiaries targeted with general food distributions were 5.6 million people) was available in 2020, which included more registered in 2010 (9 million people), in 2012 (4 million people), than 210,000 urban households in the context of the short-term and in 2015 (3 million people). response to the COVID-19 pandemic. Safety net spending and coverage is characterized by an In 2020, the national adaptive social protection system has inefficient program mix between long-term predictable been effectively scaled-up to respond to the economic impact transfers, and shock responses11. Even during normal years, of COVID-19. The ASNP 2 provided emergency cash transfers when no major crisis strikes, long-term safety net coverage is to 150,000 affected households (approx. 1,050,000 individuals) only one third of shock response coverage. Implementing shock in urban areas and 250,000 households (approx. 1,750,000 responses in the same areas year after year and targeting the individuals) in rural areas. WFP and UNICEF contributed to same beneficiaries is an ineffective way to address chronic this scale up, targeting 120,000 households (approx. 840,000 poverty and the deep causes of recurrent food crises. individuals). The scale-up of the adaptive social protection system allowed for covering about 65 percent of the 5.6 million The proportion of safety net assistance delivered through individuals impacted by the crisis. The ASNP 2 will extend the cash transfers targeted to the poor has gradually increased support to 30,000 COVID-affected households beyond the with the launch of the Adaptive Safety Net Project (ASNP1). emergency phase. Between 2013 and 2019, the ASNP1 provided cash transfers to 1,088,213 individuals in poor and fragile areas in the eight 10 Schnitzer, 2018. 11 Public expenditure review (PER) of the social protection sector (2020) NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 65 3.2. Household coping strategy To cope with shocks, the most commonly used strategy The strategy of using savings is significantly underused in for Nigerien household is to rely on friends and family, Niger compared with other Sahel Countries. In Niger, less sale of livestock, and savings, a sign that coverage formal than 30 percent of the population rely on savings to recover society nets programs are limited in Niger (Box 3). Chapter from shocks compared to over 40 percent in Burkina-Faso, 2 shows that 95 percent of Nigerien households own some Mali, and Chad. As mentioned in chapter 1, households’ type of livestock, yet on average, less than 20 percent of the capacity to save has also declined over time. total stock of animals were sold in the market to generate income in the last 12 months. This reflects the fact that To get through difficult times, rural households sell Nigerien keep livestock as one of the risk-coping strategies livestock. 22 percent of rural households and 4 percent and saving mechanisms. The pattern is consistent with the of urban households stated that they sold their livestock literature on livestock assets in SSA countries (see Pica to mitigate the shock. 7 percent of the rural households Ciamarra et al, 2011, Maziku et al. 2017, and Jha 2019). reported they ever reduced consumption in order to manage a certain risk (Figure 64). No significant difference is observed for the use of other strategies. Figure 62. Most households rely on support from family Figure 63. Savings is underused in Niger compared to other and friends, sale of livestock, and savings to manage risks. Sahelian countries. 0.7 No strategy 0.12 0.6 Engaged in spiritual 0.06 activities 0.5 Sold livestock 0.19 0.4 Migrated 0.07 0.3 Purchased less 0.06 0.2 expensive food Reduced consumption 0.06 0.1 Support from relatives 0 0.19 or friends Burkina Mali Niger Senegal Chad Used savings 0.18 Faso Use savings 0 0.05 0.1 0.15 0.2 0.25 Support from relatives and friends Source: World Bank staff calculation based on EHCVM 2018/19. 66 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Figure 64. Rural households sell livestock Figure 65. Poor are more likely to get support from family or friends 0.25 0.3 0.22 0.20 0.25 0.2 0.19 0.17 0.2 0.15 0.14 0.13 0.12 0.15 0.1 0.07 0.1 0.05 0.04 0.03 0.05 0 0 Used savings Support from relatives or friends Reduced consumption Sold livestock No strategy Used savings Support from relatives or friends Reduced consumption Sold livestock No strategy Rural Urban Top 60% Bottom 40% Source: World Bank staff calculation based on EHCVM 2018/19. Coping strategies vary across different categories of however, 17.4 percent of households didn’t take any actions, shocks. To study the households coping behavior by each higher than the other shock categories. In Diffa, the most type of shock, the 22 shocks discussed in the previous often used strategy is no strategy at all. Support from section can be grouped into 7 categories.12 Figure 66 government and engage in spiritual activity are also used presents the most frequently used coping strategies by frequently. In Maradi, the newly emerging conflict zone, type of shock. When coping with employment shocks, the majority of households don’t have means to mitigate natural hazards, health, household issues, and price risk. Households in Tillaberi have to reply on government shocks, using saving is the most frequently adopted support most of the times. strategy by households. With the surge of conflict shocks, 12 Household issues: divorce/separation, suspension of regular transfers from other households. Prices: decline of farm products prices, high prices of agricultural inputs, high prices of food products. Natural hazards: drought, flood, crop disease, animal disease, locust attacks or other crop pests, landslide. Employment: Significant loss of household’s non-farm income, bankruptcy of household non-farm enterprise, significant loss of wage income, loss of wage employment. Health: illness/accident of family member, death of family member. Crime and safety: theft of money/assets/crops/livestock, farmer/Livestock farmer conflict, armed Conflict/Violence/Insecurity. Other: other shocks not classified above. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 67 Figure 66. Coping strategies vary by shock categories 30.0 25.0 20.0 15.0 10.0 5.0 0.0 Crime and safety Employment Health Household Natural hazards Other Prices issues Used savings Support from relatives or friends Reduced consumption Sold livestock No strategy Source: World Bank staff calculation using EHCVM 2018/19 68 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 3.3. Effects of shocks on welfare In this section, we use multivariate regressions to explore Our multivariate regressions suggest that conflicts are the correlation between different types of shocks and negatively correlated to household food consumption welfare. Households’ idiosyncratic health shock is defined in rural areas. Experiencing any conflict in the past three as having a death of a household member in the past years is associated with a decline of 17 percent in rural 3 years. This information is self-reported from EHCVM areas. The significant impact in rural areas is consistent 2018/19. Drought is calculated from CHIRPS data, which with Ibáñez, A. M., & Vélez, C. E. (2008). They study the provides a measure of monthly rainfall for each 5.7 by 5.7 impact of conflict on welfare in Columbia and find that km pixel in the country. Households affected by drought the welfare losses caused by displacement are 37% of the are defined as those living in areas with precipitation levels net present value of rural lifetime aggregate consumption. below one standard deviation from its 10-year monthly average for at least one month in the planting season. Two Droughts show a negative and significant correlation other definitions of drought – severe drought identified with consumption of affected households in both rural as two consecutive months below their respective 10- and urban areas. Households experiencing drought in year average rainfall during the planting season, and urban areas have seen their consumption declined by as self-reported drought shock from EHCVM 2018/19 – were much as 25 percent, and their food consumption reduced used as a sensitivity test. Conflict shock is measured as by 22 percent. Meanwhile, drought-affected households in the logged number of fatalities within 25 km radius of rural areas have their consumption 9 percent lower than households. Data on conflicts and fatalities come from the unaffected households. According to Gerber (2017), the Armed Conflict Location & Event Data Project (ACLED). For impacts and costs of droughts can be extensive in urban robust test, we apply different radius, 15 km and 20 km, as areas. In addition to specific industries (e.g. food and well as self-reported conflict experienced by households beverage), this also puts the service sector (e.g. tourism) from EHCVM 2018/19. Our regressions control for a rich set at risk and could spark social tensions. The urban costs of of household characteristics (e.g. household size, female- droughts will continue to grow in the future due to climate headed household, household age, household head change and expanding urbanization, and are magnified by education level, and the number of people in different relatively higher levels of returns from urban compared age groups), household access to infrastructure (water, with agricultural water use. electricity, improved sanitation), households’ productive assets (land size, livestock, agricultural input), and regional fixed effect. Annex 4 provides more technical details. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 69 However, health shocks is adversely associated with be negatively correlated with welfare. Wagstaff (2007) household consumption in urban areas only. Urban suggests that urban households were more vulnerable households who experienced the death of a household in terms of reduced earned income; while Atake (2018) member have food consumption 13 percent lower than finds that the poorest households, such as those in Sub- other urban households. Our findings is consistent with Saharan French-Speaking Africa countries (SSAF) may be the literature where health shocks are often found to the most vulnerable to health shocks. Figure 67. Conflicts and droughts are significantly Figure 68. The magnitude of correlation between shock and associated with lower household consumption food consumption is even larger. Urban 0.0253 Urban 0.0398 Conflict Conflict Rural -0.0555 Rural -0.166** Urban -0.25*** Urban -0.217** Drought Drought Rural -0.09*** Rural -0.0961*** Urban -0.079 Urban -0.125* Death hh Death hh Rural 0.0128 Rural 0.0297 -0.3 -0.25 -0.2 -0.15 -0.1 -0.05 0 0.05 -0.25 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 Source: World Bank staff calculation based on EHCVM 2018/19 70 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Box 4 : Evidence on the impact of safety nets Several evaluations have been conducted on the impact of and the main beneficiaries of the accompanying measures. safety net in Niger. Below the main lessons learned. Placing women at the centre of the safety net implementation increased their visibility and productive inclusion, hence Cash transfer programs targeting poor households can mitigate strengthening their empowerment. Despite their positive the adverse effects of climate shocks and fostered resilience impact, accompanying measures did not significantly improve by facilitating savings and economic diversification. The children’s physical growth or cognitive development. Evaluation impact evaluation13 carried out in six communes in the regions results suggest that this type of impact may require longer term of Dosso and Maradi during the first phase of the cash transfers interventions and that training, even when combined with cash (2013-15) showed that cash transfers improved household transfers, may not be enough. Improvements in relevant services welfare and food security. Importantly, cash transfers helped are necessary to give beneficiaries the necessary support for households to mitigate the adverse effects of climatic shocks healthy child development. and protect consumption levels better than the control group. Several mechanisms contributed to enhance resilience among Productive inclusion measures integrated into the safety cash transfer beneficiaries. They were more likely to participate net system yielded cost-effective results and produced in savings groups and save more. They were also engaged in considerable impact on households’ income, food security, more diversified agricultural and non-agricultural activities, resilience and well-being compared to cash-only. The impact and had higher earnings from these activities during droughts, evaluation15 showed that productive inclusion measures helped which suggests income-smoothing. Savings and diversification to boost investments and diversify off-farm income-generating of income-generating activities reduced the need for adverse activities, which led to strong increases in revenues and profits. coping mechanisms. On the other hand, the evaluation found It also showed higher levels of total consumption and food limited impact on asset accumulation, with few differences in security among beneficiaries, who also experienced improved household durables and livestock. mental health and a greater sense of worth. The program created a safe space and fostered group dynamics and peer learning. The combination of cash transfers with human capital Beneficiaries experienced stronger social and financial support, accompanying measures provided added value in terms of higher trust, and greater capacity for collective action. Women improving practices aimed at promoting human development. reported greater aspiration and decision-making power over The impact evaluation14 showed a positive impact on women’s their own resources. The program improved gender attitudes knowledge and practices related to children’s health, nutrition, and perceptions of women engaging in economic activities. and development. Women were the cash transfer recipients 13 Premand P., Stoeffler Q. (2020) Do Cash Transfers Foster Resilience? Evidence from Rural Niger 14 Premand P., Barry O. (2020). Behavioral Change Promotion, Cash Transfers and Early Childhood Development Experimental Evidence from a Government Program in a Low-Income Setting 15 Briefing paper: “Productive inclusion measures in the Sahel. An integrated approach to promote the economic activities of poor and vulnerable households”. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 71 3.4. Conclusion The most important shocks Nigerien households face their food consumption, by selling livestock, or by doing are natural hazards, health related shocks, high food nothing at all. All these coping mechanisms are sub- prices, and insecurity. This chapter uses descriptive optimal and could potentially have long term negative statistics and econometric techniques to provide an impact on the households. In the case of a reduction of assessment of the extent to which various shocks affects food consumption for example, it has been demonstrated households in Niger. Poor rainfall has huge consequences in the literature that a reduction of food consumption will on poverty, especially, given that in Niger, most poor rely increase the level of food insecurity which in turn could on rainfed substance agriculture for livelihood. Similarly, affect early childhood development. recrudescence of conflicts in recent years have contributed to difficulties for households to conduct normal economic Going forward, the country should establish a nationwide activities, resulting in lower income. This chapters provide well-targeted and adaptative social protection system in strong evidence that drought and conflicts result in lower order to efficiently reduce the negative impact of shocks household consumption. on households. Evidence from the literature suggest the successful impact of such approach (Box 4). More recently, In a context were formal social assistance program are Thomas Bossuroy & Patrick Premand (2020) shows that very limited, households’ resorts to sub-optimal coping safety nets programs can be enhanced to further boost mechanisms which in turn can have long lasting effects, economic inclusion and resilience, and that productive including on early childhood development. When a shock inclusion programs integrated to safety nets have great occurred, most households cope by using their savings, potential to protect and promote livelihoods. by asking support from friends of family, by reducing 72 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 4 HUMAN CAPITAL NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 73 This chapter presents an updated description of human capital - education, health, and food security - in Niger. Building on the recently available survey, a mixed of descriptive and econometric analysis is used to provide an overview of the human capital landscape in Niger. 4.1. Achievements in human capital index The Human Capital Index (HCI), first introduced by the components include: mortality rate of children under 5 World Bank in 2018, is an aggregate measure of key years of age, the number of years of school attendance components of human capital in a country. It is defined by age 18, harmonized test scores, number of children as the expected future productivity, relative to the current age 15 who survive until age 60, and the rate of stunting health and education outcomes. Constructed around three for children under 5 years of age. The index calculates a components, namely survival, education and health, the single value between 0 and 1, based on how much each index applies five key measures that global research has indicator contributes to the productivity of an adult linked to productivity: Child Survival, School Enrollment, (World Bank, 2018). Quality of Learning, Healthy Growth and Adult Survival. The indicators applied to measure performance under these Table 11. Human Capital Index and its components across benchmark countries Indicator Burkina Faso Chad Mali Niger Sub-Saharan Africa Low Income HCI Component 1: Survival Probability of Survival to Age 5 0.924 0.881 0.902 0.916 0.934 0.928 HCI Component 2: School 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 Age 15 to 60 0.761 0.646 0.750 0.767 0.735 0.747 Fraction of Children Under 5 Not Stunted 0.751 0.602 0.731 0.515 0.688 0.654 Human Capital Index (HCI) 2020 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). 74 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION In 2020, Niger was ranked 155th out of 157 countries The spatial distribution of the HCI shows important in the global ranking of the HCI. 16 The HCI for Niger region disparities, with two regions who are leading estimated at 0.32, indicates that a typical child born today (Niamey and Agadez) and the rest remaining behind. can expect to attain only 32% of their productive potential Niamey, the region that includes the capital city of the as an adult (World Bank, 2020a). When compared to country, has the highest HCI estimated at 0.49 (Figure 69). neighboring countries in the Sahel, Niger performs equally Agadez then follows with an HCI of 0.39. The remaining poorly (Table 11). In the region, Niger and Mali tied at a regions are all have an HCI with values ranging between close range with a HCI averaging 0.32 greater than the HCI 0.29 and 0.32. Consequently, improvement of the country’s for Chad (0.30), but much lower than that of and Burkina performance will require closing the spatial gaps. Faso (0.38). Across these four benchmark countries, Niger ranks lowest in three indicators: Harmonized Test Scores, In Niger, poverty is associated with low human capital. Survival Rate from Age 15 to 60, and Fraction of Children As illustrated under Figure 69, there is a clear and strong Under 5 Not Stunted. correlation between poverty and human capital index. This is not surprising, as the HCI summarize the key Between 2018 and 2020, Niger experienced a decline in endowments that an individual will bring to the labor HCI, a clear move in the wrong direction. Due to limited market. A high HCI reflect better endowment, and more data availability, it is not possible to have a longer trend chance to access quality jobs and to be more productive. of the HCI in Niger. Estimations from the World Bank suggest that human capital index of the country shifted downward from 0.318 in 2018 to 0.316 in 2020 (Table 12). This reduction was experienced by both boys and girls. The poor performance was mainly driven by a deterioration of the nutrition status of children. Table 12. Human Capital Index (HCI) and its components for the year 2018 and the year 2020 Male Female Male+Female Indicator 2018 2020 2018 2020 2018 2020 HCI Component 1: Survival Probability of Survival to Age 5 0.910 0.913 0.917 0.920 0.913 0.916 HCI Component 2: School Expected Years of School 5.962 5.936 5.059 5.045 5.518 5.498 Harmonized Test Scores 307.091 307.091 302.089 302.089 304.922 304.922 HCI Component 3: Health Survival Rate from Age 15 to 60 0.745 0.752 0.773 0.781 0.759 0.767 Fraction of Children Under 5 Not Stunted 0.572 0.497 0.602 0.532 0.587 0.515 Human Capital Index (HCI) 2020 0.320 0.318 0.315 0.313 0.318 0.316 Uncertainty Interval [0.30,0.34] [0.29,0.34] [0.29,0.33] [0.29,0.33] [0.30,0.34] [0.29,0.33] Note: HCI 2018 back-calculated - Estimate 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). 16 The Central African Republic ranked last with a HCI of 0.29. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 75 Figure 69. Human capital and poverty Human capital across region Human capital and poverty 0.6 60.0 0.5 0.49 50.0 Dosso, Maradi, 0.3, 48.4 0.3, y = 1256.4x2 - 1193x + 289.6 46.1 0.3, Tillaberi, R² = 0.938 Poverty headcount (%) 0.4 0.39 40.0 Niger, 0.3, 40.8 42.3 Tahoua, 0.3, 0.32 0.32 35.7 0.3, 34.0 Diffa, 0.31 0.31 0.30 0.30 0.29 30.0 0.3 20.0 0.2 Agadez, 0.4, 14.6 10.0 0.1 Niamey, 0.5, 6.9 0.0 0.0 0.30 0.35 0.40 0.45 0.50 Human Capital Index Tahoua Niamey Agadez Diffa Niger Dosso Maradi Tillaberi Zinder Source: World Bank staff calculation based on EHCVM 2018/19. 76 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 4.2. Education Despite progress in recent years, Niger continues to Compared to its regional peers, investment in education experience both low schooling rate and high gender remains low. Niger has the second lowest expenditure on disparities in primary and secondary education. Table education as a percentage of GDP (3.5%) when compared 13 highlights the high dropout rate with 45% of students to neighboring countries (Table 13) and the lowest share enrolled in primary school not continuing with enrollment of government expenditure on education (13.2%). The at the secondary school level. The same picture is low government expenditures have a consequence on observed across the four neighboring countries with less the ability of the state to provide quality education. The than a third of the students attending secondary school as limited resources are not enough to cover the cost related compared to over half of the students attending primary to human resources, physical infrastructures, learning school. Niger ranks second lowest on net enrollment materials, scholarship, etc. Yet all these are critical factors rate for secondary school and percentage of secondary that affect satisfactory academic achievement level. and primary school students who are female. Beyond enrollment rates, Niger also faces a challenge of poor access to quality education with the lowest performance on the harmonized test score (Table 11). Table 13. Education investments and outcomes across benchmark countries Government Expenditure on Net enrollment Net enrollment % of primary % of primary expenditure on education as % rate, primary (%) rate, secondary students who are students who education as % of of government (%) female are female GDP expenditure 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. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 77 The probability of being enrolled in a school is higher for enrollment of 6 percentage points. The gap between boys boys compared to girls with a bigger gap at the tertiary and girls’ enrollment is much higher for the age group 18- related age group. The highest enrollment rates is 14 years old, 9 percentage points. In general, enrollment recorded among the 7-12 age group at 54% while the lowest rates decline once students turned 19 years old possibly were for 3-6 age group indicating that early childhood due to girls getting married and boys starting to shoulder education trails behind for both girls and boys (Figure the responsibilities of bread winner within the household. 70). Among the 13-15 age group, 57% of boys were enrolled However, girls seems to be dropping out much faster. in school compared to 51% of girls resulting in a gap in Figure 70. School enrollment rates by gender and age group 3-6 year 7-12 year 13-15 year 16-18 year 19-24 year 0.00 0.10 0.20 0.30 0.40 0.50 0.60 Percent Boys Girls Source: World Bank staff calculation using data from EHCVM 2018/19 78 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Households in the highest wealth quintiles are more wealth quintile, respectively. Early childhood education likely to send their children to school but early childhood is low across all wealth quintiles with the rates hovering education enrollments are low across all the wealth around 6%-19%. The gap in enrollment rates across quintiles. School enrollment rates are the highest among wealth quintiles is the lowest for ages 3-6 indicating that the 7-15 age group and they are more likely to be from challenges in accessing Early Childhood Education Center wealthier households (Figure 71). In fact, enrollment rates is more evenly distributed. However, enrollment rates for ages 7-12 are the highest among households in the drop significantly for the tertiary level age group (19-24 age highest wealth quintile at 74% in contrast to an enrollment group) and with the highest wealth quintile exhibiting the rate of 40% and 46% for the lowest and second lowest highest rate of 22%. Figure 71. School enrollment rates by age group and wealth quintile 0.80 0.70 0.60 Percentage of children enrolled in school (%) 0.50 0.40 0.30 0.20 0.10 0.00 3-6 years 7-12 years 13-15 years 16-18 years 19-24 years Age group Lowest Second Middle Fourth Highest Source: World Bank staff calculation using data from EHCVM 2018/19 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 79 There are important spatial differences in primary with 92 percent of children aged 7 to 12 who are attending school enrollment (children aged 7-12). Primary school school. The region of Agadez and Dosso also performed enrollment rates for urban children are much higher than well, with an enrollment rate of 64 percent each. On the among rural students with a difference of approximately other extreme, enrollment is much lower for Zinder and 41 percentage point between urban and rural areas. There Diffa, 44 and 39 percent respectively. This suggests that are also important differences across regions (Figure 72). there is need for policies to equalize access to primary The region containing the capital city of Niamey stands out, school education at the country level. Figure 72. Primary school-age (7–12 years) enrollment rate by region 1 0.9 0.8 Percentage of children enrolled in school (%) 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Agadez Diffa Dosso Maradi Tahoua Tillaberi Zinder Niamey Source: World Bank staff calculation using data from EHCVM 2018/19 80 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION For both girls and boys, family disavowal is by far the School enrollment is generally associated with a lower main reasons for never attending a formal school. Among participation in paid and unpaid work but there are boys aged 7 to 24 years old, 42% never attended school gendered differences. As they grow older, girls are because there their family didn’t want (Figure 73). The more likely to spend more time more time on domestic corresponding number for girls is slightly higher (42%). work compared to boys (Figure 74). This pattern is more Families appear to refuse to send girls to school at a pronounced for girls between the ages of 16 and 24 and slightly greater rate compared to boys. This is in addition not enrolled in school. On average, a girl between 19-24 to girls reporting not being sent to school because of their and out of school spent 34 hours a week taking care of gender (7%). Issues related to unavailability of school a child, collecting wood and water compared to 8 hours facility, and tradeoff between school and work are also for a boy in the same category. The analysis also suggests important constraints to schooling. The tradeoff between that there is a tradeoff, and that going to school slightly schooling and work, and issues related to the perceived reduces the number of hours girls spent on domestic importance of school are more pronounced for boys. The chores. However, their level of involvement in domestic fact that parents are refusing to send their children to activities remained far greater than that of their male school is likely due to a clear tradeoff between expected peers in school and not in school. return to school (child probability to graduate and to find a good job), and immediate benefit via child labor or early marriage. To affect parents’ decision in a positive manner, several factors must be aligned, including improved supply of quality education, job creation to ensure better transition from school to quality jobs. Figure 73. Distribution of reasons for having never attended a formal school (ages 7–24) Girls 7% 16% 46% 14% 1% 10% 7% Boys 15% 42% 14% 1% 15% 11% 0% 20% 40% 60% 80% 100% Being a girl No school/too far away Family refuse Paid/Unpaid work Too expensive/not enough money School not suitable/useful Others Source: World Bank staff calculation using data from EHCVM 2018/19 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 81 Figure 74. Hours of domestic work and work outside home in the past week, disaggregated by gender, age group, and school enrollment (mean values) Hours of domestic work Work outside home Boys Boys 10.0 70.0% 9.0 60.0% 8.0 7.0 50.0% Work ourside home (%) Hours domestic work 6.0 40.0% 5.0 4.0 30.0% 3.0 20.0% 2.0 1.0 10.0% 0 0.0% 3-6 7-12 13-15 16-18 19-24 Total 3-6 7-12 13-15 16-18 19-24 Total years years years years years years years years years years Enrolled in school in 2018/19? Enrolled in school in 2018/19? Yes No Yes No Hours of domestic work Work outside home Girls Girls 40.0 40.0% 35.0 35.0% 30.0 30.0% Work ourside home (%) Hours domestic work 25.0 25.0% 20.0 20.0% 15.0 15.0% 10.0 10.0% 5.0 5.0% 0.0 0.0% 3-6 7-12 13-15 16-18 19-24 Total 3-6 7-12 13-15 16-18 19-24 Total years years years years years years years years years years Enrolled in school in 2018/19? Enrolled in school in 2018/19? Yes No Yes No Source: World Bank staff calculation using data from EHCVM 2018/19 82 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION As they transition into adulthood, both boys and Deterioration of classrooms, and frequency of tuition girls incur greater responsibilities that translate into fees are by far the most important problems raised by increased participation in work outside of the home. parents. In fact, 38% of public-school parents reported Indeed, girls 19-24 who were not enrolled in school spent that their schools faced issues of deterioration of the the most time on domestic work (38 hours) and boys of classrooms (Figure 75). This problem is also important for the same age group and not in school (58%) were the community schools. The frequency of tuition fees is the most likely to report participating in income generating second most important issue raised by parents, especially activities including farm work, self-employment, and wage for public and community schools. Absenteeism of work (Figure 74). Across all the age groups, enrollment teacher is an important problem for community school, in school was associated with a lower involvement in and to some extent, public schools as well. Other issues work outside of the home. This potentially indicate that are equally important for public and community schools: the decision to keep children in school has significant lack of teachers, overcrowding, lack of supply. A couple financial implications, especially for families in need of of other issues are predominantly mentioned for public both the income and extra labor provided by their school schools only: poor education, lack of bathrooms, and lack aged children. of equipments. It is important to note that compared to public and community schools, private schools have less problems, and parents tend to be more satisfied by the services provided in these private schools. Figure 75. Problems reported at schools, disaggregated by school type (%) Deteriorated classroom Frequency of tuition fees Lack of bathroom Lack of equipment Lack of supplies Lack of teachers Overcrowded classrooms Poor education Teacher absence 0% 5% 10% 15% 20% 25% 30% 35% 40% Public Private non-religious Communautary and others Source: World Bank staff calculation using data from EHCVM 2018/19 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 83 Being a girl and residing in a rural area significantly school. Additionally, the wealth level of families is a key reduced the likelihood of school enrollment while determinant of the school enrollment as households in being of primary school age improves the probability the higher wealth quintiles are significantly more likely to of being in school. To estimate the correlates of school send their children to school. As one would expect, the enrollment we use a multivariate analysis (logistic model) proximity of a school either in the village or neighborhood presented in Figure 76. The results indicate that girls and increases the likelihood of enrollment. Surprisingly, even rural children are the least likely to be enrolled in school. when parents report many problems at schools it does In contrast, primary school age children (ages 7-12) have not negatively affect enrollment rates. This could be due a higher likelihood of being in school compared to any to the fact that families often don’t have many options for other age group. Having siblings in school, or one of the where to send their children for school. parents with at least a primary school education are also correlated with a higher likelihood of being enrolled in Figure 76. Correlates of the likelihood of being enrolled in school (logit marginal effects with 95% confidence intervals) Average Marginal Effects with 95% CIs 1. female 1. child_of_head 1. nage_category 2. nage_category 3. nage_category 1. father_primary education Effects with respect to 1. mother_primary education 1. other_children_in_school 2. wealth_quintile 3. wealth_quintile 4. wealth_quintile 5. wealth_quintile 1. proper_school_in_village 1. many_school_problems 1. rural -.2 0 .2 .4 Effects on the Pr (In_School_2018) Source: World Bank staff calculation using data from EHCVM 2018/19 Notes: Base group for age is 3–6 years old; ‘Father – primary school’ is an indicator for the individual’s father having attended at least some primary school; ‘Other children in school’ is an indicator for there being other household members enrolled in school; ‘School in village/neighborhood’ is an indicator for whether there is a school in the village that is appropriate for this individual, given their school attainment or age; ‘Many school problems’ is an indicator for whether local households report an above-median number of problems associated with the schools. This regression is limited to ages 3-18 because the survey captured local school presence (an explanatory variable) only up to high school. 84 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Across age groups, men exhibited higher literacy rates all age groups, the literacy for male is much higher for compared to women. Consistent with increased enrollment men compared to women. The gap varies between 19 and rates in recent years, literacy rate is much higher for the 29 percentage points. The gap seems to have reduced for younger cohorts. For instance, the literacy rate among the the youngest cohorts. However, it remains high, and effort 15-25 age group is 46.5 percent, compared to only 16.9% should be made to increase enrollment and survival for for those aged 65 and above (Figure 77). In addition, for women, in order to reduce this gender gap. Figure 77. Literacy rates by gender and age group 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% 15-25 years 25-35 35-45 45-55 55-65 >65 years Male Female Source: World Bank staff calculation using data from EHCVM 2018/19 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 85 4.3. Health The high rate of early marriage is a potential driver Under five mortality is higher among the poorest and of high rate of maternal mortality. The proportion of vaccination coverage increases with welfare quintile. As deliveries assisted by a skilled birth attendant in Niger illustrated in figure 79, Under-five mortality is much lower was estimated at 32.6% in 2012 (DHS). A catalyst for the in urban areas (83) compared to rural areas (163). A total high rate of maternal mortality is possibly the high rate of 55 percent of children aged 12 to 23 months are fully of early marriages particularly in rural regions (Figure vaccinated. There are some disparities across location 78). This is because young adolescents face a higher and welfare. Rural areas have higher rates of children risk of complications during pregnancy and delivery stunted and lower vaccination rates compared urban and therefore high probability of maternal mortality as ones. There is also a clear and strong correlation between compared to older women. In Niger, the median age at welfare quintiles and health outcomes, with children from first marriage for women aged 20 to 49 is estimated at wealthier households being in a better position. 15.8. Given the low adoption of contraceptive methods, the probability of most young married women to have a baby before they turn 18 years old is very high. The median age at first marriage is much higher for those living in urban areas (18.1) and those in the wealthiest quintile (17.5). Figure 78. Maternal health across subpopulations Median age of women at first marriage Married women using any method of Total fertility rate (aged 15-49) (ages 20-49) contraception (15-49) 18.5 35 9 18 30 8 17.5 7 25 17 6 16.5 20 5 16 15 4 15.5 3 10 15 2 14.5 5 1 14 0 0 Highest Highest Highest Total Lowest Total Lowest Total Lowest Urban Second Middle Fourth Urban Second Middle Fourth Urban Second Middle Fourth Rural Rural Rural Residence Wealth quintile Residence Wealth quintile Residence Wealth quintile Source: DHS 2012 86 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Figure 79. Child health across subpopulations Under-five mortality rate (‰) Children stunted Received all eight basic vaccinations 180 60.0 80.0 160 70.0 50.0 140 60.0 120 40.0 50.0 100 30.0 40.0 80 60 20.0 30.0 40 20.0 10.0 20 10.0 0 0.0 0.0 Highest Highest Highest Total Lowest Total Lowest Total Lowest Urban Second Middle Fourth Urban Second Middle Fourth Urban Second Middle Fourth Rural Rural Rural Residence Wealth quintile Residence Wealth quintile Residence Wealth quintile Source: DHS 2012 Morbidity follows life cycle, with children and elderly likely to fall sick, and to visit health facilities. But overall, more likely to fall sick. Clearly, there is a U shape babies and children are more likely to seek health care. relationship between age and the probability of falling This is a sign that auto-medication may be at plays when sick (Figure 80). There are no major differences in men its comes to adults and older people, preferring to visit and women at both extreme ends of the age distribution, a health center only when the situation is very serious however, for the working age population, women are more (Figure 81). Figure 80. Experiences of illness and rates of visiting a health facility, by gender and age group Experienced illness in the past 30 days (%) Visited health facility (among those with illness) (%) 60.0 80.0 70.0 50.0 60.0 40.0 50.0 30.0 40.0 30.0 20.0 20.0 10.0 10.0 0.0 0.0 0-5 6-10 10-15 15-25 25-35 35-45 45-55 55-65 >65 0-5 6-10 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 EHCVM 2018/19 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 87 Figure 81. Reasons for not visiting a health facility when ill, by wealth quintile Lowest 15.3 59.6 20.5 4.6 Second 16.2 63.9 15.4 4.5 Middle 22.2 58.8 15.1 3.9 Fourth 22.6 61.7 11.6 4.2 Highest 19.1 70.4 7.4 3.1 Niger 19.2 62.8 14.1 4.0 0% 20% 40% 60% 80% 100% Unncessary Self-medication Lack of financial measure Other Source: World Bank staff calculation using data from EHCVM 2018/19 While accessing health services, there are several issues waiting times, the lack of medicine, and the high cost. The that household have to dealt with, the most important largest disparity in number of complaints between private one being long waiting time, the lack of medicine, and the and public health facilities is for long waiting times and high cost. Figure 82 illustrates that public health facilities lack of medicine (with a difference of 18 and 17 percentage are perceived overall to have more problems than private points respectively). When it comes to cost related issues, ones. The problems patients noted most when visiting the difference between public and private is minor. both public and private health facility visits were long 88 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Figure 82. Problems reported at public and private health facilities Facility not clean Long waiting time Staff not qualified Too expensive No medicine Inefficient treatment Bad reception Lack of staff Other problem 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% Public Private All Source: World Bank staff calculation using data from EHCVM 2018/19 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 89 4.4. Food security Chronic food insecurity is prevalent in Niger. Due to Comparison of food security indicators with neighboring data limitations, the country is not ranked in the Global countries highlighted Niger’ poor performance. Compared Hunger Index. However, estimations using the 2020 EHCVM to its peers, Niger is performing poorly in relation to suggest that close to two in five households (38.4%) are average dietary energy supply adequacy, share of dietary in a situation of severe food insecurity (Figure 83). An energy supply derived from cereals, roots and tubers, and additional one in five households (21.4%) does experience access to basic drinking water serves (Table 14). moderate food insecurity. Overall, three in five households is experiencing a form of food insecurity. Food insecurity is more pronounced for rural households. In rural areas, 64.6 percent of households are experiencing a form of food insecurity, against 36.5 percent in urban areas. Figure 83. Distribution of food security categories Urban 46.6 16.9 15.5 21.0 Rural 20.4 15.0 22.6 42.0 Total 24.9 15.3 21.4 38.4 0% 20% 40% 60% 80% 100% Food secure Mildly food secure Moderately food secure Severely food secure Source: World Bank staff calculation using data from EHCVM 2018/19 90 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Table 14. Food security indicators across benchmark countries Average dietary Share of dietary Gross domestic Per capita food People using at least energy supply energy supply derived product per capita supply variability basic drinking adequacy from cereals, roots (in purchasing power water serves (%) and tubers equivalent) Burkina Faso 122 64 2,190 30 48 Chad 95 62 1,580 31 39 Mali 135 68 2,327 36 78 Niger 121 62 1,219 40 50 Source: FAO (2020) Note: These numbers generally refer to values for years between 2017 and 2019, with the share of dietary energy supply derived from cereals, roots and tubers reflecting the period 2015-17. Per capita food supply variability is the standard deviation of per capita food supply (in dietary energy) over the previous five years. Household food security is generally better off in Niamey Increasing levels of wealth are associated with greater and Agadez regions. The Niamey region, which includes food security while rural households, female headed the capital city, is the region with the highest food security ones, and those whose head is engaged in agriculture scores. At the opposite end of the distribution Maradi are the most likely to be food insecure. In effect results has the lowest score. Regions with highest monetary of a multivariate analysis (logistic model) support that poverty rates tend to also be the one with the lowest food wealthier households were the least likely to be food security. This point to affordability of food as being one of insecure as were households with access to asphalt road the possible causes of malnutrition. Of course, issues of (Figure 85). In addition, households headed by individuals availability of the food in the market could also be at play: engaged in the agriculture sector or holding a secondary the rarer, the higher the price. education degree had a higher likelihood of being food secure. Female headed households have a lower likelihood of being food secure. Residing in a rural area significantly increases the precarity of households’ food security. Figure 84. Average household food security across regions (0-8 scale, 0 = insecure and 8 = secure) 7.00 6.43 6.00 5.49 4.97 5.00 4.55 4.32 4.36 4.44 4.18 4.00 3.92 3.00 2.00 1.00 0.00 Agadez Diffa Dosso Maradi Tahoua Tillaberi Zinder Niamey Niger Source: World Bank staff calculation from EHCVM 2014 and EHCVM 2018/19. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 91 Figure 85. Correlates of being food secure (logit marginal effects with 95% confidence intervals) Average Marginal Effects with 95% CIs 2. wealth_quintile 3. wealth_quintile 4. wealth_quintile 5. wealth_quintile 1. fhh Effects with respect to dependency ratio 1. head_agriculture 1. primary 1. secondary 1. hhandig ln_population 1. market_permanent 1. road_asphalt 1. rural -.2 -.1 0 .1 .2 Effects on the Pr (Food_Secure) Source: World Bank staff calculation using data from EHCVM 2018/19 Notes: ‘Dependency ratio’ is the proportion of household members below 15 years or above 64 years old; ‘Head – agriculture’ is an indicator for the household head’s main occupation being crop or livestock agriculture; ‘Head – primary school’ and ‘Head – secondary school’ are indicators for the household head’s completion of some primary and secondary school; ‘Head – handicap’ is an indicator that the household head has a physical or mental handicap; ‘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. 92 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 4.5. Conclusion Performance on human capital is very low. To be As illustrated in the literature, investment in early successful, the fight against poverty should put a childhood development will have to be part of the strong emphasis in building human capital. This chapter package of investment designed to build human capital provides a diagnostic of the human capital index in Niger, in the country. The benefits associated with intervening with a focus on each of its sub-component: education, early on have being illustrated (Wodon 2017), with the right health, and food security. The HCI for Burkina Faso is policies early in the lives of children, beginning with the estimated at 0.32, which indicates that a typical child born provision of quality pre-natal services, and an emphasis today can expect to attain only 32% of their productive on early childhood development. It is demonstrated potential as an adult. This is lower when compared to the that early childhood development improves academic Sub-Saharan Africa average (40%). The poor performance performance, and developed other social skills (such as on education on health is driven by both supply and socialization, enthusiasm for lifelong learning, teamwork, demand issues. For both girls and boys, family disavowal resilience among others) that will prove critical in is by far the main reasons for never attending a formal. the society. Deterioration of classrooms, and frequency of tuition fees are by far the most important problems raised by parents. For those who are already in the active population, policy When it comes to health service, evidence suggest that makers should consider the expansion of programs such auto-medication may be at plays when it comes to adults as the Sahel Women’s Empowerment and Demographic and older people, they would opt to visit a health center Dividend (SWEDD). SWEDD gives women and girls across only when the situation is very serious. Households West Africa vocational training so that they can earn more mentioned long waiting time, lack of medicine, and high and have a better future. The project also strengthens girls’ cost as key constrains in assessing health services. Due to education, increases access to reproductive health services, poor productivity of agriculture, chronic food insecurity is and engages whole communities on issues including child prevalent in Niger, and the Covid-19 seems to have worsen marriage, human rights and gender equality. Given that the situation. If well designed, the fight against poverty, poverty seems to affect men and women equally, variants and the needed improvement of human capital will benefit of such program could be considered so as to benefit boys from a nationwide pro-poor targeted social assistance and men as well. program, that will help improve access to education and health service, and also boost agricultural productivity. But this should be coupled with improvement of the quantity and quality of services in these areas, including advisory services to farmers, improvement of quantity and quality of education and health services. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 93 REFERENCES Aguiar, A., Narayanan, B., McDougall, R., 2016. “An Overview Bernard, T., Hidrobo, M., Le Port, A., & Rawat, R. (2018). of the GTAP 9 Data Base.” Journal of Global Economic “Nutrition-based incentives in dairy contract farming Analysis 1, no. 1 (June 3,2016): 181-208. https://jgea.org/ in Northern Senegal.” American Journal of Agricultural resources/jgea/ojs/index.php/jgea/article/view/23 Economics, 101(2), 404-435. Alderman, H., Hoddinott, J., & Kinsey, B. (2006). Long term Cornelsen, L., Green, R., Turner, R., Dangour, A. D., Shankar, consequences of early childhood malnutrition. Oxford B., Mazzocchi, M., & Smith, R. D. (2015). What happens to Economic Papers, 58, 450–474. patterns of food consumption when food prices change? Evidence from a systematic review and meta‐analysis of Alem, Y., & Söderbom, M. (2012). Household-level food price elasticities gglobally. Health economics, 24(12), consumption in urban Ethiopia: the effects of a large food 1548-1559. price shock. . World development, 40(1), 146-162. De Brauw, A., Mueller, V., & Lee, H. L. (2014). “The Role of Andreyeva, T., Long, M. W., & Brownell, K. D. (2010). The Rural-Urban Migration in the Structural Transformation of impact of food prices on consumption: a systematic Sub-Saharan Africa.” World Development 63, 33-42 review of research on the price elasticity of demand for food. . American journal of public health, , 100(2), 216-222. Decker et al. 2020. COVID-19: A pivotal moment to support women farmers, https://blogs.worldbank.org/ Arouna, A., Michler, J. D., & Lokossou, J. C. (2019). “Contract developmenttalk/covid-19-pivotal-moment-support- farming and rural transformation: Evidence from a field women-farmers experiment in Benin.” Working paper no. w25665. National Bureau of Economic Research. Del Ninno, C., & Lundberg, M. (2005). Treading water: The long-term impact of the 1998 flood on nutrition in Atake, E. H. (2018). Health shocks in Sub-Saharan Africa: Bangladesh. Economics & Human Biology, 3, 67–96. are the poor and uninsured households more vulnerable? Health economics review, 8(1), 26. Dell, M., Jones, B., & Olken, B. (2014). What do we learn from the weather? The new climate-economy literature. Baez, J., de la Fuente, A., & Santos, I. (2010). Do natural Journal of Economic Literature, 52(3), 740–798. disasters affect human capital? An assessment based on existing empirical evidence. (Discussion Paper No. 5164). Devereux, S. (2007). The impact of droughts and floods Bonn, Germany: IZA. on food security and policy options to alleviate negative effects. Agricultural Economics, 37, 47-58. Bartram, J., & Cairncross, S. (2010). Hygiene, sanitation, and water: forgotten foundations of health. PLoS medicine, 7(11). 94 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Enahoro, D., Njiru, N., Thornton, P., & Staal, S. S. (2019). A Jamison, E. A., Jamison, D. T., & Hanushek, E. A. (2007). review of projections of demand and supply of livestock- The effects of education quality on income growth and derived foods and the implications for livestock sector mortality decline. Economics of Education Review, 26(6), management in LSIL focus countries. 771-788. Fafchamps, M. (1992). “Cash Crop Production, Food Price Kazianga, H., & Udry, C. (2006). Consumption smoothing? Volatility and Rural Market Integration in the Third World.” Livestock, insurance and drought in rural Niger . Journal American Journal of Agricultural Economics, 74(1): 90-99 of Development Economics, 79(2), 413–446. Gertler, P., & Gruber, J. (2002). Insuring consumption Larson, D. F. (2016). On the determinants of low against illness . American Economic Review. productivity in maize farming in Uganda: The role of markets, fertilizer use and gender. Tokyo: Springer. Giugale, Marcelo and Nga Thi Viet Nguyen (2014). “Money to the People: Estimates of the Potential Scale of Lindelow, M., & Wagstaff, A. (2005). Health shocks in China: Direct Dividend Payments in Africa”. Center for Global are the poor and uninsured less protected? World Bank Development. Policy Research Working Paper, 3740. Gyimah-Brempong, K., Paddison, O., & Mitiku, W. (2006). Mahajan, K. (2019). Back to the plough: Women managers Higher education and economic growth in Africa. . The and farm productivity in India. World Development, 124, Journal of Development Studies, 42(3), 509-529. 104633. Hess, G. D. (2003). The economic welfare cost of conflict: Mara, D., Lane, J., Scott, B., & Trouba, D. (2010). Sanitation an empirical assessment. and health. . PLoS medicine, 7(11). Hill, R., Nikoloski, Z., & Tao, T. (2019). Assessing sources of Meghir, C., Mobarak, A. M., Mommaerts, C. D., & Morten, M. vulnerability to poverty in Lesotho . (2019). “Migration and Informal Insurance.” NBER Working Paper No. 26082. https://www.ilo.org/wcmsp5/groups/public/@ dgreports/@dcomm/documents/briefingnote Murray, U., Gebremedhin, Z., Brychkova, G., & Spillane, C. (2016). Smallholder farmers and climate smart Ibáñez, A. M., & Vélez, C. E. (2008). Civil conflict and forced agriculture: Technology and labor-productivity constraints migration: The micro determinants and welfare losses amongst women smallholders in Malawi. . Gender, of displacement in Colombia. World Development, 36(4), Technology and Development, 20(2), 117-148. 659-676. Naab, J. B., Mahama, G. Y., Yahaya, I., & Prasad, P. V. (2017). INSD, 2019. Annuaire Statistique 2018, Niger Institut Conservation agriculture improves soil quality, crop yield, national de la statistique et de la démographie and incomes of smallholder farmers in North Western Ghana., . Frontiers in plant science, 8, 996. International Labor Organization. (2020). ILO Monitor: Covid-19 and the World of Work. Third edition. Korkoyah Ojogho, O. (2010). Determinants of food insecurity among and Wreh 2015. https://www-cdn.oxfam.org/s3fs-public/ arable farmers in Edo State, Nigeria. Agricultural Journal, file_attachments/rr-ebola-impact-women-men-liberia- 5(3), 151-156. 010715-en.pdf NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 95 Organization for Economic Cooperation and Development Wagstaff, A. (2007). The economic consequences of (OECD). 2008. “Livestock and regional market in the Sahel health shocks: evidence from Vietnam. Journal of Health and West Africa Potentials and challenges.” Economics , 26(1): 82-100. Pica-Ciamarra, U., Tasciotti, L., Otte, J., & Zezza, A. (2011). Watts, M. (1987). Drought, environment and food Livestock assets, livestock income and rural households: security: some reflections on peasants, pastoralists and cross-country evidence from household surveys. commoditization in dryland West Africa. In Drought and hunger in Africa (pp. 171-211). Ragasa, C., & Chapoto, A. (2017). Moving in the right direction? The role of price subsidies in fertilizer use and Woldemicael, G. (2010). Do women with higher autonomy maize productivity in Ghana. Food security, 9(2), 329-353. seek more maternal health care? Evidence from Eritrea and Ethiopia. . Health care for women international, 31(7), Robinson, S., Cattaneo, A., El-Said, M., 2000. Updating and 599-620. estimating a Social Account Matrix using cross entropy methods, IFPRI TMD Discussion Papers No.58, IFPRI: World Bank (2020), Macro-Poverty Outlook, Spring 2020 Washington DC. Edition. World Bank: Washington DC. Somé, J. W., & Jones, A. D. (2018). “The influence of crop World Bank (2020). “Rural Income Diagnostic in Niger”. production and socioeconomic factors on seasonal household dietary diversity in Niger.” PLoS ONE 13(5): Wouterse, F. (2012). “Migration and Rural Welfare: The e0195685. https://doi.org/ 10.1371/journal.pone.0195685 impact of Potential Policy Reforms in Europe.“ World Development 40(12), 2427-2439 Theriault, V., Smale, M., & Haider, H. (2018). Economic incentives to use fertilizer on maize under differing Zeufack, Albert G.; Calderon, Cesar; Kambou, Gerard; agro-ecological conditions in Niger. . Food Security, 10(5), Djiofack, Calvin Z.; Kubota, Megumi; Korman, Vijdan; Cantu 1263-1277. Canales, Catalina. 2020. Africa’s Pulse, No. 21, Spring 2020 : An Analysis of Issues Shaping Africa’s Economic Future. UNECA (2019). Niger ASM country profile. Available here World Bank, Washington, DC. © World Bank. https:// https://knowledge.uneca.org/ASM/Niger openknowledge.worldbank.org/handle/10986/33541 License: CC BY 3.0 IGO. United Nations Food and Agriculture Organization (FAO) (2019). AQUASTAT 2019. Rome: FAO. USAID 2020. https://www.usaid.gov/niger/food-assistance van der Mensbrugghe, D., (2019), MANAGE Model Documentation, unpublished model documentation, available upon request van Ittersum, M. K. (2016). Can sub-Saharan Africa feed itself? Proceedings of the National Academy of Sciences of the United States of America , vol. 113,52: 14964-14969. 96 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION ANNEX 1: TECHNICAL NOTE ON POVERTY MEASUREMENT BASED ON EHCVM 2018/19 DATA The main objective of this Harmonized Living Conditions Poverty measurement is the process of generating poverty Household Survey (EHCVM) is to build capacity in the indicators from survey data. Poverty measurement design, implementation, processing and analysis of involves three steps: (i) constructing an indicator for survey data for poverty assessment. EHCVM 2018/10 data measuring welfare; (ii) constructing a poverty line; and (iii) were collected in two waves, the first from September to aggregating the data to produce poverty indicators. This December 2018, the second from April to July 2019. Each note explains the methodological choices made for the wave collected half the sample. The two-wave approach measurement of poverty. The first section explains the was chosen in order to take into account of the seasonality approach used to construct the consumption aggregate of consumption (both in terms of habits and levels since the welfare indicator used is normalized annual of consumption). per capita household consumption. The second section explains the methodological approach used to construct the poverty line. The third section analyzes the transition from the consumption aggregate to the welfare indicator by applying different deflators. Lastly, initial results are briefly presented in the fourth section. Figure A1.1 summarizes the components of poverty measurement. • Food purchases Food • Food from home production consumption • Food received from others • Food consumed away from home • Household expenses Non-food • Personal expenses consumption • Health expenses Consumption • Education expenses aggregates • Purchase value Value of usage of • Replacement value durable assets • Depreciation rate • Interest rate Poverty • Lodging characteristics measures 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 living across space Spatial deflator NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 97 1. CONSUMPTION AGGREGATE Moreover, since consumption is reported in non-standard units (NSU) during data collection, we must also find a way of converting these NSUs into standard units (SU) The consumption aggregate represents annual household before applying prices. Obviously, if the data on NSUs is of consumption. It is calculated by aggregating food average or poor quality, this also affects the quality of the consumption, non-food consumption in non-durable consumption data obtained after valuation. goods and services, the use value of durable goods, and the imputed rent of owner-occupied and rent-free households. Three scenarios have been tested to value food consumption within the household Food consumption is measured over the last seven days (the reference period) preceding the enumerator’s Scenario A: Unit values combined with market prices. visit. It is the sum of household food consumption at For a given product, when a household purchased the home (purchases made and actually consumed, self- product during the last 30 days prior to the enumerator’s consumption of the household’s own production, and visit, the valuation of consumption (including auto- gifts received and actually consumed) and meals taken consumption and, if applicable, gifts) is calculated using away from home. In this survey, food consumption in the the unit value of acquisition. In this case, there is generally household is measured in quantities and meals taken no need to convert quantities into SUs provided the outside the household are reported as values. Food unit of acquisition and the unit of consumption are the consumption within the household is annualized by same, which happened here in two thirds of cases. If the multiplying the quantities consumed by 365/7. units are different, conversion to SUs is necessary before consumption can be valued.17 For households that did not purchase the product within the last 30 days prior to The trickiest question therefore concerns the valuation data collection, valuation is based on market prices. This of household food consumption (purchases, auto- is done sequentially. Consumption is first valued using consumption and gifts). The survey is designed to the average price calculated by regional and residential use two price vectors: the unit values of the products area; if the price is available at this geographical level for purchased, and the prices recorded in the markets of this product, the calculation ends. If the information is the localities where the sampled households live. The missing at the previous level, the average price calculated first set of information (unit values) is available if the at the level of the Agro-Ecological Zone (AEZ) and place of product was purchased by the household within 30 days residence is used.18 If information is still missing at this prior to data collection. When a product was purchased level, the price calculated by place of residence (urban during this period, in addition to providing information or rural) at the national level is used. If the information on consumption broken down into purchases, auto- is missing at the previous level, the average price for the consumption, and gifts, the questionnaire also provides region is used followed by the average price for the AEZ, information on the last purchase (quantity purchased and and finally the national price. It is important to note that corresponding value), which makes it possible to derive all prices are calculated by wave. In other words, Wave 1 the unit value of acquisition. If the product was purchased prices are not used to value the quantities of a household more than 30 days before the interviewer’s visit, the value surveyed in Wave 2 and vice-versa. of the purchase is not provided and therefore no unit value can be obtained. Obviously if consumption of the product within the household comes exclusively from auto-consumption and gifts, no unit value is available either. 17 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 observa- tions, only the unit of consumption is available as there have been no acquisitions in the last 30 days. 18 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. 98 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Scenario B: Market prices only. In this case, consumption Table A1.1 shows the breakdown of food consumption of a given product is first valued using the average price according to the different scenarios outlined above. The calculated for the region/residential area; if the price is results show that valuation by unit values alone (Scenario available at this geographical level for this product, the C) yields food consumption levels close to the case where calculation is ends. If the information is missing at the quantities are valued by a combination of unit value and previous level, the average price calculated at the level market price (Scenario A, adopted in the Ouagadougou of the Agro-Ecological Zone (AEZ)/ residential area is workshop conducted in January/February 2020). The used. If information is still missing at this level, the price difference between scenarios A and C in terms of median calculated by place of residence (urban or rural) at the distribution is 2 percent in Niger. The difference between national level is used. If the information is missing at the Scenarios B and C is extremely high at 94 percent. previous level, the average price for the region is used followed by the average price of the AEZ, and finally the From these results, it emerges that there is a choice national price. As before, it should be noted that all prices between Scenarios A or C on the one hand and Scenario are calculated by wave. It is also important to note that if B on the other. A good indicator in making the decision the choice is made to value quantities by market price, the is the share of food consumption in total consumption, conversion factors from NSUs to SUs must always be used. depending on which scenario is chosen. The share of food consumption, including meals taken away from home, was Scenario C: Unit values only. This scenario consists of 67% in Niger according to the EHCVM 2014 survey. These using unit values exclusively instead of prices. For a given statistics are compatible with Scenarios A and C, where product, when a household has acquired the product by the share of food consumption (including meals taken purchase in the last 30 days before the survey agent’s visit, away from home) is XXX percent for Scenario A and YYY the valuation of consumption (including auto-consumption percent for Scenario C. However, if Scenario B is used, the and, if applicable, gifts) is calculated using the unit value share of food consumption is abnormally high. of acquisition, as in Scenario A. For households that have not purchased the product in the last 30 days prior to On the basis of these two assessment criteria, it follows collection, valuation is calculated using the unit values of that Scenario B should not be selected. Meanwhile, further households that have purchased the product. A vector of work is being done using Scenarios A and C. Since the unit values is constructed using all possible combinations quantities valued are the same for all three scenarios, of product and unit of acquisition. The sequence of the the weakness of Scenario B lies in the quality of the price quantity valuation process is the same as in the previous data, which may not take sufficient account of differences scenario, the only difference being that market prices are in product quality. The other possible difficulty with replaced by unit values. As information on the product or Scenario B is the use of NSUs, another data source that unit combination is used, conversion factors for converting necessarily introduces noise. This will play a role in the NSUs to SUs are not needed. choice between Scenarios A and C. Since the latter has the advantage of not using NSUs, this is the scenario chosen here. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 99 Table A1.1 Breakdown of per capita food consumption according to various price assumptions for valuing declared quantities Consumption distribution Scenarios National average 1st quartile 2nd quartile 3rd quartile A 133,448 70,668 103,126 157,432 B 254,040 86,433 137,646 251,765 C 130,624 69,098 100,908 153,150 Meals taken away from home. Consumption of meals are excluded from the calculation of food consumption taken away from home is given in terms of value for and will be calculated by use value instead. Expenditure the last seven days before the interviewer’s visit to the on festivities and ceremonies as well as pilgrimages, household. It is provided for each individual (for meals which are considered exceptional expenditure, are also taken individually) and for the household as a whole for excluded. The only case in which holiday expenditure meals taken collectively by several household’s members. is used is for expenditure on clothing and footwear for The total value declared by the household is annualized religious holidays such as Christmas, New Year, Easter, end by multiplying it by 365/7. of Ramadan, Tabaski, etc. The reason for this choice lies in the fact that the clothing acquired during these holidays At the end of the above valuation process, consumption is real household consumption and not prestige or taken within the household is added to the meals taken conspicuous expenditure, which amounts to a transfer to away from home to give the total food consumption of other households. It is also important to stress the classic the household. debate over whether expenditure on education (school fees, costs of supplies, etc.) and health (consultations, Non-food consumption. Non-food consumption of non- medical examinations, medication, hospitalization) durable goods and services (including education and constitutes investment in human capital or consumption. health) is measured in value terms over a reference The choice was made to include them, as has long been period of 7 days, 30 days, 3 months, 6 months, and 12 the practice in WAEMU member countries. Nevertheless, months depending on the anticipated frequency of expenditure on therapeutic medical devices (crutches, consumption of each type of good. The value reported wheelchairs, dentures, prescription glasses, etc.) was during the reference period is multiplied by a factor excluded from the consumption aggregate. Even if these taking into account the frequency, or 365/7, 12, 4, 2, items were to be included, they would be treated like and 1, respectively. The important point is to define durable goods. durable goods as well as exceptional expenditure in order to exclude them in the aggregation of non-food Use value of durable goods. Durable goods are those consumption. Durable goods are defined as means of that render services to the household over a long period transportation (car, motorcycle, bicycle, etc.), household of time after their acquisition. For these goods, the use appliances (television, refrigerator, freezer, oven, washing that is made of them is considered consumption by machine, dishwasher, air conditioner, music system, radio, the household. It is therefore necessary to estimate fans, etc.), large pieces of furniture (sofa and armchair set, this consumption, which is called “use value.” All goods dining table and chairs, bookcase, other cupboards, etc.), regarded as durable goods have been defined above. In and electronic appliances and other goods (computer, addition, real estate (land, buildings) and goods mainly telephone, mobile phone, cameras, musical instruments intended for economic production (dugout canoes and such as guitar or piano, motorized gardening equipment, outboards, hunting rifles, etc.) were ignored. valuable jewelry and watches, carpets, etc.). These goods 100 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION For goods regarded as durable, the use value is a function Imputed rent of owner-occupied households. The final of the acquisition value, the age of the goods, the inflation component of the consumption aggregate is the imputed rate, the real interest rate, and economic depreciation. The rent of owner-occupied and rent-free households. For acquisition value and the age of the good were provided households, housing is an investment good; when a during the survey, an annual inflation rate of 1% and a real household has built a dwelling, it consumes it by occupying interest rate of 2% were used for all durable goods, and it. The general approach for estimating imputed rent is the the only unknown parameter was the depreciation rate. econometric approach. In some cases, where the number For each good and each household owning it, if vrempla of observations was too small, an alternative approach is the value of the asset replacement cost, vacqui is the was chosen, which is explained below. acquisition value, and age the age of the asset in whole years, the formula for calculating the depreciation rate The econometric approach is based on the following (depret) is as follows: 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- occupied and rent-free households. The explained variable of the model is the logarithm of the rent, the explanatory variables typically being: type of dwelling, number of The median depreciation rate (mdpret) of the asset for rooms, type of walls, type of roof, type of floor, type of all households is then calculated. Finally, if s12q03 is the toilet, presence of electricity in the dwelling, presence of number of goods of a given type owned by the household running water in the dwelling, mode of garbage disposal, and s12q08 is the acquisition price of such goods, the use mode of sewage disposal, and other community variables value of a given good (depan) is obtained by applying the such as the existence of a paved road in the locality, the following formula: most common mode of transportation in the locality, etc. The model is estimated using the stepwise procedure, which consists of gradually introducing the variables into the model and retaining only those that are significant. To account for differences in the housing market, the model is estimated separately for the country’s capital city, The sum of this variable (depan) for all assets owned by a other urban areas, and rural areas. In the capital and other household provides the aggregate of the use value of the urban areas, the econometric approach is systematically household’s durable goods. implemented. In rural areas, the housing market is tight, as shown by the low number of renter households in the It is important to note that adjustments are made to the samples. There are only 43 households with rental value data before calculation: (i) for goods less than one year in rural Niger. Thus, the econometric approach cannot old, age was assumed to be 0.5; (ii) for goods older than produce satisfactory results in rural areas. An alternative 20 years (less than 3% of observations), age was limited to approach is used. This consists of calculating the median 20 years; (iii) when the number of goods was not reported rent of tenants according to the number of rooms, and this and other information was present, the number of rent is imputed to owner-occupied households occupying observations of the good was imputed by the mode value; a dwelling with the same number of rooms. Here, given the and (iv) the acquisition value of outliers was adjusted small number of tenant households, the number of rooms before proceeding with the calculations (see adjustment variable is recoded into three modalities, for example (1 of outliers below). room, 2 rooms, and 3 or more rooms) NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 101 Table A1.2. Comparison of actual rent and imputed rent Actual rent Imputed rent # Obs Min Median Mean Max # Obs Min Median Mean Max Capital 278 18,000 300,000 386,171 1,697,054 418 44,394 354,384 442,114 2,169,331 Other urban areas 185 18,000 180,000 185,503 1,440,000 696 44,614 140,362 202,460 2,283,217 Rural areas 27 24,000 120,000 154,049 360,000 4,420 60,000 127,278 150,000 150,000 A type of validation was conducted by comparing the Outlier adjustment. It is always difficult to distinguish distributions of actual and imputed rents, and the results between what is an outlier (abnormally high or abnormally are shown in Tables A2, which cover the country’s capital low value) and what is simply an atypical value. Improper city, other urban areas, and rural areas, respectively. The outlier adjustment can reduce real inequalities in the comparison between actual and imputed rents shows population. For this reason, adjustments should be made quite significant differences in the capital. At median level, with caution. Here, adjustments were made in two stages: the gap is 42% in Niger. However, this difference does not abnormally low values, and abnormally high values. Values always suggest that the adjustment is bad. Indeed, if, that are too low are defined as zero food or zero non-food for example, the characteristics of owners’ houses are consumption. Household consumption was calculated better than those of renters, the difference is justified. according to four main consumption functions: (i) food A simple test of averages was carried out in the case of consumption, including meals taken away from home; Niger, and it appears that in the capital (Ouagadougou), (ii) non-food consumption without use value of durable houses occupied by owner-occupied households are on goods and imputed rent; (iii) use value of durable goods; average of better quality than those occupied by renters. and (iv) imputed rent. Households with a zero i or a zero Differences can also be justified by the fact that the fit of ii component were removed from the databases. The logic the model is not always very good; for example, surveys do is simple: it is unlikely that a household has zero food not take into account the quality of the materials used to consumption; rather, this household did not complete construct dwellings, and disparities in quality are greater the interview (in cases of a one-person household, in capital cities. Imputed rent estimates are of better where the householder is often absent) or refused to quality in the rest of the urban area and in rural areas, complete it. Similarly, a household cannot have zero at least if quality is measured as the differences between annual non-food consumption, whatever its standard of actual and imputed rent. In urban areas, Table A1.2 shows living. It is necessary to buy even basic goods for everyday that the median imputed rent is 23% lower than actual consumption (household soap, matches, etc.). As a second rent in urban areas. For rural areas, the difference is 6 step, an adjustment was made for abnormally large values. percent only. Contrary to the previous case, this adjustment is made per consumption item. The logarithm of the consumption 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 the interquartile range (this value is called the “maximum allowed”). These values are adjusted by replacing the value with the maximum allowed (or “trimming”). This choice is made to minimize the impact on inequality. 102 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 2. POVERTY LINE Having adopted a reference population, a national basket was constructed covering 85% of the most consumed food products in this reference population, excluding meals The poverty line is the value of the welfare indicator that away from home20. The basket was constructed based allows individuals to satisfy their minimum vital needs. not on nominal expenditure but on annual expenditure The approach used to construct the poverty line is that of for each product adjusted by the spatial deflator (see the the cost of basic needs19. A poverty line is constructed in calculation of the spatial deflator in Section 3). This ensures two steps: (a) calculating the food poverty line; (b) deriving that differences in prices do not affect the procedure for a total poverty line by applying to the food line a share of constructing the poverty line. Before finalizing the basket, non-food expenditure. it was verified that it represented at least 70% of the food consumption in each region or Agro-Ecological Zone (AEZ). With regard to the food poverty line, a basket of food items To obtain the food line, the basket was valued using unit providing each individual with 2,300 kilocalories (which is values from the consumption records, the same unit within the range of the internationally accepted standard values that were used to value food consumption. These for food consumption) is determined. The valuation of unit values are filled in during the survey as different non- this basket provides the food poverty line. Three factors standard consumption units (bottle, basin, plate, heap, are important in carrying out this task: (i) the reference etc.). The unit values collected in NSUs are then converted population for determining the basket; (ii) how the basket to SUs using the conversion factors from the NSU survey is constructed; and (iii) the price vector used to value that took place before the main data collection. the basket. Two non-food poverty lines were calculated, and these On the first point, the reference population must be led to two overall poverty lines. However, the non-food households around the poverty line. The objective is to poverty lines were not calculated directly; instead the have a reference population that has, as much as possible, total poverty line is calculated using the share of food the consumption habits of households that are neither consumption of households around the food poverty line. too poor nor too well-off. Given that the poverty lines are The first option consists of determining the non-food around 40% in the subregion, the interval from the second component of the poverty line as the share of non-food to third and the seventh to eighth decile is an acceptable consumption of households whose total consumption is range. Deciles 3 to 8 were used for all countries. equal to the poverty line. The second is to determine the non-food 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. 19 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. 20 Meals taken outside of the household cannot be used in the construction of the poverty line. The reason is that the process requires a correspon- dence between quantities consumed and calorie intake, and this information is not available for meals taken outside. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 103 Table A1.3. Assumptions Used in the Construction of Poverty Lines Deciles of Quantity of Approach to Minimum Maximum Gini without Gini with outlier reference kilocalories of calculating spatial deflators spatial deflators outlier adjustment population consumption poverty line adjustment Niger 3 to 8 2300 average 0.896 1.259 0.401 0.382 For the first of the two poverty lines (zref_min), households live. The welfare indicator must therefore take all these around the food poverty line are defined as those with factors into account. total consumption within plus or minus 10% of the food poverty line; if there are no households in this range, Household composition and size. The first element to households within plus or minus 20% of the food poverty consider is the size and composition of households. Here, line are used. If we call zali the previously calculated food household composition was ignored and only size was poverty lines, and alpha_min the share of household taken into account. Household composition should be food consumption whose total per capita consumption is reflected by an equivalence scale, and there is no consensus just equal to the food poverty line as defined above, the regarding the best approach to deriving an equivalence minimum poverty line is given by: scale. Moreover, virtually all countries concerned have adopted the practice of only taking household size into account. Thus the consumption aggregate is divided by household size to yield annual per capita consumption. For the second of the two thresholds (zref_max), Nevertheless, for the purpose of carrying out sensitivity households around the food poverty line are defined as tests such as ranking regions in terms of poverty levels, those with food consumption within plus or minus 10% of two equivalence scales were calculated since it is easy to the food poverty line; as before, if there are no households produce poverty figures using either one. in this range, households falling within plus or minus 20% of the food poverty line are used. If we call alpha_max the Time deflator. The second element to consider is the share of household food consumption whose per capita time at which the data is collected in the household. Nine food consumption is just equal to the food poverty line months elapsed between the start of data collection and as defined above, the maximum non-food poverty line is completion of this process. Data collection for the first given by: wave took place in several of the countries immediately following the harvest, and data collection for the second wave took place in the period well away from the harvest. During the collection period, consumer prices changed. Consumption was normalized using a time index. To do this, the national household final consumer price index is an effective tool. WAEMU countries have regional indices 3. HOUSEHOLD COMPOSITION, TEMPORAL AND that could have been used as part of this process, but SPATIAL DEFLATORS, AND INDICATORS OF WELFARE their coverage is limited as they tend to focus on country capitals, with secondary cities and rural areas less well The consumption aggregate is not an indicator of welfare represented. In fact, some countries collect regional prices because it does not allow for a fair comparison between for the purpose of calculating the national index but do households. Households are of different sizes and not calculate purely regional indices. As a result, there is compositions and face different prices depending on when some uncertainty regarding the quality of these regional the data was collected and where household members indices, and the national index is more reliable. 104 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION To calculate time deflators, if we call IPCi the consumer the poverty line of the AEZ and areas j, def_spa the spatial price index at month i, i=1, ..., n the period of n collection deflator of the AEZ and areas j is the ratio of the threshold months, we can calculate IPC as the average index during of AEZ or area j to the national threshold: the collection period by: Finally, for a household k surveyed in month i and The time deflator for each collection month is given by: belonging to AEZ/area of residence j, if we call dtotk the total annual consumption of the household and hhsizek the household size, the welfare measure indicator is: It was pointed out above that the time deflator is applied to the annual consumption of each product before the construction of the poverty line. Thus, for a household k surveyed in month i, the annual expenditure of product m (depan) is normalized by the following formula: 4. SYNTHESIS OF RESULTS The indicator of poverty easiest to calculate is the incidence of poverty, which is the percentage of people living below the poverty line. The incidence of poverty depends on the chosen poverty line. Each country should Spatial deflator. It is advisable to also apply a spatial have a national poverty line. This line is important for deflator so as to take into account disparities in the cost monitoring and evaluating public policies to combat of living between different regions and localities in the poverty in the country. The line depends on national country. A natural candidate is the regional price index, or standards, including consumption preferences and the at least the prices underlying these calculations. However, cost of living. The incidence of poverty according to the as noted above, prices collected at regional level show low national poverty line is that used for poverty diagnoses in coverage of small urban centers and rural areas. A test was national documents such as development plans or poverty conducted in order to use these as a deflator. As poverty reduction strategies. However, this poverty incidence is rates of over 70% were obtained in some countries, the not directly comparable with that of any other country idea was abandoned. The poverty lines constructed by because it depends on the national poverty line, which AEZ and area of residence were used as spatial deflators. takes into account specific norms and preferences. The approach to constructing poverty lines by AEZ and area of residence was the same as that for constructing For international comparisons and the monitoring of the national poverty line and the same national basket the Sustainable Development Goals (SDG), international was used. This basket was valued using the average unit poverty lines are more appropriate. The extreme poverty values of the AEZ and areas for the food poverty line. The line is US$1.90 per person per day at 2011 purchasing power non-food poverty line was also constructed by AEZ and parity (PPP). Here, this line was converted in FCFA taking area using the same approach as above. In other words, into account the increase in the cost of living as measured for the non-food poverty line, given that the average of by inflation between 2011 and 2018. It is important to note the minimum and maximum poverty lines was used as the that the first SDG target (Eradicating Extreme Poverty by non-food poverty line, the same approach was followed. 2030) uses the above extreme poverty line. Clearly, in the case of Côte d’Ivoire, where the minimum poverty line at the national level was used, the minimum poverty line by AEZ and area was also used. If we call zzaej NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 105 Table A1.4: Poverty measures Nominal Real per capita National Poverty rate – Poverty rate – Poverty rate – International Gini Index Inequality per capita expenditure poverty line National Urban Rural poverty rate index P90/P10 expenditure ($1.90 2011 PPP) Niger 259012.5 263225.7 181781.6 40.8 11.8 46.8 41.2 0.350 8.62 The main results are shown in Table A4. This table shows day at 2011 PPP prices. Columns 10 and 11 provide two the welfare indicator level as standardized annual per measures of inequality: the Gini index, and the ratio of capita consumption, the incidence of poverty according to average per capita consumption of the richest 10% to that the selected thresholds, and inequality indicators. of the poorest 10%. Columns 2 and 3 of Table 4 show per nominal capita The EHCVM survey is important for Niger. The survey was consumption and real per capita consumption, designed not only to produce poverty indicators but, more respectively. The difference between these two variables importantly, to generate data for in-depth analytical work is that they take into account differences in the cost of designed to assess poverty in its many dimensions. On the living between localities in each country. The reference basis of the production of the poverty figures, one of the is the national level. The fact that nominal per capita objectives of the project has been achieved; it is now a consumption is higher than real per capita consumption in matter for the countries concerned to add value to the most countries suggests that the cost of living is relatively data. Nevertheless, it is important to stress that survey higher in localities where nominal consumption is highest data is never perfect, especially in African countries, (notably capital cities and major urban centers). Columns where the level of literacy remains average. In this project, 4 to 7 show the national poverty results, the poverty line, the price data is of average quality. An assessment of the and the poverty rates for the country as a whole and for survey will allow for lessons to be learned from this first urban and rural areas in particular. Column 8 illustrates round in the work in order to improve the next one. the international poverty rate at $1.90 per person per 106 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION ANNEX 2: RECONSTRUCTION OF POVERTY ESTIMATES IN 2014 TO OBTAIN COMPARABLE POVERTY TREND BETWEEN 2014 AND 2018 The Niger EHCVM 2018/19 is part of the joint effort of survey design and implementation, there is a major between the World Bank and the WAEMU Commission break of comparability of the methodology to calculate to conduct new nationally-representative household consumption aggregates. surveys in each of the 8 WAEMU member countries. For Niger, the advantages of the newly introduced survey, Thus, to obtain comparable poverty estimates between EHCVM 2018/19, are threefold: (i) it follows international 2014 and 2018, we applied the 2018 methodology as standards for poverty measures, (ii) it is comparable to described in Annex 1 to recalculate Nigerien household the same household surveys conducted in other WAEMU consumption aggregates in 2014. We also used the same countries, and (iii) it will be comparable to the next EHCVM poverty line in 2018 at 2014 prices based on the official CPI. survey planned in 2021/22. While the EHCVM 2018/19 is comparable to the previous survey, EHCVM 2014, in terms Table A2.1. Poverty trend 2014 official 2014 reconstructed for comparability 2018 official Headcount Depth Severity Headcount Depth Severity Headcount Depth Severity National 44.1 13.1 5.4 46.2 14.7 6.3 40.8 11.2 4.3 Urban 8.6 1.4 0.4 8.0 1.4 0.4 11.8 2.4 0.7 Rural 51.0 15.4 6.3 53.7 17.2 7.5 46.8 13.0 5.0 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 107 Table A2.2. Household budget share from the reconstructed data based on EHCVM 2014 Category Food Health Education Rent Housing Transportation Durable Personal/ Other assets household items National 50.6% 9.9% 1.4% 3.6% 6.3% 5.7% 4.7% 10.0% 4.3% Urban 40.0% 6.6% 2.9% 8.9% 9.1% 8.3% 6.8% 9.2% 3.5% Rural 57.8% 12.0% 0.4% 0.1% 4.3% 3.9% 3.3% 10.6% 4.9% Agadez 56.0% 4.6% 1.7% 3.1% 5.6% 7.2% 4.7% 8.8% 4.0% Diffa 63.7% 4.9% 0.5% 4.7% 4.8% 1.8% 3.0% 8.8% 4.0% Dosso 53.8% 7.7% 1.2% 2.0% 6.2% 5.3% 4.2% 11.7% 5.2% Maradi 57.4% 6.0% 0.8% 2.1% 3.8% 5.3% 2.9% 12.2% 6.4% Tahoua 55.8% 7.3% 0.8% 2.3% 6.0% 5.1% 3.2% 11.3% 4.4% Tillaberi 47.3% 25.9% 0.4% 0.6% 5.5% 2.3% 5.0% 8.0% 2.9% Zinder 55.0% 8.9% 0.7% 1.8% 5.3% 4.4% 4.0% 10.5% 5.7% Niamey 34.9% 5.4% 4.1% 10.7% 10.2% 10.8% 8.4% 8.4% 2.6% Table A2.3. Household budget share from the official data EHCVM 2018/19 Category Food Health Education Rent Housing Transportation Durable Personal/ Other assets household items National 57.8% 3.5% 1.0% 9.7% 9.9% 4.5% 4.1% 6.1% 3.2% Urban 48.2% 3.5% 2.6% 8.5% 12.5% 7.6% 7.0% 6.8% 3.4% Rural 62.9% 3.5% 0.3% 10.4% 8.5% 2.9% 2.6% 5.8% 3.2% Agadez 53.4% 3.0% 1.0% 8.3% 12.1% 6.5% 5.4% 6.5% 3.8% Diffa 60.0% 3.1% 0.5% 9.1% 9.6% 3.5% 2.7% 7.2% 4.2% Dosso 59.5% 4.5% 0.6% 9.6% 9.7% 3.2% 2.9% 6.8% 3.3% Maradi 60.6% 3.0% 0.9% 10.2% 9.1% 3.4% 3.8% 5.7% 3.2% Tahoua 60.6% 3.8% 0.4% 9.6% 8.8% 4.2% 3.3% 5.9% 3.5% Tillaberi 61.7% 4.4% 0.6% 9.3% 9.2% 2.9% 3.1% 5.6% 3.2% Zinder 63.1% 2.9% 0.4% 10.3% 8.5% 3.1% 3.0% 5.7% 3.0% Niamey 42.2% 2.9% 3.5% 9.9% 13.6% 9.7% 8.1% 7.1% 2.9% 108 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION ANNEX 3: CONSTRUCTION OF MULTI-DEPRIVATION INDEX The Multidimensional Poverty Index (MPI), launched evenly across the dimensions within each category. Table by the United Nation’s Development Program’s Human 1 shows all the dimensions of the index. Many dimensions Development Reports Office in 2010, is designed to quantify are household-based: if the household is deprived in any and measure nonmonetary dimensions of poverty. The of the dimensions, all household members are considered original concept has three dimensions (Health, Education deprived. The cross-dimensional cut-off is 0.40; that is and Standard of living) and 10 indicators: (1) Nutrition; households are considered multi-dimensionally poor if (2) Child Mortality; (3) Years of Schooling; (4) School the weighted sum of deprivation scores is larger than 0.40. Attendance; (5) Cooking Fuel; (6) Sanitation; (7) Drinking Water; (8) Electricity; (9) Flooring; and (10) Asset Ownership. The proportion of the population that is multidimensionally However, its components may vary across countries as they poor is the incidence of poverty, or headcount ratio (H). can be affected by public policies and data availability. The average proportion of indicators in which poor people are deprived is described as the intensity of their poverty The MPI for Burkina Faso includes 15 and 16 dimensions, (A). The MPI is calculated by multiplying the incidence of respectively. They are grouped into six broad categories: poverty by the average intensity of poverty across the poor education, childhood and youth, health, access to basic (MPI = M0 = H x A); as a result, it reflects both the share of services, housing conditions, and asset ownership. Each of people in poverty and the degree to which they the categories has a weight of 0.166, which is distributed are deprived. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 109 Table A3.1. Categories and dimensions of the Burkina Faso and Niger multidimensional poverty index Category Dimension Deprived if Education Educational achievement None of the household members 15 years or older has completed six years of schooling (primary completed). Literacy Any person older than 15 years or older in a household is illiterate. Childhood and youth School attendance Any child 6 to 14 years old does not attend school. Children behind grade Any child 7 to 17 years old is behind the normal grade for his/her age. Child labour Any child 7 to 17 years old works. Health Waste management Household employs unrecommended waste management methods such as incinerate, burn, and stock in nature. Health services Any person who fell sick or ill in the last 30 days did not receive (*Not available for Burkina Faso) specialized health services. Access to Basic services Cooking fuel Household uses solid fuels and/or solid biomass fuels for cooking, such as charcoal, wood, straw, shrubs, grass, agricultural crop, and animal waste. Drinking water Household does not have access to improved drinking water (according to the SDG guideline) or safe drinking water is at least a 30-minute walk from home (roundtrip). Sanitation Household’s sanitation facility is not improved (according to the SDG guideline) or it is improved but shared with other households. Electricity Household has no electricity, generator, or solar panel. Housing conditions Floor Household has floors made of natural or rudimentary materials such as mud, wood, straw, metal sheet, sand, and animal wastes. Wall Household has walls made of natural or rudimentary materials such as wood, metal sheet, sand, animal wastes, and straw. Roof Household has roofs made of natural or rudimentary materials such as mud, straw, rustic mat, wood planks, reused wood, and unburnt bricks. Overcrowding There are more than 3 people per sleeping room. Assets ownership Assets ownership Household does not own more than one of: radio, gas/electric cooker, standing fan, TV, satellite dish/decoder, generator, telephone, bike, motorbike, refrigerator, or computer and does not own a car or truck. 110 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Table A3.2. MPI results for Niger in 2014, national and regional National Agadez Diffa Dosso Maradi Tahoua Tillaberi Zinder Niamey H 0.902 0.771 0.910 0.950 0.953 0.922 0.957 0.968 0.249 M0 0.696 0.576 0.724 0.735 0.761 0.727 0.708 0.743 0.142 Number of 22,667 2,241 2,021 2,631 3,002 2,563 2,667 3,135 4,407 observations Table A3.3. MPI results for Niger in 2018, national and regional National Agadez Diffa Dosso Maradi Tahoua Tillaberi Zinder Niamey H 0.860 0.671 0.886 0.913 0.884 0.906 0.904 0.912 0.222 M0 0.606 0.431 0.623 0.608 0.646 0.637 0.634 0.658 0.126 Number of 35,393 3,375 3,546 3,948 5,752 4,608 4,927 5,459 3,778 observations Table A3.4. Result of sensitivity test to the value of the cutoff, for Niger k = Cutoff k = 1/15 k = 2/15 k = 3/15 k = 4/15 k = 5/15 k = 6/15 k = 7/15 k = 8/15 k = 9/15 k = 10/15 k = 11/15 k = 12/15 k = 13/15 k = 14/15 H: 2014 0.992 0.978 0.969 0.947 0.926 0.902 0.869 0.832 0.767 0.706 0.579 0.418 0.284 0.103 H: 2018 0.988 0.969 0.956 0.922 0.897 0.860 0.817 0.772 0.664 0.545 0.374 0.203 0.101 0.025 H: Change from -0.004 -0.009 -0.014 -0.024 -0.030 -0.043 -0.052 -0.061 -0.104 -0.161 -0.206 -0.215 -0.183 -0.077 2014 to 2014 M0: 2014 0.720 0.718 0.717 0.711 0.705 0.696 0.682 0.663 0.626 0.588 0.499 0.376 0.265 0.101 M0: 2018 0.639 0.637 0.635 0.627 0.619 0.606 0.587 0.564 0.503 0.428 0.308 0.177 0.092 0.024 M0: Change from -0.081 -0.081 -0.082 -0.084 -0.086 -0.091 -0.095 -0.099 -0.124 -0.160 -0.191 -0.199 -0.173 -0.077 2014 to 2018 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 111 Figure A3.1. Result of sensitivity test to the value of the cutoff, for Niger 1.000 0.900 0.800 0.700 0.600 0.500 0.400 0.300 0.200 0.100 0.000 0.00 0.20 0.40 0.60 0.80 1.00 Cutoff H:2014 H:2018 0.800 0.700 0.600 0.500 0.400 0.300 0.200 0.100 0.000 0.00 0.20 0.40 0.60 0.80 1.00 M0:2014 M0:2018 112 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION ANNEX 4. DECOMPOSITION RESULTS Table A4.1. Growth and redistribution decomposition of poverty changes in Niger between 2014 and 2018 Change in poverty headcount 2014 2018 Actual change Growth Redistribution Residual National 46.2 40.8 -5.4 -2.9 -2.5 0.0 Urban 8.0 11.8 3.8 -1.0 4.9 0.0 Rural 53.7 46.8 -6.9 -2.7 -4.2 0.0 Table A4.2. Non-linear Oaxaca-Blinder decomposition of poverty change in Niger between 2014 and 2018 Overall Total Endowment Return to endowment Interaction Poverty rates in 2018 0.408*** (0.000) Poverty rates in 2014 0.463*** (0.000) Change in poverty rates -0.054*** (0.000) Endowments 0.002*** (0.000) Coefficients -0.059*** (0.000) Interaction 0.003*** (0.000) Rural -0.000*** -0.187*** -0.002*** (0.000) (0.001) (0.000) Agadez 0.000*** 0.011*** 0.001*** (0.000) (0.000) (0.000) Diffa 0.000*** 0.002*** 0.000*** (0.000) (0.000) (0.000) Dosso -0.000* 0.002*** 0.000* (0.000) (0.000) (0.000) NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 113 Overall Total Endowment Return to endowment Interaction Maradi -0.000*** 0.004*** -0.000*** (0.000) (0.000) (0.000) Tahoua 0.000*** 0.032*** 0.001*** (0.000) (0.000) (0.000) Tillaberi 0.000*** 0.017*** 0.000*** (0.000) (0.000) (0.000) Zinder -0.000*** 0.011*** -0.001*** (0.000) (0.000) (0.000) Number of kids age 0-5 0.001*** 0.030*** -0.002*** (0.000) (0.000) (0.000) Number of kids age 6-14 -0.005*** 0.029*** 0.005*** (0.000) (0.001) (0.001) Number of adults age 15-64 -0.001*** 0.007*** 0.001*** (0.000) (0.001) (0.000) Number of adults age 65+ -0.000*** -0.010*** -0.002*** (0.000) (0.000) (0.000) Gender of head: Female -0.000*** 0.009*** -0.002*** (0.000) (0.000) (0.000) Age of head 0.001*** 0.134*** 0.005*** (0.000) (0.001) (0.001) Marital status of head: Never married -0.000*** 0.002*** 0.002*** (0.000) (0.000) (0.000) Marital status of head: -0.001*** 0.000*** -0.002*** Widowed/Divorced/Separated (0.000) (0.000) (0.000) Head having health problem 0.000*** -0.001*** -0.000*** (0.000) (0.000) (0.000) Number of family members with no 0.000*** 0.055*** 0.006*** education (0.000) (0.001) (0.001) Number of family members with 0.000*** -0.010*** -0.000*** primary education (0.000) (0.000) (0.000) Number of family members with secondary (dropped) (dropped) (dropped) or above education Education of head: Primary 0.000*** -0.002*** 0.000*** (0.000) (0.000) (0.000) Education of head: Secondary or above -0.000*** 0.006*** -0.001*** (0.000) (0.000) (0.000) Ownership of house -0.000*** -0.021*** -0.001*** (0.000) (0.001) (0.000) Ownership of TV 0.001*** 0.013*** 0.001*** (0.000) (0.000) (0.000) 114 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Overall Total Endowment Return to endowment Interaction Ownership of fan 0.000 -0.031*** -0.000*** (0.000) (0.000) (0.000) Ownership of radio 0.003*** 0.021*** 0.008*** (0.000) (0.000) (0.001) Ownership of motorbike -0.001*** -0.006*** 0.001*** (0.000) (0.000) (0.000) Ownership of phone -0.001*** -0.007*** 0.001*** (0.000) (0.000) (0.000) Walls with improved materials 0.003*** -0.010*** 0.014*** (0.000) (0.000) (0.002) Roofs with improved materials -0.001*** 0.001*** -0.000*** (0.000) (0.000) (0.000) Floor with improved materials -0.000*** -0.001*** 0.000*** (0.000) (0.000) (0.000) Access to electricity 0.000*** -0.008*** -0.000*** (0.000) (0.000) (0.000) Access to drinking water 0.000*** -0.044*** 0.007*** (0.000) (0.000) (0.001) Access to improved sanitation facility 0.001*** 0.011*** 0.003*** (0.000) (0.000) (0.000) Access to improved waste management 0.000*** -0.008*** 0.003*** (0.000) (0.000) (0.000) Having family shock 0.001*** -0.002*** 0.003*** (0.000) (0.000) (0.000) Having natural disaster shock -0.000 0.008*** -0.001*** (0.000) (0.000) (0.000) Having economic shock -0.001*** -0.032*** -0.011*** (0.000) (0.000) (0.002) Number of employed family members 0.000*** -0.080*** -0.035*** (0.000) (0.000) (0.005) Employment status of head: Employed 0.000*** -0.097*** 0.001*** (0.000) (0.001) (0.000) Sector of head’s employment: Industry 0.000*** -0.001*** 0.004*** (0.000) (0.000) (0.001) Sector of head’s employment: Services 0.000*** -0.007*** -0.000*** (0.000) (0.000) (0.000) _cons 0.099*** (0.002) Number of observations 9,640 Adjusted R2 note: .01 - ***; .05 - **; .1 - *; standard errors in brackets NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 115 Table A4.3. Non-linear Oaxaca-Blinder decomposition of poverty change in Niger between 2014 and 2018 (presented in % of poverty change) Overall Total Endowment Return to endowment Interaction Change in poverty rates 100.0% Endowments -3.0% Coefficients 109.2% Interaction -6.2% Rural 0.5% 343.1% 3.0% Agadez -0.4% -19.9% -1.4% Diffa -0.1% -2.9% -0.1% Dosso 0.0% -2.8% 0.0% Maradi 0.1% -7.7% 0.2% Tahoua -0.4% -58.6% -1.3% Tillaberi -0.1% -32.1% -0.3% Zinder 0.2% -21.1% 1.0% Number of kids age 0-5 -2.2% -55.6% 3.9% Number of kids age 6-14 8.4% -52.7% -8.9% Number of adults age 15-64 1.4% -12.8% -0.9% Number of adults age 65+ 0.3% 17.8% 3.4% Gender of head: Female 0.2% -16.1% 4.0% Age of head -1.4% -247.3% -9.6% Marital status of head: Never married 0.1% -3.5% -3.0% Marital status of head: 1.0% -0.6% 3.9% Widowed/Divorced/Separated Head having health problem -0.5% 2.4% 0.1% Number of family members with -0.2% -101.9% -10.7% no education Number of family members with 0.0% 18.5% 0.2% primary education Number of family members with secondary 0.0% 0.0% 0.0% or above education Education of head: Primary -0.2% 2.8% -0.7% Education of head: Secondary or above 0.5% -11.4% 1.8% Ownership of house 0.5% 37.7% 2.7% 116 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Overall Total Endowment Return to endowment Interaction Ownership of TV -1.0% -23.8% -2.2% Ownership of fan 0.0% 57.6% 0.5% Ownership of radio -5.8% -39.1% -15.2% Ownership of motorbike 1.2% 11.1% -2.4% Ownership of phone 1.5% 13.2% -1.0% Walls with improved materials -5.3% 18.2% -25.4% Roofs with improved materials 1.7% -2.1% 0.6% Floor with improved materials 0.4% 1.4% -0.2% Access to electricity -0.1% 15.3% 0.5% Access to drinking water -0.9% 80.3% -12.6% Access to improved sanitation facility -1.6% -19.8% -4.7% Access to improved waste management -0.5% 15.4% -5.6% Having family shock -1.5% 4.4% -5.6% Having natural disaster shock 0.0% -15.4% 2.0% Having economic shock 2.1% 58.0% 20.3% Number of employed family members -0.5% 146.8% 64.8% Employment status of head: Employed -0.2% 179.3% -1.0% Sector of head’s employment: Industry -0.4% 2.4% -6.6% Sector of head’s employment: Services -0.1% 12.6% 0.3% cons -181.8% NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 117 Table A4.4 Unconditional quantile regressions for change in consumption growth among households in the bottom 40 percent of the consumption distribution Total Endowment Return to endowment Interaction Log consumption in 2018 12.099*** (0.000) Log consumption in 2014 12.014*** (0.000) Change in Log consumption 0.085*** (0.000) Endowments 0.026*** (0.000) Coefficients 0.087*** (0.000) Interaction -0.027*** (0.000) Rural 0.002*** 0.246*** -0.002*** (0.000) (0.001) (0.000) Agadez -0.001*** -0.008*** 0.001*** (0.000) (0.000) (0.000) Diffa -0.000*** -0.004*** 0.000*** (0.000) (0.000) (0.000) Dosso 0.000* 0.010*** -0.000* (0.000) (0.000) (0.000) Maradi 0.000*** -0.003*** -0.000*** (0.000) (0.000) (0.000) Tahoua -0.002*** -0.040*** 0.001*** (0.000) (0.000) (0.000) Tillaberi -0.000*** -0.021*** 0.000*** (0.000) (0.000) (0.000) Zinder 0.001*** -0.015*** -0.001*** (0.000) (0.000) (0.000) Number of kids age 0-5 -0.008*** -0.118*** -0.009*** (0.000) (0.001) (0.000) Number of kids age 6-14 0.033*** -0.163*** 0.030*** (0.000) (0.001) (0.000) Number of adults age 15-64 -0.007*** -0.260*** 0.021*** (0.000) (0.001) (0.000) Number of adults age 65+ -0.000 -0.000 0.000 (0.000) (0.000) (0.000) Gender of head: Female -0.001*** -0.012*** -0.003*** (0.000) (0.000) (0.000) 118 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Total Endowment Return to endowment Interaction Age of head -0.008*** -0.282*** 0.012*** (0.000) (0.001) (0.000) Marital status of head: Never married 0.002*** -0.005*** 0.005*** (0.000) (0.000) (0.000) Marital status of head: -0.002*** 0.001*** 0.004*** Widowed/Divorced/Separated (0.000) (0.000) (0.000) Head having health problem -0.001*** 0.017*** -0.001*** (0.000) (0.000) (0.000) Number of family members with 0.015*** 0.291*** -0.034*** no education (0.000) (0.002) (0.000) Number of family members with 0.000*** 0.115*** -0.002*** primary education (0.000) (0.001) (0.000) Number of family members with secondary (dropped) 0.029*** 0.000*** or above education (0.000) (0.000) Education of head: Primary 0.001*** -0.003*** -0.001*** (0.000) (0.000) (0.000) Education of head: Secondary or above 0.001*** -0.004*** -0.001*** (0.000) (0.000) (0.000) Ownership of house 0.000 -0.002** 0.000** (0.000) (0.001) (0.000) Walls with improved materials -0.021*** 0.014*** 0.021*** (0.000) (0.000) (0.000) Roofs with improved materials 0.009*** -0.005*** -0.002*** (0.000) (0.000) (0.000) Floor with improved materials 0.002*** 0.001*** 0.000*** (0.000) (0.000) (0.000) Access to electricity -0.000*** 0.014*** -0.000*** (0.000) (0.000) (0.000) Access to drinking water -0.004*** 0.071*** 0.012*** (0.000) (0.000) (0.000) Access to improved sanitation facility -0.003*** -0.011*** 0.003*** (0.000) (0.000) (0.000) Access to improved waste management -0.002*** 0.006*** 0.002*** (0.000) (0.000) (0.000) Having family shock -0.010*** 0.007*** 0.009*** (0.000) (0.000) (0.000) Having natural disaster shock -0.001*** -0.004*** -0.001*** (0.000) (0.000) (0.000) Having economic shock 0.012*** 0.052*** -0.020*** (0.000) (0.000) (0.000) NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 119 Total Endowment Return to endowment Interaction Number of employed family members 0.021*** 0.168*** -0.082*** (0.000) (0.000) (0.000) Employment status of head: Employed -0.000*** 0.064*** 0.000*** (0.000) (0.001) (0.000) Sector of head’s employment: Industry -0.000*** 0.002*** 0.007*** (0.000) (0.000) (0.000) Sector of head’s employment: Services -0.000*** 0.015*** -0.000*** (0.000) (0.000) (0.000) _cons -0.079*** (0.002) Number of observations 9,640 Adjusted R2 note: .01 - ***; .05 - **; .1 - *; standard errors in brackets Table A4.5 Unconditional quantile regressions for change in consumption growth among households in the bottom 40 percent of the consumption distribution (presented in share of change in consumption growth) Total Endowment Return to endowment Interaction Change in Log consumption 100.0% Endowments 30.1% Coefficients 102.1% Interaction 32.2% Rural 2.0% 216.1% -2.1% Agadez -1.0% -6.6% 0.5% Diffa -0.3% -3.1% 0.1% Dosso 0.0% 8.6% 0.0% Maradi 0.1% -2.8% -0.1% Tahoua -1.5% -35.1% 0.9% Tillaberi -0.2% -18.0% 0.2% Zinder 0.8% -12.8% -0.7% Number of kids age 0-5 -7.0% -104.0% -8.1% Number of kids age 6-14 29.2% -143.1% 26.8% Number of adults age 15-64 -5.7% -227.9% 18.2% 120 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Total Endowment Return to endowment Interaction Number of adults age 65+ 0.0% -0.1% 0.0% Gender of head: Female -0.8% -10.4% -2.8% Age of head -6.7% -248.1% 10.7% Marital status of head: Never married 1.7% -4.4% 4.1% Marital status of head: -1.8% 0.5% 3.9% Widowed/Divorced/Separated Head having health problem -1.1% 15.3% -0.7% Number of family members with 12.8% 255.7% -29.7% no education Number of family members with 0.1% 101.2% -1.3% primary education Number of family members with secondary 0.0% 25.4% 0.3% or above education Education of head: Primary 0.4% -2.9% -0.8% Education of head: Secondary or above 1.1% -3.8% -0.7% Ownership of house 0.0% -1.4% 0.1% Walls with improved materials -18.9% 11.9% 18.4% Roofs with improved materials 7.6% -4.2% -1.4% Floor with improved materials 1.7% 1.2% 0.2% Access to electricity -0.3% 12.3% -0.4% Access to drinking water -3.6% 62.4% 10.8% Access to improved sanitation facility -2.4% -9.3% 2.4% Access to improved waste management -1.6% 5.4% 2.2% Having family shock -9.0% 5.9% 8.3% Having natural disaster shock -1.1% -3.3% -0.5% Having economic shock 10.4% 45.5% -17.6% Number of employed family members 18.6% 147.5% -71.8% Employment status of head: Employed -0.3% 56.1% 0.3% Sector of head’s employment: Industry -0.3% 2.1% 6.5% Sector of head’s employment: Services -0.3% 13.4% -0.4% _cons -69.0% NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 121 ANNEX 5. IMPACTS OF SHOCKS ON HOUSEHOLDS’ WELFARE 1. BACKGROUND nutrition level and destruction of assets. It also indirectly breaks the social relation and cohesion, affects both the exchange and employment market, changes the structure Assessing vulnerability in different households is of political institutions, and slows economic growth. important. It provides policymakers with guidance on how Hess (2003) found that on average individuals who live to efficiently allocate resources to prevent households in a country that has experienced some conflict would from falling into poverty in the future. Unlike measuring permanently give up to approximately 8 percent of their poverty, which is ex-post using household income or current level of consumption to live in a purely peaceful consumption, measuring vulnerability is conceptually and world. Jennings & Sanchez-Pages (2017) conclude that empirically much more complex. It quantifies the threat when the threat is severe, social capital and welfare are of poverty in the future. Therefore, vulnerability analysis likely to fall and this effect of an external threat on social has two components: the likelihood of shocks occurring, capital is stronger in poor economies. and the impact of shocks on welfare (Calvo & Dercon 2013; Hohberg et al. 2018; Hoddinott & Quisumbing 2003). A large body of literature have studied the impact of weather shocks on people’s wellbeing ((Baez, de la Fuente, Both idiosyncratic and covariate shocks have serious & Santos, 2010) (Dell, Jones, & Olken, 2014)). The impact impacts on poor households’ income and overall well- can be short-run and long-run. Kazianga & Udry (2006) being. In Niger, the most commonly reported shocks find that droughts and erratic rainfall halve crop income over the period 2018-2019 are drought and illness/ and reduce consumption significantly among affected death of family members, followed by illness of family households. The negative impact also carries over the members and high food prices (Figure 1). Rural and urban longer term. Children who become stunted due to droughts households experience different types of shocks. In rural or floods often do not fully recover later in life, resulting areas, 35.6% of families report drought as the most severe in lower school attainment and earnings in adulthood shock. For urban families, on the other hand, illness and (Alderman, Hoddinott, & Kinsey, 2006). Moreover, the death of family members are the most frequently reported impact often hit the deprived population most (Del Ninno shocks. 38.4% of the urban households reported these two & Lundberg, 2005). Health shocks are also found to be shocks as the most severe shocks they experienced in the negatively correlated with welfare. In Indonesia, Gertler past three years. Drought is less of a problem for urban and Gruber (2002) found that households faced with people since nearly 95% of the household income is non- health shocks are unable to fully ensure consumption. agriculture income. It’s worth noting that although fewer Lindelow and Wagstaff (2005) found that negative health households experienced conflict/violence and reported it shocks are associated with a significant reduction in as a shock, the impact is severe as it affects the household income in China. Impact heterogeneity is also found by welfare both directly and indirectly (Justino, 2011). Its direct Wagstaff (2007) and Atake (2018). Wagstaff (2007) discovers impact includes illness of family member, reduction in 122 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION that urban households were more vulnerable in terms 2. DATA of reduced earned income; while Atake (2018) finds that the poorest households, such as those in Sub-Saharan The Harmonized Household Living Standards Survey French-Speaking Africa countries (SSAF) may be the most 2018/19 for Niger is used for this study, which is nationally vulnerable to health shocks. representative. It contains two rounds, with the first round conducted from September to December in 2018 and the The rest of this paper is organized as follows: section 2 second round carried out between April and July in 2019. It describes the data and limitations; section 3 presents the provides households self-reported shock data and various methodology; results and robustness checks are discussed welfare indicators. in section 4. Section 5 concludes. Figure 2. Average region-level welfare measures NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 123 124 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION The welfare indicators used in this study include both The survey data indicates that if a person feels sick, only monetary measure and non-monetary measure. More 56.5% of the time would he or she go to the hospital. specifically, the following dependent variables are used: Of those who didn’t seek medical advice in the past 30 percentage of people feeling sick in the past 30 days, days, 61% of the rural population and 73% of the urban percentage of them going to the hospital, primary and population self-medicate themselves. secondary school attendance rate, income per capita, consumption per capita, food consumption per capita, The school enrollment rate is used as an education farm productivity, FIES, and food insecurity rate. Figure 2 indicator. Education in Niger is structured in much the shows these measures averaged at the regional level. same way as in the rest of the world: primary, secondary, and higher education. As of 2018, despite the efforts to Monetary indicators - per capita household income and improve education, the country had the third lowest consumptions - are generally regarded as the best proxies adult literacy rate in the world (31%), comparing to the for household welfare. In Niger, urban households make world average of 86.3%. On average, only 45% of the kids 27% more annual income than rural households, allowing between 6 and 18 years old in the household are enrolled them to have more savings to cope with adverse events. in primary or secondary schools. The situation is much The income structure also differs in urban and rural better in urban households, being 77% of children aged areas. Figure 2.1 shows that urban households depend on 6 to 18 are enrolled in primary and secondary schools. salary and enterprise income. In the capital city Niamey, In rural households, this rate is about 39%. The biggest almost all of the household income is gained from non- problem the education sector encounter is lack of books agriculture activities. On the other hand, rural households and equipment. Of those who reported a certain problem rely mainly on agriculture and agriculture income. In the in education, the majority is rural population, meaning south, where the main food crops are planted, agriculture that rural people face problems more frequently and are income is the most important source of income. thus more likely to drop out of school. Table 1 presents summary statistics of all the variables used in our analysis. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 125 Table A5.1. Summary Statistics Rural Urban All Dependent variables Feel sick in the past 30 days 0.406 0.383 0.402 Go to hospital if sick in the past 30 days 0.550 0.599 0.558 Primary and secondary attendance rate 0.306 0.672 0.366 Logged income per capita 0.079 0.243 0.107 Logged consumption per capita 12.29 12.99 12.41 Logged food consumption per capita 11.79 12.33 11.89 Logged farm productivity 0.001 0.014 0.002 FIES 3.729 2.110 3.452 Food insecurity rate 0.117 0.020 0.100 Shocks Self-reported death shock 0.078 0.081 0.078 Self-reported drought shock 0.333 0.036 0.283 Self-reported conflict shock 0.013 0.020 0.014 Drought shock (CHIRPS) 0.071 0.123 0.080 Severe drought shock (CHIRPS) 0.018 0.071 0.027 Logged number of fatalities within 15km 0.330 0.558 0.369 Logged number of fatalities within 20km 0.340 0.572 0.380 Logged number of fatalities within 25km 0.351 0.631 0.399 Household characteristics Rural 0.832 Household size 5.97 5.92 5.96 Dependency ratio 157.18 107.14 148.58 Enrollment rate 0.31 0.67 0.37 Age 0-14 3.35 2.72 3.24 Age 15-64 2.46 3.04 2.56 Age 65+ 0.15 0.16 0.16 Agriculture household 0.72 0.15 0.62 % ever experienced a shock 0.71 0.50 0.68 126 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Rural Urban All Head of household characteristics Average age 43.32 46.75 43.91 % women 0.18 0.16 0.17 % primary education 0.09 0.19 0.11 % secondary education 0.05 0.17 0.07 % tertiary education 0.00 0.09 0.02 Literary rate (French) 0.12 0.43 0.17 Literacy rate 0.31 0.54 0.35 Housing characteristics % have access to electricity 0.07 0.58 0.16 % have access to water 0.02 0.50 0.10 % have toilet 0.01 0.26 0.05 As shown in part 2 of Table 1, drought is calculated using fatalities as the conflict shock measurement. In addition, CHIRPS data, which provides a measure of monthly rainfall a dummy from EHCVM 2018/19 which is self-reported by for each 5.7 by 5.7 km pixel in the country. Households household if it experienced any armed conflict/violence/ were matched to the average of its 4 nearest rainfall insecurity in the past 3 years is also used to quantify observations using the GPS coordinates in the EHCVM conflict shock. 2018/19. If during the planting season (June to September 2018), any month’s rainfall falls below 1 standard deviation Households self-reported shock on the death of a below its 10-year monthly average, it’s considered a household member in the past 3 years from EHCVM 2018/19 drought shock to that household. I also define severe is used to define idiosyncratic health shock. As introduced drought as two consecutive months below their respective earlier, high food piece is also an important shock to 10-year average rainfall during the planting season. Self- households. Many literatures have discovered negative reported drought shock from EHCVM 2018/19 is also used, impact of high price level on consumptions ( (Alem & which indicates whether the household has experienced Söderbom, 2012) (Cornelsen, et al., 2015) (Andreyeva, Long, any drought shock in the past 3 years. & Brownell, 2010)). A perfect dataset would include several years’ worth of observations for each household, and Data on conflict events come from the Armed Conflict even better, information on what could happen and how Location & Event Data Project (ACLED). It collects the dates, probable this is/was in differing states of the world. In the actors, locations, fatalities, and modalities of all reported absence of such scenarios, and of panel data, price shock political violence and protest events across Africa and is not explored in this study. many other regions. Conflict events reported in ACLED data include battles, explosions/remote violence, protests, 3. METHODOLOGY riots, strategic developments, and violence against civilians, as well as their associated number of fatalities. In this paper, we follow the model in (Hill, Nikoloski, & I match household to the conflict events within 15km, Tao, 2019) to examine the impact of shocks on household 20km, 25 km radius and calculated the logged number of welfare and other outcomes. NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 127 To capture the impact of shocks on households’ welfare geography. This study relies on the following observation and outcomes in education, health, and food security, and to address this: although variation in the probability to further explore the possible heterogeneity of impact, distribution of Si may not be considered exogenous two regression models are adopted in this paper. The two to welfare across households, the timing of a shock models estimated are given by equation (1) and (2) below: conditional on its distribution is exogenous (Thomas et al., 2010; Anttila-Hughes and Hsiang, 2013). Therefore, the historical distribution of rainfall data is considered in the construction of drought indicators in this study. Moreover, idiosyncratic health shock is difficult to measure yi is a welfare indicator of household i including whether exogenously. This possible endogeneity is taken care of household members feel sick; if sick, whether they go to by including a rich set of household characteristics that the hospital; primary and secondary school enrollment could corelated with differences in self-reporting of rate; food insecurity experience scale; food consumption the shock. per capita; total consumption per capita, and income per capita. Si is a vector comprising both idiosyncratic shocks 4. RESULTS AND ROBUSTNESS CHECKS of the death of a family member and covariate shocks drought and conflict. Regression analysis is used to quantify the impact of shocks on various welfare measures. As a first step, no A rich set of household characteristics are represented by geographic difference is considered. Results are presented Xi , including household size, female-headed household, in Table 2. household age, household head education level, and the number of people in different age groups. Hi is a set of If the household ever experienced the death of family variables capturing access to basic infrastructures like members in the past 3 years, the probability of household water, electricity, and improved sanitation. Ai includes the members feeling sick in the past 30 days is increased by land size, pesticide and fertilizer usage, and the value of 4.4 percentage points. Drought has the expected negative owned livestock. Ri takes care of the region fixed effect. impact on health but not statistically significant. (Mara, Equation (2) also includes the interactions of rural dummy Lane, Scott, & Trouba, 2010) and (Bartram & Cairncross, and shock variables to identify which area bears 2010) find that improved sanitation has significant impacts more impact. on health. In this study, although negative correlation is found between improved sanitation and health, it’s not The main identification challenge in assessing the impact significant. Femake-headed households are 4 percentage of shocks is endogeneity. Poor households with limited points more likely to be sick. coping mechanisms are more likely to report shocks than the non-poor households, who tend to have good If feeling sick in the past 30 days, people facing conflict mitigation and coping strategies in place. To address the shock in the past 3 years are more likely to go to the endogeneity issue, I use weather shock indicators derived hospital (9 percentage points increase in likelihood). by exogenously measured rainfall and conflict shock Household having access to electricity increases the indicators derived from ACLED dataset. These indicators chance of people seeking medical help by 6 percentage are matched to household survey by GPS coordinates. points. The survey data shows households with access to electricity or toilet have higher annual income (Figure 3). Even when measured exogenously, the probability of Households with access to basic infrastructure generally occurrence of a shock is not exogenous to consumption have higher education level, which explains why these per capita across households. Arid areas are more likely households are more likely to seek health services to be poor and hit by frequent droughts due to their (Woldemicael, 2010). 128 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Table A5.2. Regression Results – Basic Model (1) (2) (3) (4) (5) (6) (7) (8) (9) VARIABLES sick_30d med_30d attend_pct lnincpc lnexp lnfoodexp ltfp fies food_security hhsize 0.0455*** -0.00310 -0.0458** -0.0152*** 0.00221 0.00344 -0.000666 0.214 -0.0106 (0.0136) (0.0201) (0.0196) (0.00490) (0.0225) (0.0237) (0.000728) (0.139) (0.0181) hhh_fem 0.0407*** -0.000494 0.0554*** -0.0226*** -0.0257 -0.0468 -0.00120 0.351*** 0.0189 (0.0148) (0.0222) (0.0212) (0.00378) (0.0262) (0.0313) (0.00112) (0.129) (0.0201) hhh_age -0.000879** -0.000754 0.00355*** 0.000317** -0.00175** -0.000985 -1.88e-05 -0.00461 0.00166*** (0.000426) (0.000655) (0.000623) (0.000150) (0.000769) (0.000862) (1.59e-05) (0.00428) (0.000611) hhh_primary 0.00356 0.0378 0.0393* 0.0139** 0.0977*** 0.0862*** -0.00243 -0.202 -0.0118 (0.0174) (0.0246) (0.0222) (0.00676) (0.0273) (0.0300) (0.00213) (0.146) (0.0212) hhh_ 0.0198 0.0907*** 0.175*** 0.0361*** 0.175*** 0.125*** -0.00258 -0.352** -0.0517* secondary (0.0198) (0.0270) (0.0305) (0.00868) (0.0365) (0.0394) (0.00222) (0.167) (0.0272) hhh_tertiary -0.0340 0.0148 0.202*** 0.254*** 0.585*** 0.394*** -0.00957 -1.021*** -0.00468 (0.0265) (0.0548) (0.0397) (0.0353) (0.0469) (0.0487) (0.00951) (0.265) (0.0198) num_kids0005 -0.0511*** 0.0128 0.0429** 0.00789 -0.101*** -0.0990*** 0.000530 -0.118 0.0641*** (0.0139) (0.0206) (0.0201) (0.00515) (0.0233) (0.0249) (0.000530) (0.146) (0.0193) num_kids0614 -0.0551*** 0.00649 0.0921*** 0.00825* -0.0733*** -0.0681*** 0.000514 -0.169 0.0376** (0.0137) (0.0204) (0.0195) (0.00475) (0.0223) (0.0238) (0.000547) (0.138) (0.0181) num_1564 -0.0584*** 0.00148 0.0338* 0.0107** -0.0175 -0.0256 0.000853 -0.219 0.0126 (0.0133) (0.0200) (0.0200) (0.00492) (0.0228) (0.0243) (0.000958) (0.134) (0.0187) access_elec 0.0114 0.0568** 0.219*** 0.0311*** 0.365*** 0.309*** 0.00393 -1.005*** -0.0978*** (0.0191) (0.0262) (0.0255) (0.0296) (0.0328) (0.0336) (0.00343) (0.140) (0.0135) access_water -0.0134 0.0279 0.103*** 0.298*** 0.220*** 0.218*** 0.0196 -0.737*** -0.0527*** (0.0205) (0.0301) (0.0293) (0.0389) (0.0395) (0.0408) (0.0191) (0.194) (0.0163) NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 129 (1) (2) (3) (4) (5) (6) (7) (8) (9) VARIABLES sick_30d med_30d attend_pct lnincpc lnexp lnfoodexp ltfp fies food_security type_ -0.0233 0.0207 0.0462 0.277*** 0.190*** 0.189*** -0.00847 -1.059*** 0.00881 sanitation (0.0196) (0.0324) (0.0308) (0.0386) (0.0413) (0.0431) (0.00844) (0.173) (0.0217) landsize 0.00382** 0.00169 0.00735** 0.00801** 0.00615* -0.000299 -0.0420*** -0.00562*** (0.00191) (0.00255) (0.00294) (0.00337) (0.00338) (0.000192) (0.0146) (0.00206) use_pesti 0.00112 0.00721 0.0686** 0.0765** 0.103*** 0.000263 0.260 -0.0584** (0.0165) (0.0263) (0.0275) (0.0310) (0.0334) (0.000637) (0.168) (0.0236) use_fer 0.00510 -0.0285 0.117*** 0.138*** 0.152*** -0.000487 -0.435*** -0.0273 (0.0156) (0.0242) (0.0264) (0.0294) (0.0310) (0.000976) (0.152) (0.0230) livestock -0.00161* -0.00260* 0.00916*** 0.0112*** 0.0137*** 7.59e-05 -0.0687*** -0.00633*** (0.000901) (0.00141) (0.00156) (0.00177) (0.00184) (7.48e-05) (0.00830) (0.00129) death_hh 0.0534*** 0.00425 -0.0178 -0.00279 0.00304 -0.00805 1.63e-05 0.487*** -0.0166 (0.0185) (0.0255) (0.0252) (0.0249) (0.0289) (0.0305) (0.000603) (0.177) (0.0227) ddrought -0.00973 0.0231 -0.00596 -0.108*** -0.113*** -0.107*** -0.000803 0.743*** 0.0613*** (0.0112) (0.0172) (0.0159) (0.0173) (0.0206) (0.0217) (0.000775) (0.103) (0.0158) conflict -0.0635** 0.0943* 0.00753 -0.0155 -0.0918 -0.108* 0.00145 1.237*** 0.0363 (0.0307) (0.0539) (0.0543) (0.0460) (0.0640) (0.0633) (0.00107) (0.299) (0.0426) Constant 0.410*** 0.642*** 0.0903*** 12.74*** 12.16*** 12.17*** 0.0589*** 3.735*** -0.0820** (0.0239) (0.0401) (0.0329) (0.0433) (0.0480) (0.0510) (0.0131) (0.240) (0.0322) Observations 6,022 4,915 4,584 6,022 6,022 6,022 3,612 6,022 6,022 R-squared 0.063 0.031 0.196 0.486 0.323 0.296 0.051 0.158 0.141 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 130 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Table A5.3. Regression Results – Full Model (10) (11) (12) (13) (14) (15) (16) (17) VARIABLES sick_30d med_30d attend_pct lnincpc lnexp lnfoodexp fies food_security death_hh_rural 0.0654*** 0.0176 -0.0161 -0.00705 0.0128 0.0297 0.568*** -0.0318 (0.0214) (0.0289) (0.0283) (0.00441) (0.0270) (0.0314) (0.205) (0.0261) death_hh_ -0.00408 -0.0604 -0.00762 0.00534 -0.0790 -0.125* 0.109 0.0551 urban (0.0290) (0.0473) (0.0419) (0.0181) (0.0614) (0.0701) (0.273) (0.0399) drought_rural -0.00738 0.0245 0.0131 -0.00813** -0.0900*** -0.0961*** 0.715*** 0.0548*** (0.0114) (0.0176) (0.0162) (0.00364) (0.0176) (0.0210) (0.106) (0.0163) drought_urban -0.0739* 0.00291 0.138*** -0.0163 -0.250*** -0.217** 0.650 0.0763 (0.0420) (0.0695) (0.0530) (0.0256) (0.0661) (0.0851) (0.514) (0.0517) conflict_rural -0.0769** 0.0591 0.0224 0.00140 -0.0555 -0.166** 1.395*** 0.0667 (0.0367) (0.0668) (0.0650) (0.00950) (0.0510) (0.0745) (0.348) (0.0531) conflict_urban -0.0305 0.177*** -0.120* -0.0121 0.0253 0.0398 0.950* -0.0151 (0.0413) (0.0515) (0.0702) (0.0199) (0.0719) (0.0755) (0.515) (0.0317) rururb -0.0168 -0.0213 -0.247*** -0.0358*** -0.213*** -0.219*** 0.385** 0.0960*** (0.0197) (0.0284) (0.0257) (0.00819) (0.0301) (0.0327) (0.166) (0.0167) hhsize 0.0455*** -0.00329 -0.0441** -0.0152*** 0.00222 0.00299 0.215 -0.0103 (0.0136) (0.0202) (0.0194) (0.00490) (0.0221) (0.0235) (0.138) (0.0181) hhh_fem 0.0419*** 0.00141 0.0516** -0.0220*** -0.0205 -0.0403 0.351*** 0.0156 (0.0149) (0.0223) (0.0210) (0.00375) (0.0259) (0.0309) (0.128) (0.0199) hhh_age -0.000882** -0.000770 0.00326*** 0.000271* -0.00199*** -0.00122 -0.00404 0.00176*** (0.000427) (0.000656) (0.000618) (0.000149) (0.000762) (0.000857) (0.00428) (0.000611) hhh_primary 0.00436 0.0379 0.0245 0.0119* 0.0893*** 0.0776*** -0.174 -0.00784 (0.0174) (0.0245) (0.0220) (0.00654) (0.0269) (0.0296) (0.145) (0.0210) hhh_secondary 0.0198 0.0908*** 0.180*** 0.0364*** 0.177*** 0.127*** -0.357** -0.0523* (0.0198) (0.0271) (0.0304) (0.00869) (0.0357) (0.0387) (0.167) (0.0269) hhh_tertiary -0.0366 0.0135 0.220*** 0.257*** 0.593*** 0.402*** -1.072*** -0.00778 (0.0268) (0.0551) (0.0383) (0.0352) (0.0482) (0.0515) (0.262) (0.0204) NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 131 (10) (11) (12) (13) (14) (15) (16) (17) VARIABLES sick_30d med_30d attend_pct lnincpc lnexp lnfoodexp fies food_security num_kids0005 -0.0510*** 0.0132 0.0446** 0.00822 -0.0988*** -0.0965*** -0.122 0.0629*** (0.0139) (0.0206) (0.0199) (0.00514) (0.0230) (0.0248) (0.145) (0.0192) num_kids0614 -0.0552*** 0.00660 0.0906*** 0.00820* -0.0735*** -0.0680*** -0.170 0.0374** (0.0137) (0.0204) (0.0192) (0.00473) (0.0219) (0.0237) (0.137) (0.0181) num_1564 -0.0584*** 0.00153 0.0283 0.00983** -0.0214 -0.0290 -0.208 0.0139 (0.0132) (0.0200) (0.0197) (0.00492) (0.0224) (0.0241) (0.134) (0.0187) access_elec 0.0108 0.0548** 0.164*** 0.0243*** 0.329*** 0.272*** -0.922*** -0.0816*** (0.0198) (0.0271) (0.0274) (0.00811) (0.0302) (0.0330) (0.145) (0.0134) access_water -0.0200 0.0187 0.0228 0.0547*** 0.231*** 0.149*** -0.619*** -0.0214 (0.0208) (0.0304) (0.0288) (0.0107) (0.0374) (0.0383) (0.193) (0.0154) type_sanitation -0.0233 0.0191 -0.0111 0.0748*** 0.242*** 0.155*** -0.969*** 0.0238 (0.0199) (0.0327) (0.0300) (0.0154) (0.0395) (0.0425) (0.175) (0.0218) landsize 0.00387** 0.00191 0.00133** 0.00987*** 0.0106*** -0.0481*** -0.00678*** (0.00192) (0.00256) (0.000536) (0.00294) (0.00339) (0.0147) (0.00209) use_pesti 0.00150 0.00760 0.00973* 0.0722*** 0.0797*** 0.252 -0.0594** (0.0165) (0.0263) (0.00521) (0.0273) (0.0308) (0.168) (0.0235) use_fer 0.00543 -0.0286 0.0136** 0.120*** 0.140*** -0.440*** -0.0283 (0.0156) (0.0242) (0.00529) (0.0263) (0.0294) (0.152) (0.0229) livestock -0.00143 -0.00235 0.000767** 0.0111*** 0.0132*** -0.0720*** -0.00721*** (0.000911) (0.00145) (0.000333) (0.00154) (0.00176) (0.00839) (0.00132) Constant 0.421*** 0.659*** 0.289*** 0.141*** 12.89*** 12.32*** 3.461*** -0.151*** (0.0275) (0.0442) (0.0393) (0.0125) (0.0477) (0.0536) (0.265) (0.0351) Observations 6,022 4,915 4,584 6,022 6,022 6,022 6,022 6,022 R-squared 0.064 0.032 0.224 0.301 0.494 0.332 0.160 0.146 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 132 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Figure 3. Household income and education level by access to infrastructure 1,800.000 100% 1,600.000 90% 1,400.000 80% 70% 1,200.000 60% 1,000.000 50% 800.000 40% 600,000 30% 400.000 20% 10% 200,000 0% - Access to No access to Toilet No toilet Access to No access to Toilet No toilet electricity electricity electricity electricity No education Primary Secondary Tertiary Source: Author’s calculation The primary and secondary school enrollment rate is fertilizer and pesticide increase agriculture productivity. defined for the 2017/2018 school year. None of the shocks Experiencing drought in the past three years decreases have significant relationship with enrolment rate. Having total consumption and food consumption by 10% and 11% access to electricity significantly increases enrolment by respectively. Access to infrastructures continues to show 21.9%. Water is associated with 10 percent increase in a strong positive impact. Female-headed household is a enrollment. In fact, across all regressions, having electricity negative factor for annual income and total consumption, and water makes people better off. though not significantly. Drought decreases income per capita by 1 percentage All three shocks are shown to have a significant negative point. No statistically significant relationships are found impact on food security. If a household experiences death between income and other shocks; perhaps because of family member in the past 3 years, its FIES will increase income includes non-agriculture income like livestock, by 0.5 units, meaning more severe food insecurity. If this salary, and enterprise. Take livestock, for instance, household also reports conflict and drought shocks, its the rainfall over multiple seasons is important. Post- FIES would further increase 2. Drought also increases secondary education increases annual income per capita food insecurity rate by 6 percentage points. Female- and by 25% compared with no education. Secondary education old-headed households experience more severe food increases annual income per capita by 4% compared with insecurity. On the other hand, higher education help ease no education. Therefore, better education helps income the problem (Ojogho, 2010). Furthermore, owning livestock growth (Gyimah-Brempong, Paddison, & Mitiku, 2006) significantly decreases the level of food insecurity by (Jamison, Jamison, & Hanushek, 2007). Moreover, fertilizer providing households with additional means to cope with use is positively correlated with income potentially shocks and mitigate impacts. because about 57% of income is agriculture income and NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 133 Table 3 shows the results of the full model which explores through its impact on rural households. Severe drought different impacts in rural and urban areas. The impact in both rural and urban areas affect health negatively. It of family member death on the probability of people decreases the likelihood of going to hospital when sick in feeling sick in the past 30 days is mainly driven by the rural and urban areas by 13% and 11%, respectively. The impact on rural households. (Watts, 1987) already shows negative impact of severe drought on total consumption the linkage between drought and higher food insecurity. is mainly from its urban impact. Drought shock would increase FIES further by 0.7 in the rural area. Drought has bigger impact on rural household’s Tables 6-8 show results using alternative conflict per capita food consumption, being 3% more. Conflict indicators derived from ACLED data. If conflict is defined generally decreases all dimensions of people’s welfare as the number of fatalities within 15km and 20 km of (Ibáñez & Vélez, 2008). The probability of people going to households, it will decrease school enrollment rate by the hospital after feeling sick is lowered by 8 percentage 2.1 percentage points and 1.9 percentage points in rural points in the rural area. FIES is further reduced by 1.4 in area. If conflict is defined as the number of fatalities rural area and 1 in urban area by conflict. within 25km of households, stronger impact is found on school enrollment rate. Health shock and drought are Alternative drought indicators are used and the results only significantly correlated with higher FIES and are presented in Table 4 and 5. In general, the negative lower consumptions. impacts of drought on consumptions is primarily working 134 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 5. CONCLUSION financing totaling $35 million for the Urban Water and Sanitation Project (PEAMU) helped more than 893,500 people gain access to potable water, 52% of whom were This paper contributes to the current literature by women. Moreover, 67,560 pupils of the 86,940 targeted by focusing on household vulnerability to both covariant and the project had access to an improved sanitation system idiosyncratic shocks and studying the impact of shocks in their schools, and 708 latrines were constructed for on people’s wellbeing. The results show that the most teachers.1 Based on this paper’s finding that access to commonly reported shocks in the past three years are these basic infrastructure increases welfare outcomes. drought for rural households and illness/death of family This support should be continued. Additionally, higher members for urban households. The commonly reported education level of household head help eases the drought drought shock may be a result of a bad rainfall year during impact on income, increasing investment in education is 2015-2017. Although conflict shocks are less frequently thus another important part of a plan to make Niger more reported, its impact on people’s wellbeing has grown. The resilient in the face of an increasingly uncertain world. conflict shocks are characterized by battle, explosions, and violence against civilians in Diffa and Tillaberi, and peaceful protests in Zinder and Niamey. Moreover, low- For more than 15 years, the World Bank has supported income people are the most vulnerable to health shocks. the Government of Niger’s efforts to the development The nature and prevalence also vary between regions of several sectors. As of August 2019, $0.45 billion and and household’s characteristics. Rural households suffer $90 million have been invested in water and sanitation more from weather and crop disease-related shocks. and education. The efforts put in education have Households are also more vulnerable with the increase of helped construct new classrooms, improve the quality household size. of teaching, and increase pupil retention rates. In June 2016, financing totaling $35 million for the Urban Water and Sanitation Project (PEAMU) helped more than 893,500 The findings of this paper also highlight the negative people gain access to potable water, 52% of whom were impact of shocks on education, health, food insecurity, and women. Moreover, 67,560 pupils of the 86,940 targeted by income. Drought is found to have a significant negative the project had access to an improved sanitation system impact on consumption. Health shock deteriorates in their schools, and 708 latrines were constructed for people’s health status and makes food insecurity more teachers. Based on this paper’s finding that access to severe. Finally, these impacts are larger in rural areas. these basic infrastructure increases welfare outcomes. For more than 15 years, the World Bank has supported This support should be continued. Additionally, higher the Government of Niger’s efforts to the development education level of household head help eases the drought of several sectors. As of August 2019, $0.45 billion and impact on income, increasing investment in education is $90 million have been invested in water and sanitation thus another important part of a plan to make Niger more and education. The efforts put in education have helped resilient in the face of an increasingly uncertain world. construct new classrooms, improve the quality of teaching, and increase pupil retention rates. In June 2016, 1 The World Bank. https://www.worldbank.org/en/country/niger/overview#3 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 135 Table A5.4. Regression Results – full model; alternative “drought” based on CHIRPS (1) (2) (3) (4) (5) (6) (7) (8) VARIABLES sick_30d med_30d attend_pct lnincpc lnexp lnfoodexp fies food_security death_hh_rural 0.0659*** 0.0149 -0.0164 -0.00666 0.0177 0.0344 0.542*** -0.0337 (0.0214) (0.0292) (0.0284) (0.00441) (0.0269) (0.0313) (0.206) (0.0261) death_hh_urban -0.000967 -0.0644 -0.0134 0.00690 -0.0623 -0.110 0.0606 0.0496 (0.0290) (0.0474) (0.0420) (0.0178) (0.0614) (0.0702) (0.270) (0.0400) urban_drought -0.0294 -0.0615 0.0108 -0.0242* -0.164*** -0.118** 0.262 0.0421 (0.0313) (0.0432) (0.0461) (0.0139) (0.0524) (0.0532) (0.280) (0.0306) rural_drought -0.0301 -0.138*** 0.000423 0.00115 0.0518 0.0818** -0.299 -0.0215 (0.0209) (0.0393) (0.0332) (0.00595) (0.0316) (0.0364) (0.221) (0.0314) conflict_rural -0.0805** 0.0448 0.0216 0.00172 -0.0432 -0.148* 1.269*** 0.0577 (0.0372) (0.0649) (0.0654) (0.00954) (0.0504) (0.0757) (0.352) (0.0530) conflict_urban -0.0315 0.156*** -0.123* -0.00910 0.0563 0.0725 0.858 -0.0248 (0.0419) (0.0513) (0.0710) (0.0204) (0.0755) (0.0786) (0.524) (0.0330) Table A5.5. Regression Results – full model; alternative “sdrought” based on CHIRPS (9) (10) (11) (12) (13) (14) (15) (16) VARIABLES sick_30d med_30d attend_pct lnincpc lnexp lnfoodexp fies food_security death_hh_rural 0.0663*** 0.0162 -0.0165 -0.00663 0.0174 0.0342 0.543*** -0.0335 (0.0215) (0.0292) (0.0284) (0.00442) (0.0269) (0.0313) (0.206) (0.0261) death_hh_urban -0.000628 -0.0619 -0.0145 0.00653 -0.0662 -0.112 0.0674 0.0510 (0.0291) (0.0472) (0.0420) (0.0178) (0.0607) (0.0696) (0.271) (0.0398) urban_sdrought -0.110*** -0.111* 0.0228 -0.0394** -0.167*** -0.205*** 0.354 0.0401 (0.0300) (0.0602) (0.0551) (0.0184) (0.0609) (0.0625) (0.395) (0.0385) rural_sdrought -0.0585** -0.129** -0.0575 0.00789 0.141*** 0.111** -0.485 -0.0325 (0.0227) (0.0539) (0.0482) (0.00992) (0.0484) (0.0534) (0.301) (0.0308) conflict_rural -0.0780** 0.0519 0.0181 0.00315 -0.0350 -0.146* 1.267*** 0.0566 (0.0370) (0.0661) (0.0654) (0.00961) (0.0508) (0.0759) (0.352) (0.0529) conflict_urban -0.0292 0.174*** -0.124* -0.0116 0.0319 0.0448 0.940* -0.0166 (0.0412) (0.0516) (0.0709) (0.0198) (0.0737) (0.0770) (0.517) (0.0323) 136 NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION Table 6. Regression Results – full model; logged # of fatalities within 15km (1) (2) (3) (4) (5) (6) (7) (8) VARIABLES sick_30d med_30d attend_pct lnincpc lnexp lnfoodexp fies food_security death_hh_rural 0.0628*** 0.0140 -0.0179 -0.00729 0.0142 0.0302 0.563*** -0.0345 (0.0212) (0.0290) (0.0281) (0.00444) (0.0270) (0.0314) (0.205) (0.0260) death_hh_urban -0.00399 -0.0632 -0.00899 0.00535 -0.0786 -0.125* 0.108 0.0545 (0.0291) (0.0474) (0.0419) (0.0182) (0.0618) (0.0706) (0.274) (0.0399) drought_rural -0.00627 0.0255 0.0149 -0.00806** -0.0900*** -0.0950*** 0.714*** 0.0550*** (0.0114) (0.0176) (0.0162) (0.00361) (0.0176) (0.0211) (0.106) (0.0163) drought_urban -0.0746* -0.000837 0.138*** -0.0172 -0.253*** -0.221** 0.663 0.0759 (0.0420) (0.0696) (0.0527) (0.0258) (0.0663) (0.0858) (0.511) (0.0520) urban_lf15 -0.00488 0.0105 -0.00683 -0.0101 -0.0208 -0.0257 0.181 -0.00160 (0.0128) (0.0177) (0.0163) (0.00711) (0.0170) (0.0198) (0.121) (0.00922) rural_lf15 -0.0260*** -0.0251** -0.0194* -0.00162 0.00618 -0.00455 -0.00502 -0.0229** (0.00779) (0.0122) (0.0111) (0.00208) (0.0116) (0.0137) (0.0722) (0.0103) Table 7. Regression Results - full model; logged # of fatalities within 20km (9) (10) (11) (12) (13) (14) (15) (16) VARIABLES sick_30d med_30d attend_pct lnincpc lnexp lnfoodexp fies food_security death_hh_rural 0.0625*** 0.0143 -0.0182 -0.00718 0.0143 0.0299 0.563*** -0.0347 (0.0212) (0.0290) (0.0281) (0.00444) (0.0270) (0.0314) (0.205) (0.0261) death_hh_urban -0.00423 -0.0628 -0.00924 0.00520 -0.0789 -0.125* 0.111 0.0544 (0.0291) (0.0474) (0.0420) (0.0182) (0.0618) (0.0705) (0.274) (0.0399) drought_rural -0.00651 0.0251 0.0148 -0.00812** -0.0899*** -0.0950*** 0.714*** 0.0548*** (0.0114) (0.0176) (0.0162) (0.00362) (0.0176) (0.0211) (0.106) (0.0163) drought_urban -0.0752* -0.000409 0.137*** -0.0174 -0.253*** -0.221** 0.665 0.0757 (0.0421) (0.0696) (0.0527) (0.0258) (0.0663) (0.0858) (0.511) (0.0520) urban_lf20 -0.00887 0.0151 -0.00568 -0.0120* -0.0207 -0.0249 0.185 -0.00302 (0.0129) (0.0178) (0.0163) (0.00720) (0.0170) (0.0198) (0.121) (0.00919) rural_lf20 -0.0268*** -0.0206* -0.0214** -0.000788 0.00656 -0.00586 -0.000706 -0.0224** (0.00763) (0.0120) (0.0109) (0.00205) (0.0115) (0.0135) (0.0711) (0.0101) NIGER: INVESTING IN HUMAN CAPITAL, AGRICULTURAL PRODUCTIVITY, AND SOCIAL PROTECTION FOR FASTER POVERTY REDUCTION 137 Table 8. Regression Results - full model; logged # of fatalities within 25km (17) (18) (19) (20) (21) (22) (23) (24) VARIABLES sick_30d med_30d attend_pct lnincpc lnexp lnfoodexp fies food_security death_hh_rural 0.0627*** 0.0146 -0.0183 -0.00712 0.0140 0.0300 0.563*** -0.0343 (0.0212) (0.0290) (0.0281) (0.00446) (0.0270) (0.0314) (0.205) (0.0260) death_hh_urban -0.00552 -0.0622 -0.0104 0.00552 -0.0800 -0.126* 0.120 0.0543 (0.0291) (0.0474) (0.0419) (0.0181) (0.0617) (0.0703) (0.273) (0.0398) drought_rural -0.00679 0.0249 0.0147 -0.00812** -0.0898*** -0.0950*** 0.714*** 0.0545*** (0.0114) (0.0176) (0.0162) (0.00363) (0.0176) (0.0211) (0.106) (0.0163) drought_urban -0.0762* 8.90e-05 0.137*** -0.0155 -0.253*** -0.220** 0.666 0.0764 (0.0422) (0.0696) (0.0529) (0.0256) (0.0663) (0.0857) (0.511) (0.0520) urban_lf25 -0.0216* 0.0144 -0.0160 0.00531 -0.0199 -0.0197 0.184 -0.00129 (0.0127) (0.0176) (0.0161) (0.00656) (0.0165) (0.0193) (0.119) (0.00894) rural_lf25 -0.0266*** -0.0188 -0.0238** 0.000213 0.00479 -0.00514 -0.000816 -0.0200** (0.00733) (0.0117) (0.0105) (0.00199) (0.0110) (0.0130) (0.0690) (0.00987) World Bank Poverty and Equity Global Practice, Africa Region