SOMALIA POVERTY AND EQUITY ASSESSMENT © 2024 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 the 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. Attribution—Please cite the work as follows: “World Bank. 2024. Somalia Poverty and Equity Assessment. Washington DC. © World Bank.” Photo Credits: • Front cover: SOMALIA - JANUARY 13, 2010: In the Somali village. Free Wind 2014/Shutterstock • Page xi: Mohamed Abdihakim • Page xii: Camel market in Somalia Martinez de la Varga/Shutterstock • Page xvi: Mohamed Abdihakim • Page 29: Mohamed Abdihakim CONTENTS ACKNOWLEDGEMENTS .......................................................................................................................... v ABBREVIATIONS ........................................................................................................................................ vi FIGURES .......................................................................................................................................................... vii TABLES ............................................................................................................................................................ x EXECUTIVE SUMMARY ............................................................................................................................ xi Part A ............................................................................................................................................................ xii Part B ............................................................................................................................................................ xiii Livelihoods ............................................................................................................................................. xiii Shocks ...................................................................................................................................................... xiv Nomadic .................................................................................................................................................. xv Part C ................................................................................................................................................................ xvi PART A: CORE ANALYTICS AND CROSS-COUNTRY BENCHMARKING .............................. 1 CHAPTER 1: THE INCIDENCE, NATURE, AND EVOLUTION OF POVERTY IN SOMALIA .. 1 Introduction .............................................................................................................................................. 1 Poverty remains high in Somalia, with the bulk of the poor living in urban areas ........ 3 Monetary poverty trends ..................................................................................................................... 5 Who are the poor? ................................................................................................................................... 9 Non-monetary poverty is more prevalent than monetary poverty .................................... 11 Poverty and Access to Services and Amenities ....................................................................... 11 Inequality in consumption remains relatively low, despite high inequality in opportunities ....................................................................................................................................... 15 Inequality of opportunity ..................................................................................................................... 15 The remainder of this poverty assessment will focus on three deep-dive topics ......... 16 PART B: DEEP-DIVES ............................................................................................................................. 18 CHAPTER 2: LIVELIHOOD DEEP-DIVE ........................................................................................... 18 Why is Somalia’s Labor Force Participation so low? ................................................................... 19 Limited agricultural activity and little non-agricultural labor demand? ....................... 19 The combination of limited opportunities results in most individuals working in low-return activities out of necessity ......................................................................................... 20 Dualistic Wage Employment ............................................................................................................... 23 Household enterprises: Can they contribute to poverty reduction? .................................. 26 iii SOMALIA POVERTY AND EQUITY ASSESSMENT How can the poor be more productive in activities they undertake and what prevents them from accessing productive economic opportunities? ................................ 28 CHAPTER 3: SHOCKS DEEP-DIVE ....................................................................................................... 30 What shocks are commonly reported by Somali households? .............................................. 30 How are different areas of the country affected by climatic shocks? ............................... 32 Did poor households more often report a negative impact of drought? ......................... 33 Who are vulnerable to climate shocks? .......................................................................................... 34 How Somali households cope with climatic shocks? ................................................................. 36 The relationship between displacement and climatic shocks ........................................... 37 How can households be more resilient to climatic shocks? ................................................... 38 CHAPTER 4: NOMADIC DEEP-DIVE ................................................................................................... 41 What are the welfare conditions of the nomadic population? ............................................. 42 What enables these richer nomadic households to achieve higher consumption? ...... 43 Larger herd sizes? ................................................................................................................................ 43 Commercialization .............................................................................................................................. 45 The importance of location? ........................................................................................................... 46 What can the rest learn from the richest nomadic households? .......................................... 48 PART C: POLICY RECOMMENDATIONS ........................................................................................... 50 References ....................................................................................................................................................... 55 Gender Annex ............................................................................................................................................. 62 What works in other countries .......................................................................................................... 64 Annex: Chapter 1 ....................................................................................................................................... 66 Working-Age and Household Size ................................................................................................ 66 Multidimensional Poverty Definition .......................................................................................... 67 Human Opportunity Index................................................................................................................ 68 Annex: Chapter 2 ...................................................................................................................................... 70 Labor Force Participation Regressions ........................................................................................... 70 Characteristics of “Better Jobs” .................................................................................................. 74 Household Enterprises ...................................................................................................................... 75 Potential Impact of Urban Public Works .................................................................................... 79 Annex: Chapter 3 ...................................................................................................................................... 81 Annex: Chapter 4 ...................................................................................................................................... 82 Livestock Revenue Regression ...................................................................................................... 82 iv ACKNOWLEDGEMENTS ACKNOWLEDGEMENTS This report was prepared by a World Bank team The report was prepared under the guidance of jointly led by Aphichoke Kotikula and Alastair Peter Pierella Paci (Practice Manager), Tom Bundervoet Francis Haynes, comprising Matthieu Guillaume (Lead Economist), Keith Hansen (Country Director Emmanuel Rouyer, Marie Christelle Mabeu, Shinya for Kenya, Rwanda, Somalia, and Uganda), Kristina Takamatsu, and Takaaki Masaki. The World Bank Svensson (Country Manager, Somalia), Matthias peer reviews were Obert Pimhidzai and Ruth Hill. Mayr (Senior Operations Officer, Somalia), and The report also benefitted from excellent comments Marek Hanusch (Program Leader for Kenya, from Pedro Cerdan-Infantes and Verena Phipps. Rwanda, Somalia, and Uganda). The team also benefited from the peer review of the poverty measurement and survey-to-survey The team is grateful for the financial support from methodology from Nduati Maina Kariuki and Oscar the Somalia Multi-Partner Fund and the Climate Eduardo Barriga Cabanillas. The team thanks Tom Support Facility Whole-of-Economy Program. The Bundervoet for feedback and comments throughout World Bank greatly appreciates the collaboration the process. The team would also like to thank with the Somalia National Bureau of Statistics (SNBS). everyone in the World Bank Somalia country team The work builds on the Poverty Report published by who provided feedback on earlier report drafts. SNBS and uses the Somalia Integrated Household Budget Survey (SIHBS). The report benefitted from Administrative and logistical support during the inputs from officials of the Ministry of Planning, preparation of this report was provided by Angela Investment and Economic Development, the Wangari Mwangi, Christine Khasiro Wesakania, and National Economic Council, and the SNBS. The team Martin Buchara and is gratefully acknowledged. also benefitted from the qualitative work undertaken The report was designed by Cybil Maradza. by the Heritage Institute, led by Uweis Ali. v SOMALIA POVERTY AND EQUITY ASSESSMENT ABBREVIATIONS COVID-19 Coronavirus disease 2019 DHS Demographic and Health Surveys FCV Fragility, conflict, and violence FEWS NET Famine Early Warning Systems Network FSNAU Food Security and Nutrition Analysis Unit - Somalia GDP Gross domestic product FAO Food and Agriculture Organization of the United Nations FGDs Focus Group Discussions HHEs Household enterprises HIPC Heavily Indebted Poor Countries Initiative HOI Human Opportunity Index IDP Internally displaced person IMF International Monetary Fund IPC Integrated Food Security Phase Classification kcal/person/day Kilocalories per person per day km Kilometers LFP Labor force participation LICs Low-income countries MENA Middle East and North Africa NDVI Normalized difference vegetation index NGOs Non-governmental organizations PRMN Protection & Return Monitoring Network SCD Systematic Country Diagnostic SIHBS Somali Integrated Household Budget Survey SHFPS Somali High Frequency Phone Survey SHFS-W2 Somali High Frequency Survey-Wave 2 SNBS Somalia National Bureau of Statistics SOMPA Somali Poverty and Equity Assessment SWIFT Survey of Well-being via Instant and Frequent Tracking SSA Sub-Saharan Africa TLU Tropical livestock unit UNFPA United Nations Population Fund US United States USD United States dollar WDI World Development Indicators WBGT Wet-bulb globe temperature vi FIGURES FIGURES Figure 1: Assets approach to market income ............................................................................................ 3 Figure 2: Poverty Indicators ............................................................................................................................. 3 Figure 3: Regional Poverty Map, 2022 ......................................................................................................... 4 Figure 4: Share of Poor by Region, 2022 ..................................................................................................... 4 Figure 5: Population Shares, 2022 ................................................................................................................. 5 Figure 6: Urbanization, 2022 ........................................................................................................................... 5 Figure 7: Extreme Poverty Rates ................................................................................................................... 6 Figure 8: Poverty Trends, 2017 to 2022 ...................................................................................................... 6 Figure 9: Extreme Poverty Trends, 2017 to 2022 .................................................................................... 6 Figure 10: Annual Per capita consumption growth rate, 2017-2022 .................................................. 7 Figure 11: Poor population measured by the national poverty line, 2017-2022 ............................ 8 Figure 12: Poverty gap rates using the national poverty line, 2017-2022 ........................................ 8 Figure 13: Growth-Redistribution Decomposition, 2017 to 2022 ...................................................... 9 Figure 14: Correlates of Poverty by Residency ........................................................................................... 10 Figure 15: Fertility rates among women between 15 to 49, 2020 ....................................................... 11 Figure 16: Percent of live births in last 5 years delivered by skilled provider, 2020 ...................... 11 Figure 17: Literacy Levels, 2022, 15+ .............................................................................................................. 12 Figure 18: Literacy Levels by Region ............................................................................................................... 12 Figure 19: Gross Enrollment ............................................................................................................................... 12 Figure 20: Primary Gross Enrollment by Region ......................................................................................... 12 Figure 21: Access to Electricity .......................................................................................................................... 13 Figure 22: Change in Access to Electricity, 2017 to 2022 ........................................................................ 13 Figure 23: Access to Improved Drinking Water ........................................................................................... 13 Figure 24: Trend in Improved Drinking Water, 2017 to 2022 ................................................................ 13 Figure 25: Chronic Poor by Region .................................................................................................................. 14 Figure 26: Share of Individuals by monetary and non-monetary poverty status ............................ 14 Figure 27: Gini Index, 2022 ................................................................................................................................. 15 Figure 28: Change in Gini coefficient, 2017-2022 ...................................................................................... 15 Figure 29: Shapely Decomposition of Each Circumstance, 2022 .......................................................... 16 Figure 30: Somalia Population Breakdown, 2022 ...................................................................................... 19 Figure 31: International Comparison of Labor Force Participation ..................................................... 20 Figure 32: Labor Force Participation, 2022 .................................................................................................. 20 Figure 33: LIC and MENA Agriculture Share of Employment ................................................................. 20 Figure 34: Reason for Not Searching for Work Despite Wanting to Work Among the Inactive ..... 20 Figure 35: Labor Force Participation by Sex and IDP Status .................................................................... 21 vii SOMALIA POVERTY AND EQUITY ASSESSMENT Figure 36: Employment Type by Sex and IDP Status ................................................................................. 21 Figure 37: Sector by Sex and IDP Status ........................................................................................................ 21 Figure 38: Employer Type by Sex and IDP Status ....................................................................................... 21 Figure 39: Share of Remittances sent to… ................................................................................................... 22 Figure 40: Enrollment by Remittances ........................................................................................................... 22 Figure 41: Share of Households with Revenue Sources ........................................................................... 23 Figure 42: Median Annual Household Income from Wages ................................................................... 23 Figure 43: LIC and MENA Wage Share of Employment ............................................................................ 23 Figure 44: Ratio of Wage Employment to Working-Age Population ................................................... 23 Figure 45: Share of Wage Employment by Sector ..................................................................................... 25 Figure 46: Share of Employment and Share of Jobs that are Better-Quality Wage Jobs by Employer ........................................................................................................................................ 25 Figure 47: Share of Households Receiving… ............................................................................................... 25 Figure 48: Share of Recipients by Poverty Status ....................................................................................... 25 Figure 49: Share of Households with an Enterprise .................................................................................. 26 Figure 50: Household Enterprise Owner Gender and Education ......................................................... 26 Figure 51: Share of Household Enterprises that report enough per capita profit for the household to be above the poverty line ................................................................................... 27 Figure 52: Household Enterprise Operating Location .............................................................................. 27 Figure 53: Required Job Creation, 2023-2030 ............................................................................................ 28 Figure 54: Exposure to Conflict, 2018-2022 ................................................................................................ 31 Figure 55: Regional Poverty Rate and Exposure to Conflict .................................................................. 31 Figure 56: Self-Reported Household Exposure to Conflict in 2021 or 2022 .................................... 32 Figure 57: Poverty Rate ....................................................................................................................................... 32 Figure 58: Share of Population Exposed to Any Climate Hazard .......................................................... 32 Figure 59: Share of Population Exposed to Drought ................................................................................ 33 Figure 60: Share of Population Exposed to Heat ....................................................................................... 33 Figure 61: Share of Population Exposed to Floods .................................................................................... 33 Figure 62: Share of Households Affected by Drought by Residency and Consumption Quintile, 2022 ..................................................................................................................................... 33 Figure 63: Poverty Headcount by Region and Share of Households Affected by Drought, 2022 .................................................................................................................................... 33 Figure 64: Share of Exposed Population that are Vulnerable in 1 Areas ........................................... 35 Figure 65: Regional Poverty Rates and Share of Exposed Population that are Vulnerable in 3 Areas .............................................................................................................................................. 35 Figure 66: Average Number of Vulnerable Dimensions by Region ...................................................... 36 Figure 67: Share of Households Vulnerable in each dimension ............................................................ 36 viii FIGURES Figure 68: Share of Households Responding to Drought, 2022 ........................................................... 36 Figure 69: Share of Households with Economic Responses to Drought by Residency and Consumption Quintile, 2022 ........................................................................................................ 36 Figure 70: Share of Households with Common Maladaptive Responses to Drought by Household Characteristics, 2022 ................................................................................................. 37 Figure 71: Displaced Individuals, 2016-2023 ................................................................................................ 38 Figure 72: Reason for Being an IDP ................................................................................................................. 38 Figure 73: Poverty Headcount and Gap ......................................................................................................... 38 Figure 74: Distribution of IDPs .......................................................................................................................... 38 Figure 75: Poverty Rates ..................................................................................................................................... 39 Figure 76: Access to the Internet ..................................................................................................................... 39 Figure 77: Primary Gross Enrollment .............................................................................................................. 39 Figure 78: Density .................................................................................................................................................. 40 Figure 79: Distance ................................................................................................................................................ 40 Figure 80: Climate .................................................................................................................................................. 40 Figure 81: Land Use Systems ............................................................................................................................. 41 Figure 82: Unweighted regional NDVI ............................................................................................................ 43 Figure 83: Correlation between Nomadic Poverty and NDVI Deviation ............................................ 43 Figure 84: Median Livestock and TLU Ownership ...................................................................................... 44 Figure 85: Nomadic Households by TLU per capita ................................................................................... 44 Figure 86: Type of Pastoralist ............................................................................................................................ 45 Figure 87: Pastoralist Type based on per capita income and TLU ........................................................ 45 Figure 88: Ratio of Livestock Value to Annual Consumption ................................................................. 45 Figure 89: Livestock Revenue Sources ........................................................................................................... 46 Figure 90: Average and Median Livestock Revenue .................................................................................. 46 Figure 91: Share of Households with Livestock Selling an Animal for Slaughter in the last 12 months ................................................................................................................................... 47 Figure 92: Percentage Deviation from the Long-Term NDVI in June 2022 ....................................... 47 Figure 93: Percentage Change in Median TLU from 12 months prior to the survey to date of interview ......................................................................................................................................... 48 Figure 94: Change in Average TLU per capita Ownership ....................................................................... 48 Figure 95: Mortality and Birth Rates by Drought Status .......................................................................... 49 Figure 96: Share of Livestock Owning Households with Expenditure on… ....................................... 49 Figure 97: Labor Force Status by Gender and Poverty ............................................................................... 62 Figure 98: Unemployment and Underemployment by Gender .............................................................. 62 Figure 99: Type of Employer by Gender and Poverty ................................................................................. 63 ix SOMALIA POVERTY AND EQUITY ASSESSMENT Figure 100: Type of Employment by Gender and Poverty ...................................................................... 63 Figure 101: International Comparison in Access to Electricity .............................................................. 67 Figure 102: Coverage and Human Opportunity Index, 2022 ................................................................ 69 Figure 103: D-Index, 2022 .................................................................................................................................. 69 Figure 104: Impact on the Poverty Gap ......................................................................................................... 80 Figure 105: Nomadic Poverty Rates by Region .......................................................................................... 82 Figure 106: Nomadic Population Share in each Region ........................................................................... 82 TABLES Table 1: Change in Annual Average Consumption .................................................................................. 7 Table 2: Decomposition of Change in Poverty Rate into Intra-population group and population shifts, 2017-2022 ......................................................................................................... 9 Table 3: Monetary, Multidimensional, and Chronic Poverty ................................................................ 14 Table 4: Regression Coefficients for NDVI Shock Z-Score on Monetary and Non-Monetary Indicators ................................................................................................................. 34 Table 5: Indicators to Measure Climate Vulnerability ............................................................................. 35 Table 6: Classification of Pastoralist Type .................................................................................................. 44 Table 7: Policy Recommendations ................................................................................................................ 51 Table 8: Household Size, Working Age, and the Ratio of Working-Age to Members ................. 66 Table 9: Multidimensional Poverty Definition ........................................................................................... 67 Table 10: Definition and reference groups for various opportunities ................................................ 68 Table 11: Labor Force Participation Regression, 2022 ............................................................................. 70 Table 12: Labor Force Participation Regression including perception of safety, 2022 ................ 72 Table 13: Characteristics of Better jobs ......................................................................................................... 74 Table 14: Marginal Effects from Probit Regression on Household Having an Enterprise ........... 75 Table 15: Determinants of Having an Enterprise and its impact on household welfare .............. 76 Table 16: Exploring the direct effect of household enterprise on household welfare ................ 77 Table 17: Household Enterprise Regressions .............................................................................................. 78 Table 18: Self-Reported Exposure to Shocks ............................................................................................... 81 Table 19: Poverty and 3 D's Regression ......................................................................................................... 81 Table 20: Livestock Revenue Regression ....................................................................................................... 82 x EXECUTIVE SUMMARY xi SOMALIA POVERTY AND EQUITY ASSESSMENT PART A Somalia has made macroeconomic progress in 2017 and 2022. However, while the urban poverty recent years; however, the economy remains rate did not change, poverty increased in rural and exposed to shocks, particularly climatic shocks. nomadic areas. The movement of population from The economy has shown signs of recovery in recent nomadic to urban areas, or rather from high poverty years after exposure to many shocks. In December to low poverty, countered the negative consumption 2023, Somalia achieved a historic HIPC completion growth and poverty increase in rural and nomadic point, with external debt falling to 6 percent of GDP areas. Without the population movement, poverty in 2023, and in March 2024, Somalia joined the East would have increased by around 2 percentage African Community. However, growth has been points, largely due to increased rural poverty. insufficient to increase GDP per capita, while the economy remains dependent on imports for basic Household size, IDP status, education, having a commodities. The country continues to work towards household enterprise, and receiving remittances a political settlement and still faces high levels of are all associated with poverty. Certain demographic insecurity and high exposure to climatic shocks. variables are associated with higher poverty rates, such as being an IDP, having a larger household size, Poverty remains high, especially among those and having a larger share of children. In contrast, outside of urban areas. Over half of the population the household head's education is associated with lives below the national poverty line (54%), with the lower poverty levels, especially higher levels of highest poverty rate among the nomadic population education, as is the self-reported literacy level of (78%) and the lowest rate in urban areas (46%). the household head. The presence of a household However, given the high rate of urbanization, poverty enterprise or remittances from abroad are also is concentrated in urban areas. Spatially, poverty is associated with lower poverty. In contrast, in higher in regions in Central and Southern Somalia. rural and nomadic areas, having a wage earner in the household is associated with higher poverty. Despite successive shocks, poverty remained Finally, location also shapes the spatial pattern of unchanged between 2017 and 2022. In line with poverty; areas with lower population density and the stagnation of GDP per capita, the national limited transportation access tend to experience poverty rate has also remained unchanged between higher poverty. xii EXECUTIVE SUMMARY Non-monetary poverty is even higher than driven by region and less so by poverty status. monetary poverty, with very low education Further, while there is limited difference in primary attainment and enrollment. Over three-quarters school enrollment by gender, this gap widens for of the population are considered non-monetary secondary schooling. poor, ranging from 73% in urban areas to almost universal among the nomadic population (95%). The Somalia Poverty and Equity Assessment This is largely driven by deprivations in education, will focus on three deep-dive topics. The first will followed by sanitation and flooring. Just under half focus on Somali livelihoods, given its importance the population is considered chronic poor (47%), for sustainable poverty reduction. It will look at i.e., both monetary and non-monetary poor, with the type of income, type of employment, sector the highest rates among the nomadic population of employment, and household enterprises. (74%). Only 15 percent of the Somali population is The second deep dive topic will look at shocks, neither monetary nor non-monetary poor. particularly climate shocks. It will focus on who is exposed, who is vulnerable, and what households Education has the highest inequality in access, typically do in response. Resilience to these with regional differences driving inequalities in shocks is essential for households to move out of opportunity. Secondary education has the largest poverty sustainably. The last deep dive will focus inequality based on circumstances, driven by specifically on the nomadic population, given their household poverty status and region. Enrollment high poverty rates, extreme poverty, and non- in primary education is less unequal, although monetary poverty. The chapter will look at the it remains more unequal than access to services small share of non-poor nomadic households to such as electricity, water, and sanitation. However, explore what can be applied to help the poorer the inequality in primary education is largely nomadic households. PART B LIVELIHOODS The Somali labor market has some unique The limited role of agriculture and little non- features for its income level: i) very low labor agricultural labor demand likely explains the force participation (LFP), ii) a high dependency low LFP. Agricultural employment accounts for on wage employment, and iii) a large share of less than a third of all types of employment. In this employment in household enterprises. Low- regard, Somalia is more aligned with the poorest income countries (LICs) typically have a high LFP, countries in the MENA region rather than other a high share of employment in agriculture, and LICs in sub-Saharan Africa. This is likely due to a lower share of wage employment. However, the small share of viable agricultural land, which Somalia has a very low LFP, a high dependency limits the number of agricultural opportunities. on wage employment, and a lower share of Further, in contrast with LICs, Somalia has a higher employment in agriculture. In addition, while unemployment rate than underemployment. This household enterprises are relatively rare at the also suggests a lack of jobs, with over half of the household level, they account for a sizeable share economically inactive individuals who wanted to of total employment. work citing a lack of opportunities. The limited role xiii SOMALIA POVERTY AND EQUITY ASSESSMENT of agriculture and limited non-agricultural labor share of wage employment is driven by limited demand likely results in the low LFP. non-wage opportunities rather than many wage jobs. The better jobs, determined by their contract Given the limited opportunities, most status or being entitled to paid leave, only individuals have no choice but to work in account for 14 percent of all employment and 1 occupations that offer low returns. There percent of the working-age population. These appear to be two groups of workers: i) the better better jobs are dominated by men, those with educated or connected who can access the small secondary education or above, and individuals pool of good jobs, and ii) those with no alternative from the richest urban households. They are also income source and, therefore, have no choice but concentrated in NGOs, international organizations, to work in occupations that offer low returns. and government and pay better on average. Those with alternative income, such as remittances or another working member in the household, are Household enterprises (HHEs) are responsible less likely to participate in the labor force. Social for a sizeable share of employment but rarely norms, highlighted in focus group discussions, make sufficient profit to lift the household out of limit what roles individuals deem acceptable, poverty. While only 14 percent of households had a adding additional friction to the labor market. household enterprise, they accounted for nearly half of all employment. They were more often operated by While wage employment is the most important women and those without any education. However, income source, there is evidence of a dual very few HHEs make sufficient per capita profit to lift labor market. Wage employment is the most the household above the poverty line, although they common income source across the consumption positively impact household consumption. Moreover, distribution, but the income from wages women-led enterprises face challenges: they often increases with consumption. It is important to operate from the home and are less productive. note that while the wage employment share of Overall, given the current productivity levels, HHEs total employment is high, the ratio of wage jobs may be better suited to complement other income to the working-age population in Somalia is in sources, although currently they often are the main line with other LICs. This suggests that the high source of income. SHOCKS Somalia is particularly vulnerable to climatic poverty among the nomadic population. shocks, with higher exposure among poorer regions and households. Generally, Somali Almost all households exposed to climatic shocks households are highly exposed to shocks, with over are also considered vulnerable. Vulnerability to two-thirds reporting a severe negative economic climate shocks can be defined based on variables impact from a shock in 2021 or 2022. Poorer regions linked to a household’s ability to cope with shocks: i) and poorer households were more affected by the physical propensity to experience severe income, climatic shocks, with those in central and southern asset, or health loss; and ii) the inability to cope with and Somalia more affected. A drought shock is associated recover from the losses. Most exposed households are with lower consumption and higher poverty for rural also considered vulnerable, with higher rates in poorer and urban households and only higher extreme regions. Most households lack sufficient income, xiv EXECUTIVE SUMMARY making them vulnerable to climate change. Climatic shocks are a large driver of internal displacement within Somalia. Except for 2021, Households appear to lack economic options climatic shocks have been the largest driver of to respond to the drought. Almost two-thirds displacement in Somalia for the past eight years. of households who were negatively affected by IDPs have a high poverty rate, and almost one- drought used non-economic coping mechanisms third are in the poorest urban quintile. Focus in response. Richer households were more often group discussions highlight the desire to move able to use savings or sell assets. Reducing food back to their original location, but there is a need consumption and displacement were the most to replenish assets or ensure safety and stability common maladaptive responses. before they would consider doing so. NOMADIC The nomadic population has the highest households. TLU per capita increases across the monetary and non-monetary poverty rates. The consumption distribution. Richer households sold nomadic population also has the highest rate of animals for slaughter and livestock products more extreme poverty, the largest poverty gap, and the often while earning more livestock revenue. Lastly, highest rate of inequality. They also fall behind in regional differences in access to markets and terms of literacy and enrollment, access to services climatic variables may also affect the welfare of and have the highest exposure to climatic shocks. nomadic households. However, there are a small group of nomadic households, the richest quintile, who are non-poor Supporting households in accumulating larger and whose average per capita consumption exceeds herds can help with resilience and, in turn, the average for the richest rural households. increase commercialization opportunities. Very few households have a sufficient herd size Herd size, commercialization, and location may for mobility. Mobility can allow a more diverse help these non-poor households achieve higher diet, which in turn can improve livestock's health, consumption. Most nomadic households do not productivity, and resilience. Assistance to help have enough livestock per capita to lift themselves households accumulate livestock while also out of poverty. Only a quarter of households had 4.5 promoting resilience to prevent losses can, in TLU per capita or more, the threshold for mobility. turn, help households switch their attention to A further third had less than 1 TLU per capita and commercialization. Improving access to key inputs can be considered stockless or near stockless can also help improve productivity and resilience. xv SOMALIA POVERTY AND EQUITY ASSESSMENT PART C Continued economic progress and stability are for low-skilled labor. As mentioned above, the important foundations for poverty reduction. bulk of the poor reside in urban areas. Further, Somalia has made macroeconomic progress in recent the movement toward urban areas is likely to years; however, it has been insufficient in magnitude continue, given the persistent exposure to to contribute to GDP per capita growth or poverty climatic shocks and the risk of further dropping reduction. Therefore, sustained economic growth out of nomadic households. Given the extremely and stability are priorities to facilitate a conducive low levels of education, the small pool of better- environment for poverty reduction. quality and better-paid jobs will likely remain out of reach for most of the population. Therefore, Given the limited fiscal space, policy policies should focus on increasing the labor recommendations can focus on i) harnessing demand for low-skilled workers. urbanization for improved service delivery and ii) strengthening the resilience of rural/nomadic Somalia can use its high urbanization to improve livelihoods. Due to a history of climatic shocks and its human capital service delivery, especially in conflict, Somalia has a high urbanization rate for its education. In the medium term, sustained poverty income level. As a result, the poor are concentrated in reduction will need higher levels of human capital, urban areas, which can enable more efficient service particularly in education. Given most education delivery. The second area of policy recommendations is fee-based, children from poorer households can focus on strengthening the resilience of rural are often excluded, which risks intergenerational and nomadic livelihoods. These groups have higher poverty. High urbanization can help reduce the cost poverty rates than urban areas, have experienced of service delivery as the population is concentrated a decline in consumption in recent years, and have in a smaller area. Therefore, the government must greater exposure to climatic shocks. Therefore, continue its efforts to expand the school system policies that help strengthen income generation and to increase primary school enrollment, especially buffer it against shocks will be important. among the poor. For those in or moving to urban areas, public work For those who remain in rural or nomadic areas, programs may act as a positive demand shock there is a strong need to develop more resilient xvi EXECUTIVE SUMMARY livelihoods. Exposure to climatic shocks can have and resilience. Analysis of the well-being of a negative impact on rural and nomadic livelihoods, marginalized groups reveals that these groups which in turn inhibits sustained consumption tend to experience higher levels of poverty and growth and, therefore, the movement out of lag behind in terms of education and access to poverty. Policies that help strengthen rural and services. The pastoral nomadic population faces nomadic livelihoods will be important to limit the a high poverty rate and is vulnerable to climate- negative impact of these climatic shocks. Better related challenges. IDPs are 25 percentage points management of key resources such as water, poorer than their urban counterparts and also soil, and land will be key. Further, the adoption lack access to education and services. Women of climate-smart agricultural diversification can experience significant gaps in employment, with improve resilience while also potentially benefiting only 16 percent engaged in paid labor compared food security. For the nomadic population, to 41 percent of men. Men are more likely to have interventions that help provide important inputs formal wage jobs, while women often engage in such as fodder, water, and veterinary care can household enterprises and small-scale businesses, help improve the productivity and resilience of such as street vending and home-based activities, livestock. Similarly, facilitating effective rangeland to earn income. The concentration in the informal management that enables sufficient livestock sector suggests limited access to formal job mobility will be important. opportunities. When these marginalized groups are included in the mainstream economy, they Economic inclusion is essential for reducing contribute to higher productivity and innovation. poverty in Somalia. By ensuring that all segments Furthermore, economic inclusion enhances overall of society -including women, internally displaced social cohesion and stability. Thus, empowering persons (IDPs), and pastoral nomads- have equal girls, women, and minority groups through access to economic opportunities, Somalia can inclusion can help address drivers of fragility, unlock its full potential while promoting fairness conflict, and violence (FCV). xvii SOMALIA POVERTY AND EQUITY ASSESSMENT PART A: CORE ANALYTICS AND CROSS-COUNTRY BENCHMARKING CHAPTER 1: THE INCIDENCE, NATURE, AND EVOLUTION OF POVERTY IN SOMALIA Introduction Despite this headway, Somalia continues to be extremely fragile due to an unfinished political 1. The Somali economy continues to recover settlement, continued macroeconomic challenges, from the multitude of shocks in recent years. weak institutional capacity for service delivery, and GDP is estimated to have grown by 4.2 percent in communal tension. 2023 with improving weather conditions, up from 2.7 percent in 2022.¹ However, economic growth 2. Somalia continues to work towards a political only averaged 2 percent between 2019 and 2023, settlement but still experiences high levels with a negative real GDP per capita growth rate. of insecurity. After independence in 1950, During this period, the economy has experienced Somalia was led by democratically elected civilian repeated shocks, including floods, drought, the governments until a coup d’etat in 1969 led by COVID-19 pandemic, Russia’s invasion of Ukraine, Siad Barre, who served as President until he was and rising food prices. The prolonged drought in overthrown in 1991. After the collapse of the 2021 and 2022 is estimated to have led to close to state in 1991, Somalia has faced continued conflict one-third of the population being food insecure and instability, which has hindered its progress by January 2023.² The legacy of the prolonged in securing sustainable development. In 2012, a civil war in Somalia has resulted in negligible Provisional Constitution was approved, however, domestic production capabilities. As such, Somalia political actors still need to resolve issues such is dependent on imports for basic commodities, as the sharing of resources between regions including food and fuel, which increases its and levels of government. In 2012, a Provisional vulnerability to external price shocks. However, in Constitution was approved. However, political December 2023, Somalia achieved a historic HIPC actors still need to resolve issues such as resource- completion point milestone, with the country sharing between regions and levels of government. receiving debt relief of US$4.5 billion. This resulted In addition, Al-Shabaab maintains territorial control in Somalia’s external debt falling from 64 percent in parts of southern and central Somalia. Although of GDP in 2018 to less than 6 percent of GDP in conflict-related incidents and fatalities have been 2023.³ Further, in March 2024, Somalia became on a declining trend, in 2022, there was an uptick th the 8 member of the East African Community. in incidents by 39 percent as compared to 2021. ¹ SNBS, 2024 ² By February 2023, there was a fifth consecutive failed rainy season, resulting in nearly 5 million people becoming food insecure (crisis level 3 – IPC Phase 3). https://www.ipcinfo.org/ipc-country-analysis/details-map/en/c/1156238/?iso3=SOM ³ World Bank 2024a. 1 PART A: CORE ANALYTICS AND CROSS-COUNTRY BENCHMARKING - CHAPTER 1: THE INCIDENCE, NATURE, AND EVOLUTION OF POVERTY IN SOMALIA This was due to fighting with Al-Shabaab in central report also uses the Somali High Frequency Survey- and southern Somalia.⁴ Women and girls also face Wave 2 from 2017 (SHFS-W2) and is supported by additional challenges in economic empowerment four focus group discussions (FGDs).⁷ The SOMPA and political participation.⁵ aims to identify key constraints for poverty reduction in Somalia, drawing on the asset framework to 3. Somalia is highly exposed to climatic explore factors that may inhibit household income shocks. The average annual rainfall is under 200 growth, especially among the poor. According to the millimeters for most of the country, although asset framework, a household’s ability to generate some regions have more rain (northern highlands income is dependent on (i) the assets they own, and the south). Somalia also faces high average (ii) the intensity they use these assets, and (iii) the temperatures, with most areas experiencing mean returns these assets provide. Income-earnings assets daily maximum temperatures above 30 degrees include human capital, financial and physical assets, Celsius. Most of the country has desert or semi- social capital, and natural capital. The intensity of the desert ecosystems with little grassland vegetation. use of assets includes labor force participation or the As a result, over half of Somalia’s land mass is exploitation of land through agricultural production, suitable for nomadic pastoralism. At the same and the return to these assets consists of the time, 13 percent can support cultivation, including nominal prices of factors of production. Transfers, seasonal agropastoralism and a smaller irrigated such as remittances, can also increase household agropastoralism zone located along the two main income, while income is also affected by the prices rivers. Most of the Somali population resides in of goods and services the household consumes. these agricultural areas or coastal cities. Between Lastly, asset accumulation, the intensity of usage, 2020 and 2022, Somalia experienced its longest the returns to assets, and transfers can be directly drought in four decades, with five consecutive affected by external shocks such as climatic, health, failed rainy seasons. However, as rainfall returned and employment-related shocks.⁸ to higher levels in 2023, this coincided with flooding in vulnerable areas. These two sources 5. The SOMPA will be divided into two parts: of climatic shocks cause economic losses for the the first will focus on recent poverty trends, Somali population while resulting in large numbers while the second will look at specific topics of displaced individuals.⁶ related to poverty trends. Part A of the SOMPA will consider the changes in poverty since the last 4. This Somali Poverty and Equity Assessment household budget survey conducted in 2017. It (henceforth SOMPA) follows the 2019 Somali will also decompose the trends in poverty, look Poverty and Vulnerability Assessment and is at the profile of the poor in 2022, and compare the first to draw on comprehensive data on the monetary and non-monetary poverty. Part B will Somali population's living standards. The SOMPA then build upon the trends outlined in Part A and is largely based on the 2022 Somali Integrated explore three thematic topics that may explain Household Budget Survey (SIHBS), the first the observed trends. The first thematic topic will nationally representative survey since 1985. The be livelihoods, the second will look at households’ ⁴ Between 2018 and 2022, 20,201 fatalities were caused by FCV events. This is based on data from the Armed Conflict Location and Event Data Project database: https://acleddata.com/. ⁵ World Bank 2023b. ⁶ World Bank 2023a. ⁷ Focus group discussions covered urban areas, IDPs, and nomadic individuals in Mogadishu, Burtinle, and Guricel. ⁸ López-Calva. and Rodríguez-Castelán 2016. 2 SOMALIA POVERTY AND EQUITY ASSESSMENT exposure to shocks, focusing on climatic shocks, to their rural and nomadic counterparts.⁹ The while the final deep dive will look specifically at the nomadic population had the highest rate of nomadic population. The SOMPA will conclude with poverty, with over three-quarters living below a section on policy recommendations. the poverty line.¹⁰ Despite the lowest poverty headcount rate, the poor are concentrated in Poverty remains high in Somalia, urban areas, reflecting the fact that just under with the bulk of the poor living two-thirds of the population live in urban areas in urban areas. (Box 2). The national poverty gap is 20 percent, which means that, on average, the poor consume 6. In the absence of economic growth, poverty $150 less than the poverty line. The poverty gap in Somalia remains high. In 2022, over 54 percent follows the same pattern across rural, urban, of the population lived below the poverty line, with and nomadic areas, with the largest poverty gap lower rates among urban households compared among the nomadic population (Figure 2). Figure 1: Assets approach to market income Returns to assets Transfers Accumulation Intensity of External Household market of assets use of assets Shocks income Prices Prices Source: Source: López-Calva and Rodríguez-Castelán 2016. Figure 2: Poverty Indicators A. Poverty Headcount and Poverty Gap, 2022 B. Distribution of the population living below the national poverty line by area, 2022 100% 80% 78% Urban: 64% Nomadic: 12% 65% 60% 54% 46% % 40% 37% 25% 20% 15% 20% 0% Somalia Urban Rural Nomadic Rural: 24% Poverty Headcount Poverty Gap Source: Authors’ estimates based on SIHBS 2022. ⁹ The poverty estimation methodology is described in Box 1. ¹⁰ The term “nomadic” is used interchangeably with “pastoral nomadic” in this report. 3 PART A: CORE ANALYTICS AND CROSS-COUNTRY BENCHMARKING - CHAPTER 1: THE INCIDENCE, NATURE, AND EVOLUTION OF POVERTY IN SOMALIA Box 1: Poverty estimation from the Somali Integrated Household Budget Survey (SIHBS) 2022 The poverty analysis is based on the Somali Integrated Household Budget Survey (SIHBS) carried out by the SNBS between May and July 2022. The primary objective of the SIHBS was to collect detailed information on household expenditures and consumption incurred on goods and services to monitor household welfare and measure poverty. In addition to welfare data, SIHBS collects other socio-economic information relevant to monitoring the living conditions of Somali households, such as access to basic assets, facilities, and services. The term “poverty” is used to indicate households whose per capita expenditure falls below the poverty line, i.e., households where individuals cannot afford the cost of meeting their basic needs, which include both food and non-food items (Ravallion 1994, 2016). The cost of basic needs poverty line is anchored in the cost of food that can provide the energy requirement, defined as the average number of kilocalories (per person per day) that humans need to survive. This analysis uses the standard requirement of 2,200 kcal/ person/day. Combining the costs to obtain this food requirement and necessary non-food items, the cost of basic needs poverty line is 752 USD/person/year. Source: SNBS. 2023. Poverty and inequality report. 7. Though poverty is high throughout the country, of poverty in some regions, the absolute number of there are notable spatial differences, with higher poor are concentrated in the more populated regions poverty in the central and southern regions. with large urban centers. For instance, Banadir Poverty rates range from 39 percent in Awdal to (which includes the capital, Mogadishu) has the third 87 percent in Middle Shabelle, with lower poverty lowest poverty rate but the largest share of the poor rates in the Northern regions (Figure 3). Over two- at 14 percent. In contrast, despite Middle Shabelle thirds of the population is poor in Middle Shabelle, having the highest poverty rate, it only accounts for Hiraan, Bakool, and Mudug. Despite the higher rates 4 percent of the poor in Somalia (Figure 4). Figure 3: Regional Poverty Map, 2022 Figure 4: Share of Poor by Region, 2022 # 80 to 90 70 to 80 60 to 70 50 to 60 40 to 50 30 to 40 No data Source: Authors’ estimates based on SIHBS 2022. 4 SOMALIA POVERTY AND EQUITY ASSESSMENT Box 2: Why are the bulk of the poor in urban areas? Somalia has a very high level of urbanization relative to its income level. Almost two-thirds of the Somali population reside in urban areas, with just under a quarter in rural areas and the remaining 11 percent in Nomadic areas (Figure 5). Somalia’s rate of urbanization is much higher than expected for its income level (Figure 6). A large share of this rapid urbanization is driven by a large, displaced population, who often flee rural areas because of natural shocks or conflict. However, cities also offer the prospect of better living conditions and greater employment opportunities.¹¹ Figure 5: Population Shares, 2022 Figure 6: Urbanization, 2022 100 Urban Population (%) Urban: 65% Nomadic: 11% 80 60 40 Rural: 24% 20 0 2 3 4 5 6 Log GDP per Capita, PPP (2017 $) Source: Authors’ estimates based on SIHBS 2022. Source: World Development Indicators and SIHBS 2022. 8. Over one-fifth of the population are in extreme surveys, consumption data in 2017 was imputed poverty. A household is considered extreme poor to be comparable to the 2022 consumption data.¹² if their total per capita consumption falls below Between 2017 and 2022, the national poverty rate the food poverty line. Extreme poverty is largest remained constant, with a statistically insignificant among the nomadic population, with almost half change from 55.1 to 54.4 percent. This trend is being extreme poor. Over one-quarter of the rural consistent with the negative GDP per capita growth population and 14 percent of the urban population between 2017 and 2022.¹³ However, it is important are extreme poor. There is also substantial regional to note that this stagnant poverty trend covered a variation, ranging from 11 percent in Banadir to 56 period where global poverty increased, with many percent in Middle Shabelle (Figure 7). other African countries experiencing increases in poverty.¹⁴ While poverty did not change at the Monetary poverty trends national level nor in urban areas, it increased in both rural (58.8 to 65.5 percent) and nomadic (77.4 9. The monetary poverty rate in Somalia has to 78.4 percent) areas (Figure 8).¹⁵ The extreme remained unchanged between 2017 and 2022. Due poverty rate followed a similar pattern (Figure 9). to methodological differences between the two Likewise, the poverty gap remained unchanged at ¹¹ World Bank 2021b. ¹² This was done using the survey-to-survey imputation method (see details in Box 3 and the associated background note). ¹³ The Somali economy grew at around 3 percent in 2018 and 2019. In 2020, the economy contracted by 2.6 percent due to COVID-19 pandemic. Growth recovered to 3.3 percent in 2021 largely driven by household consumption and private investment, supported by robust remittance inflows and credit growth. Economic activity in 2022 slowed to 2.4 percent on the back of the global economic slowdown, moderation in remittance inflows and the domestic drought conditions. ¹⁴ World Bank 2024b. ¹⁵ Extreme poverty is defined as the population share of those whose per capita consumption is below the food poverty line. 5 PART A: CORE ANALYTICS AND CROSS-COUNTRY BENCHMARKING - CHAPTER 1: THE INCIDENCE, NATURE, AND EVOLUTION OF POVERTY IN SOMALIA the national level while it increased for the rural 1). There was also little change in real per capita and nomadic populations (Figure 12). This suggests consumption across the national consumption that while the share of the nomadic population who distribution between 2017 and 2022, with only were poor only marginally increased, those that minor increases at the very top of the distribution. were poor were, on average, further away from However, different patterns exist across rural, urban, the poverty line. Further, an increasing share of and nomadic populations. The rural population, on the rural population became poor and additionally, average, experienced around a 2 percent decrease in the poor rural population were further away, on consumption annually between the two periods. This average, from the poverty line. explains why the poverty headcount, poverty gap, and severity of poverty all increased in rural areas. In 10. Nationally, average consumption per capita urban areas, the consumption growth was negative, did not grow, with only the poorest and richest albeit small in magnitude, at every decile of the experiencing positive consumption growth distribution. The bottom 80 percent of the nomadic between the two years. The lack of consumption population experienced a decline in consumption, growth is also consistent with the negative GDP with only the richest quintile experiencing positive per capita growth over the same period (Table consumption growth (Figure 10). Figure 7: Extreme Poverty Rates 70% 60% 50% 40% 30% 20% 10% 0% All Middle Shabelle Rural/DP Urban Nomadic Awdal Bakool Banadir Bari Galgaduud Gedo Hiraan Lower Shabelle Bay Lower Juba Waqooyi Galbeed Mudug Nugaal Sool Togdheer Sanaag All Residency Region Source: Authors estimates based on SIHBS 2022. Figure 8: Poverty Trends, 2017 to 2022 Figure 9: Extreme Poverty Trends, 2017 to 2022 47.3 90 50 43.8 78.4 77.4 80 65.5 70 40 58.9 55.2 54.4 60 28.3 30 46.1 46.1 50 21.8 21.7 21.3 40 20 14.4 14.0 30 20 10 10 0 0 National Urban Rural Nomadic National Urban Rural Nomadic 2017 2022 2017 2022 Source:Authors’ estimates based on SHFS-W2 2017 and SIHBS 2022. 6 SOMALIA POVERTY AND EQUITY ASSESSMENT Table 1: Change in Annual Average Consumption 2017 2022 Growth Per capita consumption (2022 $) 875.3 874.3 0% GDP per capita (Real 2022 Prices, $) 703 664 -6% Source: Authors’ estimates based on SIHBS 2022 and SHFS-W2 2017. Figure 10: Annual Per capita consumption growth rate, 2017-202216 National Urban .4 0 .2 0 -.5 -.2 -1 -.4 -.6 -1.5 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Decile Decile Rural Nomadic -1 3 -1.5 2 1 -2 0 -2.5 -1 -3 -2 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 Decile Quintile The horizontal dash and short dash lines show the average and median growth rates for the entire sample, respectively. The y-axes show the annualized growth rate (%). The y-axes show the quantities of real per capita consumption (RPCC) in each year. The quintiles are used for the nomadic due to the small sample size. Source: Authors’ estimates based on SIHBS 2022 and SHFS-W2 2017. 11. The movement of population from nomadic population.¹⁷ As a result, the total number of to urban areas offsets the increase in poverty in poor increased by around 1 million (Figure 11). each area, resulting in a stagnation in poverty The movement of the population from nomadic at the national level. Nationally, the population areas with higher poverty rates to areas with lower is estimated to have grown by 2 million between poverty levels decreased poverty by 2.4 percentage 2017 and 2022, with increases in the rural and points between 2017 and 2022. However, this urban population and a decrease in the nomadic was almost completely offset by decreases in ¹⁶ The growth incidence curve (GIC) displays the growth rate of household per capita consumption expenditure for every consumption quantile (i.e., households are ordered in real per capita consumption and are divided into equally sized bins). The horizontal lines show the average growth rate for each group. ¹⁷ The estimates assume that the total population increased by 2.3 and 0.5 percentage points for the urban and rural population groups and that the total population decreased by 0.8 million for the nomadic population. 7 PART A: CORE ANALYTICS AND CROSS-COUNTRY BENCHMARKING - CHAPTER 1: THE INCIDENCE, NATURE, AND EVOLUTION OF POVERTY IN SOMALIA consumption within rural and nomadic areas, no change in consumption between the two years, which in turn increased poverty by 1.7 percentage the movement of the population out of nomadic points. As a result, poverty only marginally declined areas would have decreased the poverty rate by over between 2017 and 2022 by 0.8 percentage points six percentage points, while the movement of the (Table 2). population into urban areas would have increased poverty by 3.5 percentage points. This suggests that 12. Without the movement of population, poverty those leaving nomadic areas for urban areas are poor would have increased nationally. If there had been households, as this population movement decreases no movement in population across areas between poverty in nomadic areas and increases it in urban 2017 and 2022, the reduction in consumption in areas (Table 2). This is perhaps unsurprising given rural areas alone would have increased the poverty that these households are likely to be IDPs, who rate by 1.6 percentage points, while the decreases resort to moving due to climate or conflict, and who in consumption in urban and nomadic areas would often lack access to clan-based support systems and have had little effect. Alternatively, if there had been typically lack the skills for urban livelihoods.¹⁸ Box 3: Survey to Survey Imputation Method Survey-to-survey imputation was used to estimate the change in monetary poverty, given the methodological differences between the two most recent surveys. This approach used consumption data from the SIHBS 2022 as the base and comparable data on household characteristics from the SIHBS 2022 and SHFS-W2 2017 surveys. The data on household characteristics includes variables that tend to be strongly correlated with consumption and poverty, such as household demographics, household ownership of durable goods, and housing quality. An estimation model of the relationship between the poverty correlates, and consumption is then used to impute consumption for households in the SHFS-W2. The imputation approach used the technique developed in the Survey of Well-being via the Instant and Frequent Tracking (SWIFT) approach (Yoshida et al., 2022). Figure 11: Poor population measured by the national Figure 12: Poverty gap rates using the national poverty poverty line, 2017-2022 line, 2017-2022 36.8 20 40 34.3 35 Population (Millions) 15 30 24.7 25 20.8 10 20.2 19.9 20 15.2 15.1 5 15 0 10 2017 2022 2017 2022 2017 2022 2017 2022 5 National Urban Rural Nomadic 0 National Urban Rural Nomadic Poor Non-Poor 2017 2022 Source: Authors’ estimates based on SHFS-W2 2017 and SIHBS 2022. ¹⁸ World Bank 2021b1; This finding is supported by the IDP FGD, which found most individuals became IDPs due to climate or conflict related shocks. 8 SOMALIA POVERTY AND EQUITY ASSESSMENT Table 2: Decomposition of Change in Poverty Rate into Intra-population group and population shifts, 2017-2022 Percentage change in poverty rate: Intra-area effect Arising from pop. shifts Residual Total impact by pop. groups Urban 0.0 3.5 0.0 3.4 Rural 1.6 0.3 0.0 1.9 Nomads 0.2 -6.2 -0.1 -6.1 Total 1.7 -2.4 -0.1 -0.8 Source: Authors’ estimates based on SIHBS 2022 and SHFS-W2 2017. 13. The increase in poverty for the rural population in poverty described above, there are a host of is largely due to negative average consumption other factors that are significantly correlated with growth, while for the nomadic population, it is household welfare. These factors can be identified due to a more unequal distribution. At the national using a regression model that explains a household’s level, the marginal decline in poverty was driven poverty status based on a set of demographic and equally by an increase in average consumption and a socioeconomic characteristics (Figure 14). more equitable distribution. The increase in poverty in urban areas was driven by a decrease in average Figure 13: Growth-Redistribution Decomposition, consumption, although this was partially offset by 2017 to 202219 the distribution becoming more equitable, consistent 6.6 with those moving to urban areas being poor. Most 6 5.6 of the poverty increase among the rural population was driven by a decrease in average consumption. 4 Lastly, in the nomadic population, the increase in 2.6 poverty was driven by the distribution becoming less 2 % equitable. However; this was offset partially by an 1.0 1.0 1.0 increase in average consumption (Figure 13). 0 -0.0 Who are the poor? -0.4 -0.4 -0.7 -1.0 -2 -1.6 14. Additional analysis can determine what National Urban Rural Nomadic demographic and socioeconomic household characteristics are associated with welfare. In Growth Distribution Total change in p.p addition to the spatial and regional differences Source: Authors’ estimates based on SHFS-W2 2017 and SIHBS 2022. ¹⁹ This method developed by Datt and Ravallion (1992) measures the effect of growth and redistribution in poverty reduction between over a given time period. The method asks what poverty would look like with the mean consumption level of a given year and the consumption distribution of a different year, and vice-versa. 9 PART A: CORE ANALYTICS AND CROSS-COUNTRY BENCHMARKING - CHAPTER 1: THE INCIDENCE, NATURE, AND EVOLUTION OF POVERTY IN SOMALIA Figure 14: Correlates of Poverty by Residency20 Wager Earner Household Enterprise Owns agricultural or commercial land Economic Activity Any International Remittance Domestic Remittance Only Income - None Income - Mixed Income - Non-Agriculture only Head - Employed Share Employed Head Education=Tertiary Education Head Education=Secondary Head Education = Primary Head - Literate IDP Household Size = 10+ Household Size = 7 to 9 Demographics Household Size = 4 to 6 Share of Children Female-Headed - Widowed Female-Headed - Never Married Female-Headed - Married -0.6 -0.4 -0.2 0 0.2 0.4 0.6 Significant Marginal Effects Only Nomadic Urban Rural Source: Authors’ estimates based on SIHBS 2022. Large households with many children and displaced Education protects against poverty, though few households are more likely to be poor Somalis have completed primary education 15. Larger households are more often poor in 16. Literacy and formal education of the all areas of residency, while IDP households household head are associated with lower are more often poor in urban areas. Relative to a poverty. Most of the Somali population does not household with 1 to 3 members, a larger household have any education, with only 7 percent having size is associated with higher poverty rates, with completed primary education, 5 percent having the likelihood increasing with size. Further, among completed secondary, and only 2 percent having urban households, a larger share of children is tertiary education. Relative to a household headed associated with poverty, although this is only in by an individual with no education, those with urban areas. Being an IDP is associated with higher secondary are less likely to be poor, regardless of poverty rates in urban and nomadic households, area of residency. A household head with tertiary but not rural. Relative to male-headed households, education is also associated with lower poverty households headed by a female who is widowed are in rural and urban areas, while primary education more often poor among nomadic households, and is only significant in rural areas. Likewise, in rural those with a female head who has never married and urban areas, having a literate household head are more often poor in rural areas (Figure 14). is associated with lower poverty rates. In contrast, ²⁰ Figure 18 shows the marginal effects of a probit regression, with the poverty status as the dependent variable. 10 SOMALIA POVERTY AND EQUITY ASSESSMENT literacy has no association with poverty in nomadic Non-monetary poverty is more areas (Figure 14). prevalent than monetary poverty. Running household enterprises is linked with higher Poverty and Access to Services and Amenities welfare levels, as is receiving remittances from abroad 17. International remittances are associated Health with lower poverty in rural and urban areas, while domestic remittances reduce the likelihood 18. The average woman between 15 and 49 has of poverty in nomadic households. Nationally, seven children, with very few births taking place 15 percent of households received remittances with a skilled provider. The average Somali woman from abroad, while 7 percent received domestic has 6.9 children, ranging from 6.4 in urban areas remittances. A larger share of members being to 7.3 in nomadic areas. There is little difference in employed is only associated with lower poverty fertility among women from the bottom 80 percent among rural households. Income sources outside of the wealth distribution, with only those from the of agriculture are associated with lower poverty richest households having a lower fertility rate (5.6 in rural households, although the opposite is true compared to 7 and above). Fertility also decreases among nomadic households. The importance of with education, from 7.2 children for women without remittances is also evident, although the source is any education to 3.7 among women with higher important. Among nomadic households, receiving education (Figure 15). Further, a poor household domestic remittances is associated with lower has, on average, 1.3 more members relative to a poverty, while for rural and urban households there non-poor, with household size decreasing across the is no association for domestic remittances. However, consumption distribution in rural, urban, and nomadic international remittances are associated with lower areas. However, despite additional members, poorer poverty in rural and urban areas. The ownership households do not have more working members on of a household enterprise is associated with lower average. In addition, the poorer households have a poverty in all three areas of residency. Further, lower ratio of working-aged members to household while wage earnings are typically associated with size (Table 8). Less than one-third of live births were lower poverty, in Somalia, having an individual with delivered with a skilled provider present, with the wage earnings is associated with higher poverty in largest share in urban areas (51 percent). The share rural and nomadic areas (Figure 14). of live births with a skilled provider is very low for Figure 15: Fertility rates among women between Figure 16: Percent of live births in last five years 15 to 49, 2020 delivered by skilled provider, 2020 9 90 8 80 % of Births in last 5 years 7 70 6 60 TFR (15 - 49) 5 50 4 40 3 30 2 20 1 10 0 0 All Urban Nomadic None Primary Secondary Higher Lowest 2nd All Urban Nomadic None Primary Secondary Higher Rural Lowest 3rd 4th Richest Rural 2nd 3rd 4th Richest All Residence Education Wealth All Residence Education Wealth Source: Federal Government of Somalia and UNFPA 2020. 11 PART A: CORE ANALYTICS AND CROSS-COUNTRY BENCHMARKING - CHAPTER 1: THE INCIDENCE, NATURE, AND EVOLUTION OF POVERTY IN SOMALIA nomadic women (8 percent) and the poorest 40 enrollment is very low at 41 percent for primary school percent (around 10 percent). This share increases and 31 percent for secondary school. Compared to with both wealth and education (Figure 16). other low-income African countries, Somalia has the lowest gross primary enrollment rates. Enrollment is Education lower for the poor compared to the non-poor, with a larger difference for secondary education. Enrollment 19. Literacy rates and enrollment rates are increases across the consumption distribution, and lower among the poor, women, and the nomadic primary enrollment lags far behind in nomadic areas population. Just over half of the Somali population (11 percent). For secondary education, the gap reported they are literate, with the highest rates between urban and rural areas widens (37 percent in in urban areas and lowest among the nomadic. urban and 23 percent in rural areas) (Figure 19). At the Regionally, literacy rates are also correlated with regional level, primary school enrollment follows the poverty, with poorer regions having a lower share of same pattern as literacy, with poorer regions having literate individuals (Figure 17 and Figure 18). Gross lower enrollment (Figure 20). Figure 17: Literacy Levels, 2022, 15+ Figure 18: Literacy Levels by Region 70% 100% 60% 90% 80% Poverty Headcount 50% 70% 40% 60% 15+ 50% 30% 40% 20% 30% 20% 10% 10% 0% 0% All Non-Poor Poor Rural Urban Nomadic Male Female 0% 20% 40% 60% 80% Literacy ( 15+ ) Figure 19: Gross Enrollment Figure 20: Primary Gross Enrollment by Region 70% 100% 60% 90% 50% 80% 40% Poverty Headcount 70% 30% 60% 20% 50% 10% 40% 0% 30% All Non-Poor Poor Urban Nomadic Rural Male Female Poorest 2nd 3rd 4th Richest 20% 10% 0% All Poverty Residency Gender Quintile 0% 20% 40% 60% 80% GER Primary GER Secondary Primary GER Source: Authors’ estimates based on SIHBS 2022. 12 SOMALIA POVERTY AND EQUITY ASSESSMENT Housing Characteristics relative to other low-income African countries, which is likely a reflection of its higher urbanization levels 20. Poor households have lower access to (Figure 95). electricity. Just over 60 percent of individuals had access to electricity in 2022, with higher rates for urban 21. While poor households have slightly worse individuals (80 percent), followed by rural (39 percent), access to improved drinking water, predominately and lastly nomadic (8 percent). Access to electricity is in rural areas, there is no correlation at the almost double among non-poor individuals compared regional level. 70 percent of the population has to the poor, with wider differences between the poor access to improved drinking water in the dry season. and non-poor in rural areas (28 to 61 percent) relative Once again, those in urban areas have the best access to urban areas (68 to 91 percent) (Figure 21). There to improved drinking water in the dry season. The is little difference between the poor and non-poor in non-poor also have better access, with a 9 percentage nomadic areas, with both having very low access. Over point difference compared to the poor. Under two- time, the share of individuals with access to electricity thirds of rural individuals have access to improved has improved from 49 percent in 2017 to 62 percent drinking water in the dry season. The gap between in 2022 (Figure 22). Regionally, access to electricity is the non-poor and poor is largest in rural areas at seven worse in poorer regions. Lastly, Somalia performs well percentage points. Access to improved drinking water Figure 21: Access to Electricity Figure 22: Change in Access to Electricity, 2017 to 2022 100% 100% 90% 90% 80% 80% 70% % of individuals 60% 70% % of individuals 50% 60% 40% 30% 50% 20% 40% 10% 30% 0% 20% All Non-Poor Poor All Non-Poor Poor All Non-Poor Poor All Non-Poor Poor 10% 0% Residence Rural Urban Nomadic 2017 2022 Source: Authors’ estimates based on SIHBS 2022 and SHFS-W2 2017. Figure 23: Access to Improved Drinking Water Figure 24: Trend in Improved Drinking Water, 2017 to 2022 100% 80% 80% % of individuals 70% 60% 60% 40% 50% 20% 40% 0% 30% All Non-Poor Poor All Non-Poor Poor All Non-Poor Poor All Non-Poor Poor 20% 10% All Rural Urban Nomadic 0% Rainy Dry 2017 2022 (Rainy) 2022 (Dry) Source: Authors’ estimates based on SIHBS 2022 and SHFS-W2 2017. 13 PART A: CORE ANALYTICS AND CROSS-COUNTRY BENCHMARKING - CHAPTER 1: THE INCIDENCE, NATURE, AND EVOLUTION OF POVERTY IN SOMALIA is low in nomadic areas at 31 percent in the dry season, (2023), households are defined as non-monetary though there is no difference between the poor and poor if they are deprived in non-monetary indicators non-poor (Figure 23). Over time, there has been an such as food security, housing characteristics, and improvement in the share of individuals with access education. Following the same methodology, a to improved drinking water (Figure 24). However, at household is considered to be chronic poor if it is the regional level, there is little correlation between classified as both monetary and non-monetary poor. poverty and access to drinking water. Nationally, over three-quarters of the population are considered non-monetary poor, while 47 percent 22. Very few individuals have a bank account, and are considered chronic poor. Non-monetary poverty credit is largely used for consumption. Less than follows the same pattern as monetary poverty, with 10 percent of individuals have a formal bank account, the highest rates among the nomadic population (95 with higher rates among the non-poor, urban areas, percent), followed by rural (79 percent) and urban and men. Over a quarter of households took a loan areas (73 percent) (Table 3). Only 15 percent of the in the last 12 months, with higher rates among the population is neither monetary nor non-monetary poor and nomadic households, with little difference poor (Figure 26). Chronic poverty is also concentrated across household head gender. Most loans were from in the central part of the country (Figure 25). Most traders, regardless of household characteristics, and households are deprived in the education dimension, almost all loans were mainly used for consumption. followed by sanitation and flooring. The larger share of nomadic households deprived in each dimension 23. Non-monetary poverty is higher than is also reflected by the fact that nomadic households monetary poverty.²¹ Applying the methodology are, on average, deprived in 7 dimensions compared developed by Bolch, Lopez-Calva, and Ortiz-Juarez to 4 for rural households and 3 for urban households. Table 3: Monetary, Multidimensional, and Chronic Poverty All Rural Urban Nomadic Monetary 54.4% 65.5% 46.1% 78.4% Non-Monetary 76.7% 79.4% 72.6% 94.7% Chronic 46.5% 55.4% 38.2% 74.3% Figure 25: Chronic Poor by Region Figure 26: Share of Individuals by monetary and non-monetary poverty status 100% 80% % .8 to .9 60% .7 to .8 .6 to .7 40% .5 to .6 20% .4 to .5 .3 to .4 0% .2 to .3 All Rural Urban Nomadic .1 to .2 0 to .1 All Residency No data Both Monetary Only Non-Monetary Only Neither Source: Authors’ estimates based on SIHBS 2022. ²¹ The definition of non-monetary, or multidimensional, poverty can be found in the Chapter 1 Annex. 14 SOMALIA POVERTY AND EQUITY ASSESSMENT Inequality in consumption remains contribution to the inequality, followed by region. relatively low, despite high Enrollment in primary education is less unequal than inequality in opportunities. in secondary, although it is more unequal than access to services such as water, sanitation, and electricity. 24. Inequality is relatively low in comparison to Furthermore, inequality in primary education other low-income African countries, with a higher enrollment is driven predominately by region and level among the Nomadic population. The national less so by poverty relative to secondary education. Gini index is 0.350, which puts Somalia’s inequality at Better service delivery in urban areas, notably as it the lower end of the distribution among low-income pertains to education, is a pull factor for urbanization African countries (Figure 27). Inequality, as measured among IDPs.²² Once these differences in coverage by the Gini index, is largest among the nomadic rates across characteristics are accounted for, the population. Over time, there has been no change at HOI value for primary education enrollment is 20.6 the national level in the Gini index. However, while and 10.6 for secondary education (Figure 91 and there has been no change in urban areas, there has Figure 92).²³ been an increase in inequality among both the rural (1.4 points) and nomadic (5 points) populations 26. The location of the household largely drives (Figure 28). Inequality is largely driven by differences inequality in access to services, although poverty in welfare within regions or areas of residency. still hinders access to electricity and improved sanitation. The coverage rate of access to services Inequality of opportunity. is higher than education, with improved sanitation having the lowest and improved drinking water 25. Education has the highest inequality in access, having the highest coverage. Access to electricity with location being a key driving factor. As shown shows the largest inequality among the services above, access to education is relatively low overall based on circumstance, with the area of residency in Somalia, especially for secondary education having the largest contribution, followed by region (Figure 19 and Figure 20). Enrollment in secondary and poverty status in that order. Access to improved education displays the biggest inequality based sanitation has a similar inequality level as access to on circumstance, with poverty having the largest electricity, although in that case, the region has the Figure 27: Gini Index, 2022 Figure 28: Change in Gini coefficient, 2017-2022 41.5 0.6 40 35.0 35.0 36.5 33.7 33.3 0.5 30.6 32.0 30 0.4 20 % 0.3 0.2 10 0.1 0 0.0 HFS HBS HFS HBS HFS HBS HFS HBS National Rural Urban Nomadic National Urban Rural Nomadic Source: Authors’ estimates based on SIHBS 2022 and SHFS-W2 2017. ²² A finding from the IDP FGD. ²³ A description of the human opportunity index can be found in Box 6. 15 PART A: CORE ANALYTICS AND CROSS-COUNTRY BENCHMARKING - CHAPTER 1: THE INCIDENCE, NATURE, AND EVOLUTION OF POVERTY IN SOMALIA Figure 29: Shapely Decomposition of Each Circumstance, 2022 100 80 60 40 20 0 Currently attending Primary Secondary Access Improved drinking Improved drinking Improved school NER NER to electricity water (rainy) water (dry) sanitation Residency Head Sex Head Education Region Poverty IDP Source: Authors’ estimates based on SIHBS 2022. largest contribution, followed by poverty status particularly relevant to that objective of inclusive and area of residency. IDP status also contributes and shared prosperity. to inequality in access to electricity and improved sanitation. Lastly, access to drinking water has the 28. The first deep dive will focus on income highest coverage and the lowest inequality based and employment. The profile of the poor on circumstance. The region and area of residency of shows that a household with a wage earner is the household largely drive the inequality that does positively associated with poverty in rural and exist (Figure 29). nomadic areas. At the same time, there is neither a positive nor negative association in urban The remainder of this poverty areas. Further, having an employed household assessment will focus on three head is also positively associated with poverty deep-dive topics. (Figure 14). In addition, it is well documented that Somalia has extremely low labor force 27. Somalia experiences high persistent poverty, participation.²⁶ The first deep dive will focus on with no average per capita consumption growth the income and livelihoods of Somali households, and negative per capita GDP growth between looking in detail at the differences in the type of 2017 and 2022. Economic growth remains a key income, type of employment, sectors of work, input for poverty reduction, with most examples of and household enterprises between poor and poverty reduction coinciding with economic growth.²⁴ non-poor households and individuals. However, growth alone does not guarantee poverty reduction, and in fact, sub-Saharan African countries 29. On the back of an unprecedented multi- have been less successful at converting economic season drought in 2022, the second deep dive growth into poverty reduction.²⁵ Therefore, while it will focus on households’ exposure to shocks, is important that Somalia utilizes policies that will especially climatic shocks. Households’ exposure encourage stronger economic growth, it is also to external shocks, especially repeated shocks, can important that it is done inclusively so that the poor have a substantial impact on a household’s ability can benefit from future economic growth. The to produce market income, which in turn can remainder of the report will focus on areas that are either keep a poor household poor or push a non- ²⁴ Ames, Devarajan and Izquierdo 2001; World Bank 2022a; Rodrik 2000. ²⁵ Wu et al. 2024. ²⁶ World Bank 2021a. 16 SOMALIA POVERTY AND EQUITY ASSESSMENT poor household below the poverty line.²⁷ Further, The nomadic population has the highest rates of climate change effects are already significant monetary and non-monetary poverty, with just under in Somalia, and it is likely that the incidence of three-quarters being both monetary and non-monetary extreme droughts will increase.²⁸ As a result, the poor ( Figure 2 and Table 3). Nomadic livelihoods are second deep dive will focus on who is exposed and particularly vulnerable to climatic shocks, as illustrated vulnerable to these climatic shocks, how different by the latest drought. In addition, inequality is the households respond to these shocks, and what can highest among the nomadic population, with the be done to increase their resilience. top quintile able to lift themselves above the poverty line (Figure 27). Therefore, the deep dive will make 30. The final deep dive will focus on the nomadic comparisons across the consumption distribution to households, who are, by area of residency, the determine factors that may enable the top quintile to poorest and also very exposed to climatic shocks. be above the poverty line. ²⁷ López-Calva and Rodríguez-Castelán 2016. ²⁸ World Bank 2023a. 17 PART B: DEEP-DIVES - CHAPTER 2: LIVELIHOOD DEEP-DIVE PART B: DEEP-DIVES CHAPTER 2: LIVELIHOOD DEEP-DIVE 31. The livelihood deep dive will focus on labor 64) are engaged in the labor market, that is, either and household enterprises and how they can in work-for-pay employment or unemployment contribute to poverty reduction for Somali (Figure 30). households. As outlined in the asset framework, households' capacity to generate income is based on 33. The Somalia labor market displays some the assets they own.²⁹ Poorer households’ main -and unique traits for its income level. Somalia has often only- asset is typically their labor, as they often extremely low labor force participation, especially lack financial, physical, and natural assets. For these when compared to other low-income countries households, it would seem that only employment (Figure 31). LFP is also low across all demographic can offer a sustainable route out of poverty.³⁰ groups, especially among women and the nomadic Therefore, the intensity of its use, i.e., labor force population. Compared to other low-income participation and its returns in wages, are key countries, Somalia has the lowest male LFP, the determinants of poor households’ market income. third lowest female LFP, and the fifth largest This chapter will provide a breakdown of the Somali gender gap in LFP. Although the gender gap in LFP working-age population, looking at the numbers in is smaller than in contextually similar low-income employment, unemployment, and inactivity. From countries such as Djibouti, Sudan, and Yemen, this this breakdown, the chapter will explore why labor is more driven by a lower male LFP than a higher force participation is so low in Somalia, the dualistic female LFP (Figure 31). The only group with higher nature of wage employment, and the importance of LFP is those with tertiary education. However, household enterprises (HHEs). they account for only 2 percent of the population. Interestingly, there is no difference in LFP across 32. Very few working-age people are engaged poverty status (Figure 32). A unique feature of in the labor market, especially among women. the Somalia labor market is agriculture's minor Somalia had a working-age (15 to 64) population role, accounting for only 12 percent of all work- of approximately 7 million in 2022. Over two-thirds for-pay employment. As mentioned, this differs (4.7 million) are not engaged in the labor market. from the typical pattern in low-income countries, 1.6 million Somalis were employed in work-for-pay where agriculture typically employs over half of activities, with most working in services (1.1 million) workers.³¹ In addition, Somalia is highly dependent and private wage employment (0.8 million). Around on wage employment, which accounts for over 0.4 million Somalis undertake subsistence work, half of employment. This is much higher than the while 0.3 million are unemployed. The majority of typical share in low-income countries of around the inactive do not wish to work. As a result, only one-fifth of employment.³² Lastly, HHEs are an 27 percent of the working-age population (15 to important source of employment, especially for ²⁹ López-Calva and Rodríguez-Castelán 2016. ³⁰ Fields 2012. ³¹ Merotto, D., et al, 2018. ³² Merotto, D., et al, 2018. 18 SOMALIA POVERTY AND EQUITY ASSESSMENT Figure 30: Somalia Population Breakdown, 2022 Working-Age (7.0mn) Labor Force Subsistence Not working (4.7mn) (1.9mn) (0.4mn) Unemployed In Education Employed (1.6mn) Inactive (3.6mn) (0.3mn) (1.1mn) Don't want to work Sectors: Type: (3.1mn) HH Business Owner HH Business Owner HH Agriculture Do want to work Agriculture (0.2mn) Wage (0.9mn) but didn't search (0.3mn) (0.2mn) (0.2mn) (0.4mn) Industry (0.2mn) Public (0.1mn) Services (1.1mn) Private (0.8mn) Source: Authors’ estimates based on SIHBS 2022. women, accounting for just under one-third of MENA region than other low-income countries in employment in Somalia. They are also associated sub-Saharan Africa (Figure 33). This is likely due to with lower poverty in rural, urban, and nomadic Somalia's small share of viable agricultural land and households (Figure 14). However, compared to harsh climatic conditions, which limit the number other African countries, the share of households of agricultural jobs. As a result, Somalia cannot with an enterprise is relatively low.³³ absorb a large share of labor into agricultural work, as is often the case in other LICs. Why is Somalia’s Labor Force Participation so low? 35. Outside of agriculture, the Somali economy appears to have limited labor demand. Low- Limited agricultural activity and little non- income countries typically experience low agricultural labor demand? unemployment and higher underemployment.³⁵ However, unemployment (16 percent of the labor 34. A unique feature of the Somali labor market, force) in Somalia surpasses underemployment considering its income level, is the small share of (11 percent of the labor force). The high rate of agricultural employment. Low-income countries unemployment suggests that there is a lack of typically have the largest share of agricultural economic opportunities within the economy. employment, often accounting for over half of This is supported by the fact that over half of the employment.³⁴ However, in Somalia, agriculture economically inactive individuals who wanted to only accounts for 12 percent of all work-for-pay work but did not search for employment were employment and under a third of employment discouraged workers, i.e., they were tired of if subsistence activities are included. Somalia is searching for employment or stated there were more aligned with the poorest countries in the no jobs matching their skills (Figure 34).³⁶ There ³³ Comparing to countries from the LSMS and WAEMU surveys. These countries have the following share of households with enterprises: Benin (34%); Cote d'Ivoire (44%); Guinea Bissau (64%); Mali (16%); Niger (50%); Senegal (64%); Togo (60%); Ethiopia (23%); Malawi (38%); Nigeria (60%); Tanzania (42%); Uganda (17%). ³⁴ Merotto, D., et al, 2018. ³⁵ Merotto, D., et al, 2018. ³⁶ This is likely an underestimate as the question was only asked to inactive individuals who wanted to work. If an individual had become discouraged and therefore no longer wanted to work, they would not be asked this question. 19 PART B: DEEP-DIVES - CHAPTER 2: LIVELIHOOD DEEP-DIVE Figure 31: International Comparison of Labor Figure 32: Labor Force Participation, 2022 37 Force Participation 70% 100% 60% 90% % of Working-Age 80% 50% 70% 40% LFP (15-64) 60% 50% 30% 40% 20% 30% 10% 20% LICs MENA 10% 0% Yes All Female None Incomplete Tertiary Rural Urban Nomadic Male Primary Secondary Non-Poor Poor No 0% 2.0 3.0 4.0 5.0 GPD per capita (log) All Location Sex Education Pov. Remit. Note: Yellow marker is Somalia; blue markers with red outline are both LICs and MENA countries. Source: Authors’ estimates based on SIHBS 2022 and World Development Indicators. is no gender difference in the share of inactive 36. Regression analysis suggests those who work individuals who are discouraged, although a are either those who can get the few better quarter of inactive women who want to work jobs or those who have no choice but to work in cited family responsibilities as the reason for occupations that offer little returns. Although not searching. The limited role of agriculture accounting for a small share of the population, those and limited non-agricultural labor demand likely with tertiary education are more likely to participate results in the low LFP. in the labor force, even among women. These individuals are likely those who can access the better The combination of limited opportunities jobs available in the Somali economy. In contrast, results in most individuals working in low- women IDPs and unmarried women are more likely return activities out of necessity. to participate in the labor force (Table 11). Figure 33: LIC and MENA Agriculture Share of Figure 34: Reason for Not Searching for Work Despite Employment Wanting to Work Among the Inactive 90 100% Agriculture Share of Employment (%) 80 80% 70 60 60% 50 40% 40 20% 30 20 0% All Non-Poor Poor Rural Urban Nomadic 10 0 2.0 3.0 4.0 5.0 All Poverty Residency GDP per capita (Log) Waiting Age-Related Family Responsibilities MENA LICs LIC & MENA Other Discouraged Source: Authors’ estimates based on SIHBS 2022 and World Development Indicators. ³⁷ Somalia’s data point is from 2022, while the rest are from 2019. 20 SOMALIA POVERTY AND EQUITY ASSESSMENT Box 4: IDPs and Employment Women in IDP households are more often engaged in the labor force, albeit in lower-quality employment. Both men and women from IDP households have a larger LFP relative to their non-IDP counterparts, especially for women, where LFP is 26 percent for IDPs and 17 percent for non-IDPs. IDPs also have high unemployment rates (Figure 35). Women from IDP households have a much larger share in wage employment relative to non-IDP women, with the latter more often operating HHEs (Figure 36). However, despite similar shares of wage employment for IDP women relative to men, the majority of IDP women work for other households, typically in the “other services” sector. Men IDPs also have a larger share working for other households relative to non-IDP men; however, they also have a much larger share employed in the construction sector (Figure 37 and Figure 38). This is suggestive that IDPs are often in lower-skilled occupations due to their lower levels of education and literacy. Figure 35: Labor Force Participation by Sex and Figure 36: Employment Type by Sex and IDP Status IDP Status 100% 60% 80% 50% 60% 40% 40% 30% 20% 20% 10% 0% IDP Non-IDP IDP Non-IDP 0% IDP Non-IDP IDP Non-IDP Men Women Men Women Wage HH Enterprise Owner Employed Unemployed HH Enterprise Worker HH Agriculture Figure 37: Sector by Sex and IDP Status Figure 38: Employer Type by Sex and IDP Status 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% IDP Non-IDP IDP Non-IDP IDP Non-IDP IDP Non-IDP Men Women Men Women Government Private Agriculture Private Non-Agriculture Construction Trade Food & Accm Other Services Other Hhld NGO/Int Org. Source: Authors’ estimates based on SIHBS 2022. 21 PART B: DEEP-DIVES - CHAPTER 2: LIVELIHOOD DEEP-DIVE 37. Individuals with other sources of income women’s participation in the labor force, especially often opt not to work, likely due to the lack for those from marginalized backgrounds. of opportunities and limited returns. The regression analysis also shows that remittances and 38. Remittances are more common among richer other working members in the household have a households but reduce the incentive to work. negative association with labor force participation. Inward remittances are equivalent to over 15 There is often a negative association between percent of Somalia's GDP and are high compared to remittances and female LFP in other countries, other low-income countries.³⁹ While a similar share but in Somalia, it is also negatively associated with of households in rural and urban areas received male LFP.³⁸ This suggests that if an individual can remittances, richer urban households most often achieve sufficient consumption without working, received remittances from abroad. Over a quarter they choose to remain economically inactive due of all international remittances sent went to the to the low returns and the limited number of richest urban households (Figure 39). In addition, available opportunities (Table 11). Focus group remittances from abroad are annually $200 more discussions also suggest that community norms on average, driven by larger transfer amounts restrict what type of activity an individual should rather than a larger number of transfers. Combining do, with some jobs being viewed as unsuitable for both domestic and international remittances, non- individuals from certain social groups. Security poor households received just under $300 more is also an impact factor, with an individual’s than poor households on average. In line with perception of the safety of public spaces having a previous studies, international remittances are also positive association with labor force participation correlated with higher school enrollment, especially for both men and women and those in urban areas for secondary schooling, and increased education (Table 12). Further, FGDs highlight how the fear of expenditure (Figure 40).⁴⁰ However, receiving gender-based violence is an important concern for remittances is negatively associated with LFP and Figure 39: Share of Remittances sent to… Figure 40: Enrollment by Remittances 30% 60% 25% 50% 20% 40% 15% 30% 10% 20% 5% 0% 10% Poorest 2nd 3rd 4th Richest Poorest 2nd 3rd 4th Richest Poorest 2nd 3rd 4th Richest 0% GER NER GER NER Rural Urban Nomadic Primary Secondary International Domestic None Domestic Only Any International Source: Authors’ estimates based on SIHBS 2022. ³⁸ Azizi 2018. ³⁹ Among low-income countries that had data for 2022 in the WDI database (19 out of 26 countries), Somalia has the 4th highest share of remittances as a percentage of GDP, only surpassed by The Gambia, Liberia, and Yemen. ⁴⁰ World Bank 2019. The positive relationship between remittances and education enrollment is also found in other countries: Bouoiyour and Miftah 2016; Ajefuand Ogebe 2021; Gyimah-Brempong and Asiedu 2015. 22 SOMALIA POVERTY AND EQUITY ASSESSMENT the operation of a household enterprise, suggesting income source across most of the rural and urban it reduces the incentive to work (Table 11 and Table consumption distribution (Figure 41). While the 14). The importance of remittances as an income prevalence of wage employment as a source source can be illustrated by the fact that removing of income shows only modest variations across remittance income from Somali households’ the consumption distribution, wage levels are expenditures would result in a 2 percentage point correlated with consumption. Indeed, median increase in the poverty headcount. household earnings from wages increase across the rural and urban consumption distribution, with Dualistic Wage Employment a stronger correlation in urban areas (Figure 42). Compared to low-income and MENA countries, 39. Wage employment is an important income Somalia has a high share of wage employment for its source for most households and accounts for income level (Figure 43). However, this appears to an unusually large share of employment. Wage be driven by the lack of other forms of employment, employment is the most important employment as the ratio of wage employment to the working- type in Somalia, employing the largest share of age population in Somalia is comparable to other individuals as well as being the most common low-income countries (Figure 44). Figure 41: Share of Households with Revenue Sources Figure 42: Median Annual Household Income from Wages 100% 4,000 80% 3,500 Annual USD (Median) 60% 3,000 40% 2,500 20% 2,000 0% 1,500 Poorest 2nd 3rd 4th Richest Poorest 2nd 3rd 4th Richest Poorest 2nd 3rd 4th Richest 1,000 500 0 Poorest 2nd 3rd 4th Richest Poorest 2nd 3rd 4th Richest Rural Urban Nomadic Wage Labor HH Enterprise Livestock Remittance Aid Rural Urban Source: Authors’ estimates based on SIHBS 2022 and World Development Indicators. Figure 43: LIC and MENA Wage Share of Employment Figure 44: Ratio of Wage Employment to Working-Age Population 100 Share of Wage Employment (%) 90 40% 80 35% 70 30% 60 25% 50 20% 40 30 15% 20 10% 10 5% 0 0% Niger (2018) South Sudan (2015) Benin (2018) Burundi (2013) Tajikstan (2013) Rwanda (2021) Senegal (2018) Zimbabwe (2017) Uganda (2016) Mali (2018 Nepal (2014) Comoros (2013) Liberia (2016) Gambia (2020) Somalia (2022) Madagascar (2012) Mozambique (2014) Guinea (2018) Togo (2018) Afghanistan (2013) Congo, Dem. Rep. (2018) Ethiopia (2021) Malawi(2019) Sierra Leone (2018) CAR (2008) Tanzania (2020) Burkina Faso (2009) 2.0 2.5 3.0 3.5 4.0 4.5 5.0 GDP per capita (Log) MENA LIC MENA & LIC Somalia (inc. Subsistence) Somalia (work-for-pay) Source: Authors’ estimates based on SIHBS 2022 and World Development Indicators. 23 PART B: DEEP-DIVES - CHAPTER 2: LIVELIHOOD DEEP-DIVE 40. The availability of better-wage jobs is extremely as outlined above. Respondents mentioned that hiring limited. As shown above, the regression analysis on for these better jobs is often done through connections, LFP suggests two motivations for employment, which which can, in turn, reduce accessibility. These better feed into a dual labor market: 1) a small pool of good jobs can be classified in the SIHBS data based on their jobs, which are typically only accessible to those with characteristics, namely whether they offer paid leave tertiary education, and 2) a larger pool, but still limited, and if they have a written contract. The scarcity of of low-quality, low-pay jobs which individuals take out these better jobs is demonstrated by the fact that they of necessity. The notion of a dual labor market was also account for only 14 percent of all employment and 1 mentioned during a focus group discussion in Mogadishu: percent of the working-age population. The reliance on one consisting of professional, white-collar jobs that are low-quality wage employment likely explains why having inaccessible to marginalized individuals and a second a wage earner in the household is negatively associated sector covering lower-skilled or service-orientated jobs, with poverty (Figure 14). Box 5: Guarantor system: How do lack of trust and clan dynamics create friction in Somalia’s urban labor markets? The “guarantor” practice involves trusted individuals vouching for someone’s background, character, and reliability for high-paying jobs. It serves as an informal security clearance to ensure the job candidate is not associated with terrorist groups such as al-Shabab and to establish trust in the safety of the institution’s assets and information. Historical conflicts in Somalia have led to stronger trust within clans than between different clans. While the guarantor practice serves its purpose, it, in turn, creates friction in the labor market, causing a delay between employees seeking new positions and employers looking for suitable candidates. Consequently, it takes longer for job seekers to find jobs that not only match their skills but also their clan dynamics, resulting in a high unemployment rate in Somalia. Additionally, this system leads to preferential employment within the same clan, limiting opportunities for those without established connections and those from minority clans. A 2022 report on youth unemployment reveals that 44% of youth cited nepotism and clannism as the main challenges in finding employment.⁴¹ Source: Focus Group Discussions and Heritage Institute. 41. The higher-quality jobs are better paid and percent), while over one-third are in the social sectors are offered by the government or private non- and around one-quarter are in the administrative agricultural employers. The higher-quality jobs are services sub-sector. Lastly, these jobs are the best concentrated in urban areas (79 percent) and are paid, with the largest median monthly and hourly largely occupied by men (78 percent). Furthermore, wages (Table 13). Further, although they account for under half of the employed are aged between 25 and a small share of overall employment, over half of all 34, the most educated age cohort. Over one-third of international organization jobs (66 percent) and NGO individuals in these better jobs have completed tertiary jobs (50 percent) are considered better-wage jobs, education, and 21 percent have completed secondary while 29 percent of all government jobs are better- education. Individuals from the richest 40 percent of quality wage jobs. In contrast, only 5 percent of jobs urban individuals accounted for around half of these with private non-agricultural employers and less than better jobs. Further, households with a member in 1 percent with other households are considered one of these better jobs have a much lower poverty better quality wage jobs. These two comparatively rate (32 percent). Most are either government jobs less favorable types of employers account for three- (33 percent) or private non-agricultural employers (35 quarters of all employment (Figure 45 and Figure 46). ⁴¹ Heritage Institute 2022. 24 SOMALIA POVERTY AND EQUITY ASSESSMENT Figure 45: Share of Wage Employment by Sector Figure 46: Share of Employment and Share of Jobs that are Better-Quality Wage Jobs by Employer 40% Share of Wage Employment 35% 30% International org. or a foreign embassy 25% NGO, non-profit institution, or mosque 20% 15% Other household(s)/individual 10% (ex: domestic worker) 5% Private non-agricultural entity 0% Private agricultural entity Mining & Utilities Construction Other Services Agriculture Manufacturing Admin Trade Transport Food ICT, Finance Social Government or state-owned enterprise (federal, state, local) 0% 10% 20% 30% 40% 50% 60% 70% Non-Formal Formal Better-Quality Share of Employment Source: Authors’ estimates based on SIHBS 2022. Box 6: The importance of aid 15 percent of households received a cash transfer in the 12 months before the survey, although the average value is relatively low. The share is largest among rural households (25 percent), followed by nomadic (24 percent), and a much lower share in urban households (9 percent). The share remains high across most of the rural and nomadic consumption distribution (Figure 47). The same pattern is also true for in-kind aid and charity. A larger share of cash aid is received by poor households compared to in-kind aid (Figure 48). However, the average value received by a household is relatively low at $255 for cash aid, which is only equivalent to around one-third of the per capita poverty line. Therefore, it is perhaps unsurprising that receiving aid does not harm labor force participation (Table 11) and its removal from household total consumption does not impact the poverty rate or poverty gap. Figure 47: Share of Households Receiving… Figure 48: Share of Recipients by Poverty Status 35% 100% % of Households Receiving 30% 25% 80% 20% 56% 62% 15% 60% 10% 5% 40% 0% Poorest 2nd 3rd 4th Richest Poorest 2nd 3rd 4th Richest Poorest 2nd 3rd 4th Richest 20% 38% 44% 0% Rural Urban Nomadic Cash In-Kind Cash In-Kind Charity Non-Poor Poor Source: Authors’ estimates based on SIHBS 2022. 25 PART B: DEEP-DIVES - CHAPTER 2: LIVELIHOOD DEEP-DIVE Household enterprises: Can they African countries suggests that HHEs owned by contribute to poverty reduction? more educated individuals and those in urban areas have greater job creation potential.⁴⁴ 42. Household enterprises are an important job creator in the labor market, despite being 43. Households receiving remittances or those relatively rare at the household level. As with a wage earner were less likely to have an shown above, HHEs account for 44 percent of enterprise. Households with a female head are employment.⁴² Paradoxically, despite accounting more likely to have an HHE, as are households for a large share of employment, only 14 percent with a head that has some education. In addition, of households had an HHE (Figure 49). This is IDP households less often had an enterprise. relatively low compared to other African countries. Households with domestic or international However, having a household enterprise is remittances or households with a wage earner are negatively associated with poverty (Figure 14).⁴³ also less likely to run an HHE. This suggests that Just under two-thirds of all household enterprises households do not use these other income sources were operated by non-poor households, and they to set up household enterprises (Table 14). are more often operated by women and individuals without any education (Figure 50). They, therefore, 44. Despite few household enterprises making offer alternative employment opportunities to sufficient profit to lift the household out of individuals who may struggle to gain access to poverty, they positively impact consumption. better jobs. They also have the potential to further Household enterprises operated by non-poor urban contribute to job creation. Evidence from other households are twice as productive on average Figure 49: Share of Households with an Enterprise Figure 50: Household Enterprise Owner Gender and Education 25% Male, 42% 20% 15% Poor, 38% Female, 10% 58% 5% Secondary, 11% Tertiary, 3% Primary, 0% 9% Non-Poor, All Non-Poor Poor All Non-Poor Poor All Non-Poor Poor All Non-Poor Poor 62% None, 77% All Rural Urban Nomadic Source: Authors’ estimates based on SIHBS 2022. ⁴² The share of employment by a household enterprise is calculated by summing total employment from the household enterprise module and dividing it by total employment. ⁴³ Evidence from other African countries also support the positive impact of HHEs and poverty reduction. Fox and Sohnesen 2016; Stifel 2010. ⁴⁴ Beegle and Christiaensen 2019. 26 SOMALIA POVERTY AND EQUITY ASSESSMENT than those operated by poor urban households.⁴⁵ a marketplace. Gender differences exist, with half of Male-owned household enterprises are also more female-owned HHEs operating from the household productive, as are those owned by individuals with relative to male-owned HHEs, which may limit at least completed secondary education, as found exposure to customers (Figure 52). Operating from in other African countries.⁴ However, less than 7 a marketplace has a positive association with larger percent of household enterprises make sufficient revenue per worker, although this association does profit per capita to lift the household out of not exist for profit (Table 17). This may be due to poverty poverty (Figure 51). Despite this, the higher operating costs from operating in the market presence of a household enterprise significantly area. Higher education and male ownership are increases household welfare (Table 15). Therefore, associated with larger revenues per worker, while household enterprises may be better suited to HHEs operated by poor households have lower complement other income sources at current revenues. Given that it is often the poor and women- productivity levels. However, this is rarely the case, owned enterprises who report lower revenue and with household enterprise income often being the profit, even after controlling for the location of only or largest source of income.⁴⁷ operation, other barriers may limit the success of their enterprises, such as the greater marginalization 45. Operating from a market is positively they are exposed to. They may also lack social and associated with revenue per worker but not political networks, especially women, who are often profit. Slightly more HHEs run by poor households excluded from these networks. For instance, focus operated from their household and less often from group discussions in Mogadishu highlighted that Figure 51: Share of Household Enterprises that report Figure 52: Household Enterprise Operating Location enough per capita profit for the household to be above the poverty line 100% 60% 50% 40% 50% 30% 20% 10% 0% Completed Tertiary All Non-Poor Poor All Non-Poor Poor All Non-Poor Poor Male Female None Incomplete Primary Completed Primary Completed Secondary 0% All Poor Rural Urban Male-Owned Female-Owned Non-Poor Mobile Road/Other Fixed All Rural Urban Owner Owner Education Sex Market/Industrial Site/Commercial Area Home Source: Authors’ estimates based on SIHBS 2022. ⁴⁵ The self-reported profit is used. When this is not available, profit is calculated using reported revenues and costs. In the instance revenue is not reported, it is imputed based on the enterprise and owner characteristics. ⁴⁶ Nagler and Naudé 2017 ⁴⁷ Just over a third of households with an enterprise solely rely on this income source, while for two-thirds it is the largest income source. Among these households, enterprise earnings accounted on average for 90 percent of their total income. 27 PART B: DEEP-DIVES - CHAPTER 2: LIVELIHOOD DEEP-DIVE clan dynamics often influence purchasing decisions. number of working-age individuals will reach 10.7 million by 2030. While not all working-age How can the poor be more individuals are expected to work, even to maintain productive, and what prevents the current employment-to-population ratio, on them from accessing productive average, just over 100,000 jobs must be created economic opportunities? annually until 2030. However, given Somalia’s low LFP and the importance of employment for poverty 46. Differences exist in how households engage reduction, maintaining the current employment-to- with the labor market across the consumption population ratio would be the minimum ambition. distribution. On initial inspection, there is little If the country were to increase its employment-to- difference in labor force participation, and wage population ratio to 30 percent by 2030, this would employment is the most common source of require an average of 200,000 jobs to be created employment for most. However, richer and urban each year, while reaching 40 percent would need households received remittances more often, just under 340,000 jobs a year (Figure 53). Therefore, particularly from abroad (Figure 39). In addition, while policies that help promote an environment for job employment in the services sector accounted for creation will be essential. most employment, individuals from poor households were more likely to work in low-productivity sectors 48. Improvements in education remain extremely such as agriculture, construction, and other services. important. As outlined above, literacy and enrollment Further, individuals from richer households worked are extremely low in Somalia, even relative to other for the government more often, and those from low-income countries. Further, the education of the poorer households worked for other households. household head is associated with lower poverty, as While private non-agricultural employers were the is literacy. Within the labor market itself, education largest source of employment, individuals from Figure 53: Required Job Creation, 2023-2030 richer households more often worked for larger- sized private non-agricultural employers. Finally, 600,000 while most households depend on a single worker, there are demographic differences, with poorer 500,000 households having fewer working members relative Number of New Jobs 400,000 to their size. 300,000 47. Due to Somalia’s demographic composition, 200,000 job creation will become an increasingly important issue for Somalia. In 2022, the pool 100,000 of individuals aged between 15 and 64, or the 0 2023 2024 2025 2026 2027 2028 2029 2030 working age, was around 7 million. As around half of Somalia’s population is 14 or under, the pool of working-age individuals will continually increase Current Emp to Pop (23%) Emp to Pop (30%) over time. For instance, based on the SIHBS, the Emp to Pop (40%) Empt to Pop (50%) 28 SOMALIA POVERTY AND EQUITY ASSESSMENT is often correlated with better outcomes, such as can still play an important role. For instance, employment in more productive sectors (social social protection programs can be utilized to help and administrative services), government and NGO increase the accumulation of human capital, while employment, and employment in larger private public works may offer opportunities for those who employers. Further, household enterprises with better- may initially lack the skills or experience to gain educated owners appear to be more productive. access to employment opportunities within the private sector. Lastly, interventions that promote 49. Government programs can also help the growth of household enterprises can help support areas relevant to employment. While contribute to job creation. These potential policies most employment and future job creation should will be discussed in more detail in Part C. come from the private sector, the government 29 PART B: DEEP-DIVES - CHAPTER 3: SHOCKS DEEP-DIVE CHAPTER 3: SHOCKS DEEP-DIVE 50. Somalia is particularly vulnerable to climatic more frequent, this chapter will focus on who is shocks, as highlighted by the recent multi- most exposed to climatic shocks, who is vulnerable season drought. Somalia has an arid to semi-arid to these shocks, and what households typically do in climate, with limited rainfall and high average response to these shocks. The chapter will conclude temperatures. As a result of its arid climate, with suggestions on how Somali households can over half of the country is suitable for extensive become more resilient to climatic shocks. nomadic pastoralism, while only 13 percent of the country’s total land area is suitable for cultivation. What shocks are commonly The country is also vulnerable to shocks due to reported by Somali households? its variable climate, which often has substantial consequences for the climate-dependent livelihood 52. Over two-thirds of households reported systems, such as livestock, that support most of a severe negative economic impact from any the population. Further, climate change is already shock in 2021 or 2022. After five consecutive failed impacting the country, with average temperatures rain seasons, the drought, which started in 2020, increasing, while the Gu rains have been declining was the longest and most severe in decades.⁵² in many parts of the country.⁴⁸ This is exemplified Meanwhile, food inflation reached unprecedented by the recent multi-season drought experienced levels in 2022, particularly for imported foods. For throughout 2021 and 2022, followed by extreme instance, the average annual inflation rate for cereal flooding in 2023.⁴⁹ Temperatures are forecast to prices in Mogadishu reached 69% in June 2022, up continue increasing, while changes in precipitation from 2% the previous year.⁵³ As a result, 68 percent are generally predicted to increase, although with a of households reported that they were severely much wider range of uncertainty.⁵⁰ negatively affected by a shock in 2021 or 2022, a similar share to that reported in 2017 following 51. This deep dive will look at households’ the 2016-2017 drought.⁵⁴ While poor households exposure, vulnerability, and coping strategies to were more often negatively affected by shocks, just droughts, and how these differ spatially and across under two-thirds of non-poor households also self- the consumption distribution. External shocks, such reported being negatively affected. Self-reported as drought or flooding, can have a negative impact exposure was largest in nomadic households, on a household’s ability to convert assets into market followed by rural and, lastly, urban households income.⁵¹ Improving households’ resilience to shocks (Table 18). Households in the south of the country, will be key to Somalia’s future poverty reduction. where agriculture is more common, are more often Therefore, given climatic shocks are likely to become reported to be negatively affected. ⁴⁸ World Bank 2023a. ⁴⁹ Reliefweb 2023. ⁵⁰ World Bank 2023a. ⁵¹ López-Calva and Rodríguez-Castelán 2016. ⁵² World Bank 2024c. ⁵³ IMF 2022. ⁵⁴ World Bank 2019. 30 SOMALIA POVERTY AND EQUITY ASSESSMENT Box 7: Poverty and Conflict Fragility and conflict harm economic activity. Al-Shabaab maintains territorial control in parts of southern and central Somalia. Despite a downward trend in conflict between 2018 and 2022, there was an uptick in 2021 due to fighting with Al-Shabaab in central and southern Somalia. Further, many shocks can intensify conflict as competition for scarce resources increases. Conflict can also create a vicious cycle whereby the lack of services and opportunities reinforces marginalization and breeds conflict, which in turn leads to further neglect, continued poverty, and marginalization.⁵⁵ Conflict in Somalia has also been shown to have a large negative short-term impact on consumption and, therefore, increasing poverty. The decline in consumption appears to be driven by a smaller share of household members working and earning income. By contrast, consumption for richer households appears to be unaffected.⁵⁶ In addition, conflict and insecurity have destroyed the enabling infrastructure required for domestic production and contributed to internal economic fragmentation, disrupting supply chains across the country and worsening food insecurity. Finally, areas with the potential for agricultural production are also some of the most affected by conflict.⁵⁷ Exposure to conflict is larger in the southern part of the country; however, it does not appear to be correlated with regional poverty rates. Banadir (96 percent) has the highest share of the population living within 5 km of a conflict between 2018 and 2022, likely partly due to a higher frequency of conflict and population density. Other regions with a higher share of exposure include Lower Shabelle (66 percent) and Lower Juba (61 percent) (Figure 54). However, at the regional level, there does not appear to be any correlation between poverty rates and exposure to conflict (Figure 55). Further, very few households reported a severe negative economic impact from conflict in 2021 or 2022, with higher exposure among the nomadic poor and urban non-poor (Figure 56). Figure 54: Exposure to Conflict, 2018-2022 Figure 55: Regional Poverty Rate and Exposure to Conflict 100 90 Share of Population within 5km of conflict 80 70 % (2018 - 2022) 60 .8 to 1 .7 to .8 50 .6 to .7 40 .5 to .6 .4 to .5 30 .3 to .4 20 .2 to .3 .1 to .2 10 0 to .1 0 0 20 40 60 80 100 % of population living within 5km from conflicts (2018 - 2022) (District) Poverty Headcount Source: Authors’ estimates based on SIHBS 2022 and Somalia PTI data. ⁵⁵ World Bank 2023b. ⁵⁶ Nunez-Chaim and Pape 2022. ⁵⁷ World Bank 2023b; World Bank 2021a. 31 PART B: DEEP-DIVES - CHAPTER 3: SHOCKS DEEP-DIVE price increases affected households equally across Figure 56: Self-Reported Household Exposure to location and poverty status. In contrast, drought- Conflict in 2021 or 2022 affected nomadic and poor households more 7% often. The same is true for livestock death, which is 6% concentrated among nomadic, with a larger share 5% among poor nomadic households (Table 18). 4% 3% 2% How are different areas of the 1% country affected by climatic shocks? 0% All Non-Poor Poor All Non-Poor Poor All Non-Poor Poor All Non-Poor Poor 54. Exposure to drought, floods, and heat is more common in poorer regions. Geospatial data All Poverty Nomadic Rural Urban can be used to estimate the number of households in a region exposed to climatic shocks. This allows Source: Authors’ estimates based on SIHBS 2022 and Somalia PTI data. for classifying a household as exposed to a shock regardless of whether they perceived themselves 53. Food price increases were the most common negatively affected. Regional poverty rates follow shock, followed by drought, which affected a similar pattern to the share of the population specific parts of the population more frequently. exposed to any climatic shock, with poorer regions Food price shocks affected just under half of Somali typically having greater exposure (Figure 57 and households in 2021 or 2022 (45 percent), followed Figure 58). Further, a much larger share of the by drought (35 percent) and then livestock death population is exposed to drought compared to (11 percent). All other types of shocks affected floods and heat (Figure 59, Figure 60, and Figure 61). less than 3% of households. Given the large share With around 38 percent of the population exposed of imported food in Somali diets, with domestic to drought, flood, or heat, Somalia has a similar share production satisfying only a fifth of per capita of exposure to neighboring low-income countries cereal needs, it is expected that imported food such as Ethiopia and Uganda but lower than Sudan inflation would affect most households.⁵⁸ The food and some West African low-income countries.⁵⁹ Figure 57: Poverty Rate Figure 58: Share of Population Exposed to Any Climate Hazard60 # % 80 to 90 .8 to 1 .7 to .8 70 to 80 .6 to .7 60 to 70 .5 to .6 50 to 60 .4 to .5 .3 to .4 40 to 50 .2 to .3 30 to 40 .1 to .2 No data 0 to .1 Source: Authors’ estimates based on SIHBS 2022, Worldpop, FAO, GFDRR, and Fathom (Version 3). ⁵⁸ IMF 2022. ⁵⁹ Doan et al. 2023. ⁶⁰ Definitions of exposure and data sources for each of the three climate hazards are detailed in the footnotes attached to the following figures. 32 SOMALIA POVERTY AND EQUITY ASSESSMENT Figure 59: Share of Population Figure 60: Share of Population Figure 61: Share of Population Exposed to Drought61 Exposed to Heat62 Exposed to Floods63 % % % .8 to 1 .8 to 1 .8 to 1 .7 to .8 .7 to .8 .7 to .8 .6 to .7 .6 to .7 .6 to .7 .5 to .6 .5 to .6 .5 to .6 .4 to.5 .4 to.5 .4 to.5 .3 to .4 .3 to .4 .3 to .4 .2 to .3 .2 to .3 .2 to .3 .1 to .2 .1 to .2 .1 to .2 0 to .1 0 to .1 0 to .1 Source: Authors’ estimates based on SIHBS 2022, Worldpop, FAO, GFDRR, and Fathom (Version 3). Did poor households more often between consumption and the likelihood of report a negative impact of drought? being impacted by drought is clearest among urban households and weakest among nomadic 55. Poorer households and regions more often households (Figure 62).⁶⁴ As a group, nomads are reported being negatively affected by the more likely to be affected by the drought. At the recent drought. At the household level, there is regional level, there is a weak positive correlation a clear correlation between the reported impact between poverty headcount and the share of of drought and poverty. The negative relationship households experiencing drought (Figure 63). Figure 62: Share of Households Affected by Drought by Figure 63: Poverty Headcount by Region and Share of Residency and Consumption Quintile, 2022 Households Affected by Drought, 2022 100% 100% 90% 80% 70% 80% Poverty Headcount 60% 50% 60% 40% 30% 40% 20% 10% 20% 0% Poorest 2nd 3rd 4th Richest Poorest 2nd 3rd 4th Richest Poorest 2nd 3rd 4th Richest 0% 0% 20% 40% 60% 80% 100% 120% Share of Household Experiencing Drought Rural Urban Nomadic Source: Authors’ estimates based on SIHBS 2022. ⁶¹ Drought hazard is defined as areas where at least 30% of cropland/grassland have experienced drought (VHI below 35) for the past 39 years (https://data.apps.fao.org/catalog/iso/f8568e67-46e7-425d-b779-a8504971389b). ⁶² % of people exposed to extreme heat (33C day max WBGT) (https://datacatalog.worldbank.org/int/search/dataset/0040194/ Global-extreme-heat-hazard) ⁶³ % of people exposed to river flood hazard (inundation depths of at least 50 cm during 1-in-100-year flood events). ⁶⁴ Even after removing self-reported IDPs, the poorest urban households reported the highest exposure. 33 PART B: DEEP-DIVES - CHAPTER 3: SHOCKS DEEP-DIVE 56. A negative NDVI shock at the district level poverty in rural and urban areas. An increase in the in the Deyr rainy season in 2021 had a negative Z-score of the NDVI shock in a district was associated impact on consumption and poverty.⁶⁵ Regression with lower total, food, and non-food consumption analysis shows that the larger the negative and increases in poverty, food poverty, and normalized difference vegetation index (NDVI) extreme poverty (Table 4). The negative impact of shock in November or December of 2021 was, the drought on households’ consumption and poverty larger the negative impact on consumption and is consistent with previous findings in Somalia.⁶⁶ Table 4: Regression Coefficients for NDVI Shock Z-Score on Monetary and Non-Monetary Indicators67 All Poor Non-Poor Rural Urban Consumption -0.673*** 0.023 -0.637*** -0.938** -0.849*** Food Consumption -0.446** 0.235 -0.530** -0.788* -0.750*** Non-Food Consumption -0.919*** -0.277 -0.459 -1.320* -1.074*** Food Insecurity 0.706 -1.413 1.933 3.399 1.438 Poverty 0.539*** 1.006** 0.632*** Food Poverty 0.489*** 0.751* 0.745*** Extreme Poverty 0.234** 0.919*** 0.300** Source: Authors’ estimates based on SIHBS 2022. Who are vulnerable to climate shocks? proxied using two aspects: the physical propensity to experience severe income, asset, or health loss 57. Almost all the Somali population exposed and the inability to cope with and recover from the to climatic shocks are likely vulnerable. While a losses (Table 5).⁶⁸ The population exposed to either household may reside in an area that experiences drought, heat, or flooding is multiplied by the share a climatic shock, some households may not be of the population who are considered vulnerable negatively affected due to their resilience or lack of in at least one dimension to get the share of the vulnerability. A household’s vulnerability could be exposed population who are vulnerable. ⁶⁵ A negative NDVI shock is based on the Z-score of the monthly district NDVI in relation to the monthly long-term average NDVI at the district level. The long-term average is defined as the monthly average between 2002 to 2023. The Z-score is inverted so a negative NDVI shock has a positive figure to improve the interpretation of the regression coefficients i.e. a negative coefficient means a larger negative NDVI shock is associated with a reduction in the dependent variable. ⁶⁶ Pape and Wollburg 2019. ⁶⁷ The impact of climatic shocks can be estimated using OLS. Weather shocks are exogenous variables, which implies the absence of endogeneity. Short-run deviations from long-run rainfall and temperature are plausibly exogenous (Nübler et al. 2021; Wineman et al. 2017). Therefore, OLS regressions can be used to estimate the impact of weather shocks (i.e., negative and positive rainfall shocks, temperature shocks and vegetation shocks) on household monetary and non-monetary welfare for 2022. ⁶⁸ Doan et al. 2023. 34 SOMALIA POVERTY AND EQUITY ASSESSMENT Table 5: Indicators to Measure Climate Vulnerability Vulnerability Area Dimension Physically Vulnerable Access to water The household has access to an improved water source in the dry season and the trip to collect water and return takes less than 30 minutes. Access to electricity Household has access to electricity. Unable to cope Low income Household’s per capita consumption is below 1.5 times the poverty line. Not covered by social protection The household did not report receiving aid or remittances. No access to finance The household did not have an adult with access to a bank/mobile money account. Low education The household does not have a member with at least completed primary education. Source: Doan et al. 2023. 58. Regardless of region, almost all exposed considered resilient to climatic shocks (Figure 65). households are also deprived in at least one dimension and, therefore, could be considered 59. Most households are considered vulnerable vulnerable. Relaxing the definition of vulnerability due to a lack of income, followed by education. to require a household to be vulnerable in On average, households are vulnerable in 2.6 three dimensions reduces the percentage of dimensions, with the lowest average among exposed who are vulnerable but widens regional urban households and the largest among nomadic differences. For instance, in Bakool, 88 percent households. There are also large regional differences, of the exposed are still considered vulnerable, ranging from 3.8 in Bakool to 2.1 in Banadir (Figure while this drops to 31 percent in Banadir (Figure 66). Just under three-quarters of all households are 64). Therefore, the share of the exposed but not vulnerable in the income dimension, followed by vulnerable population can be considered a proxy just under two-thirds in the education dimension for resilience. This is strongly correlated with (Figure 67). Sufficient education is important as it regional poverty, with poorer regions having enables a household to be more flexible in changing a larger share of exposed who are considered livelihoods, while sufficient income provides a buffer vulnerable, or rather a smaller share who could be to smooth consumption during shocks. Figure 64: Share of Exposed Population that are Figure 65: Regional Poverty Rates and Share of Exposed Vulnerable in 1 Areas Population that are Vulnerable in 3 Areas 100% 100% % of Exposed who are 80% 80% vulnerable Poverty Headcount 60% 40% 60% 20% 40% 0% Galgaduud Togdheer Middle Shabelle Bakool Hiraan Lower Juba Gedo Sanaag Woqooyi Galbeed Mudug Lower Shabelle Bay Sool Bari Nugaal Banadir Awdal 20% 0% 0% 20% 40% 60% 80% 100% % of Exposed who are vulnerable in 3 dimensions 1 Dimension 3 Dimensions Source: Authors’ estimates based on SIHBS 2022. 35 PART B: DEEP-DIVES - CHAPTER 3: SHOCKS DEEP-DIVE Figure 66: Average Number of Vulnerable Dimensions Figure 67: Share of Households Vulnerable in each dimension by Region 80% 4.0 70% % of households vulnerable 3.5 60% 3.0 50% 2.5 40% 2.0 30% 1.5 20% 1.0 10% 0.5 0% 0.0 Income Education Social Protection Financial Inclusion Water Electricity Galgaduud Mudug Bari Hiraan Togdheer Lower Shabelle Bakool Middle Shabelle Gedo Sool Bay Woqooyi Galbeed Nugaal Awdal Banadir Lower Juba Sanaag Source: Authors’ estimates based on SIHBS 2022. How Somali households cope with assets, regardless of the area of residency. However, climatic shocks? the ability to do so increases along the consumption distribution. By contrast, the share of households 60. Most households did not utilize any economic receiving assistance decreases with consumption in rural response after being affected by the drought. and urban areas. The poorest urban households more Over half of the households who were affected by the often reduced consumption or moved all or part of the drought took no economic action (this includes prayer, households, likely reflecting that a large share of IDPs are doing nothing, and other responses not classified at the bottom of the urban consumption distribution. elsewhere) in response (Figure 68). The most common Borrowing and increasing economic activities also form of economic response is to rely on savings or decrease with consumption in urban areas (Figure 69). Figure 68: Share of Households Responding to Figure 69: Share of Households with Economic Responses Drought, 202269 to Drought by Residency and Consumption Quintile, 2022 0% 20% 40% 60% 40% Non-economic responses 30% 20% Relied on savings/assets 10% Reduced consumption 0% Poorest 2nd 3rd 4th Richest Poorest 2nd 3rd 4th Richest Poorest 2nd 3rd 4th Richest Received assistance Borrowed Rural Urban Nomadic Moved all or part of the household Relied on savings/assets Moved all or part of the household Increased labor/firm supply Borrowed Increased labor/firm supply Received assistance Reduced consumption Source: Authors’ estimates based on SIHBS 2022. ⁶⁹ The sample for this section is households who reported having been affected by the drought shock in 2020, 2021 or 2022. “Non-economic responses” combines “Prayer”, “Do nothing” and “Other” responses not classified elsewhere. A detailed account of which survey responses are combined to create the categories discussed in this section is available in Annex 1. 36 SOMALIA POVERTY AND EQUITY ASSESSMENT 61. Reducing food consumption and displacement The relationship between displacement and were the most common maladaptive responses. climatic shocks Maladaptive responses may have long-term implications for welfare, such as harming 62. Climatic shocks are the main reason for human capital (removing children from school, internal displacement in Somalia in recent years. displacement, reduction in food consumption) or Between 2016 and 2023, on average, over 1 million supply capacity (selling out of land or livestock).⁷⁰ Somalis were displaced, with drought being the Reducing food consumption was adopted by 10 largest cause of displacement in 2017 and 2022 and percent of households affected by the drought, floods in 2020 and 2023 (Figure 71).⁷¹ Self-reported while 7 percent were displaced. By contrast, only 1 IDPs in 2022 most often cited droughts as the main percent of households sold out all land or livestock reason for displacement, with this reason being the that they owned in response to the drought, and most common among the poorest IDPs (Figure 72). less than 1 percent of households removed children from school to make them work. This is consistent 63. IDPs are poorer, both in terms of the poverty with low labor force participation rates, with most headcount and gap and are concentrated in the households having some inactive adults and low poorest urban quintile. Nationally, the poverty school enrollment rates outside of urban centers. rate is 52 percent among non-IDPs compared to Female-headed households were more often 72 percent among IDPs, while the poverty gap displaced or reduced food consumption, as were increases from 18 percent to 31 percent. Therefore, households with less educated heads (Figure 70). self-reported IDPs are more often poor, and when Figure 70: Share of Households with Common Maladaptive Responses to Drought by Household Characteristics, 2022 14% 12% 10% 8% 6% 4% 2% 0% All Nomadic Rural Urban Male Female Household Enterprises Agriculture Completed Tertiary Aid Remittances None Incomplete Primary Completed Primary Completed Secondary Labor All Residency Head sex Head education Main income source Reduced food consumption Displaced household Source: Authors’ estimates based on SIHBS 2022. ⁷⁰ Households were considered as having sold out of livestock if they owned <5 animals of any one type. This threshold was chosen based on the distribution of livestock ownership in the entire sample and literature on minimum viable flock size. ⁷¹ Data on displacement is from the Protection and Return Monitoring Network (PRMN) database, maintained by the UNHCR. (https://prmn-somalia.unhcr.org/) 37 PART B: DEEP-DIVES - CHAPTER 3: SHOCKS DEEP-DIVE they are poor, they are further, on average, from the would return if they had livestock assets. However, poverty line (Figure 73). The difference between the desire to return wasn’t uniform, with one IDPs and non-IDPs in urban areas is the widest (25 respondent able to establish and grow a small percentage points). Further, 31 percent of all IDPs business. Respondents also highlighted land tenure are in the poorest urban quintile, suggesting that issues and a lack of social capital, formal education, when IDPs do move to urban areas, they often fall at credit, and marketable skills as factors hindering the bottom of the urban consumption distribution their integration into urban areas. (Figure 74). How can households be more 64. A focus group discussion with IDPs in resilient to climatic shocks? Mogadishu highlighted that climatic shocks resulted in the loss of assets (farms and 65. Three broad methods can help promote livestock), which in turn forced them to flee economic resilience to climate change. Firstly, their homes. However, most IDPs declared a efforts can be made to reduce the biophysical impact desire to return to their original location, provided of climate change and extremes, for instance, by stability exists.⁷² Further, some stated the inability growing crop varieties that are more resilient to to return to their previous location without assets; drought. Secondly, interventions can moderate for instance, those who lost livestock stated they the socioeconomic consequences of these impacts, Figure 71: Displaced Individuals, 2016-2023 Figure 72: Reason for Being an IDP 2.0 100% 80% 1.5 60% 1.0 40% 20% All 2nd Non-Poor Poor Poorest 3rd 4th Richest 0.5 0% 0.0 2016 2017 2018 2019 2020 2021 2022 2023 All Quintile Conflict/Insecurity Drought Flood Other Conflict Drought Other Source: Authors’ estimates based on PRMN data. Source: Authors’ estimates based on SIHBS 2022. Figure 73: Poverty Headcount and Gap Figure 74: Distribution of IDPs 100% 35% 30% 80% 25% 60% 20% 40% 15% 10% 20% 5% 0% 0% Poorest 2nd 3rd 4th Richest Poorest 2nd 3rd 4th Richest Poorest 2nd 3rd 4th Richest Non-IDP IDP Non-IDP IDP Non-IDP IDP Non-IDP IDP All Rural Urban Nomadic Headcount Gap Urban Rural Nomadic Source: Authors’ estimates based on SIHBS 2022. ⁷² Federal Government of Somalia 2023. 38 SOMALIA POVERTY AND EQUITY ASSESSMENT such as encouraging alternative livelihood options, social protection, which will also benefit exposed increasing insurance uptake, or utilizing adaptive households. The latter is also made more relevant by social safety nets. Lastly, diversification towards the small share of drought-affected households who sectors that are less vulnerable to these shocks reported receiving assistance following the drought. can be considered. While economic growth can contribute to improvements in resilience, this needs 67. Households displaced by climatic shocks to be complimented by proactive and science- need support integrating into their new informed adaptation.⁷³ location. Climate-related shocks will remain a key driver of displacement, especially among the 66. Increased economic opportunity, education, poorest. These households typically move to urban and access to social protection can help improve areas and are concentrated among the poorest resilience. Poorer regions tend to be more exposed urban quintile, suggesting they have difficulty to climatic shocks and have a larger share of exposed finding economic opportunities in urban areas. This households who lack characteristics that may help is supported by the fact that IDPs have higher labor them deal with these shocks. Policies that help force participation than non-IDPs and often work households increase their market income will, in in low-quality and insecure jobs. Therefore, policies part, help households be more resilient to climatic that help better integrate displaced individuals into shocks while improving access to education and urban areas will be beneficial. Box 8: Spatial Inequalities Somalia – a country with an area of about 637,657 square kilometers – spans diverse agroecological zones, from pastoral land in the North to riverine farmland in the South. Simply looking at the national-level poverty rate masks important geographical variation and drivers for poverty reduction. Looking at the regional level maps, monetary and non-monetary poverty display a clear pattern, with lower poverty in the country's northern regions. The same pattern is true for non-monetary poverty: educational enrollment is low in the poorer regions in central and southern Somalia. On the other hand, access to the internet is high in regions with large cities around Mogadishu, Hargeisa, and Garowe. Figure 75: Poverty Rates Figure 76: Access to the Internet Figure 77: Primary Gross Enrollment % % 70 to 80 # 35 to 40 60 to 70 80 to 90 30 to 35 50 to 60 70 to 80 25 to 30 40 t0 50 60 to 70 20 to 25 30 to 40 50 to 60 15 to 20 20 to 30 40 to 50 10 to 15 10 to 20 30 to 40 0 to 10 0 to 10 No data No data No data ⁷³ World Bank 2023a. 39 PART B: DEEP-DIVES - CHAPTER 3: SHOCKS DEEP-DIVE What explains differences in welfare levels across different areas? Economic geography, a framework used to analyze spatial disparities in development, helps us understand how these disparities are connected to three major factors: density, distance, and division.⁷⁴ Figure 78: Population Density Figure 79: Distance75 Figure 80: Exposed to Climate Shock % # .8 to 1 250 to 12,000 # .7 to .8 100 to 250 .6 to .7 .6 to .7 50 to 100 .5 to .6 .5 to .6 40 to 50 .4 to .5 .4 to .5 30 to 40 .3 to .4 .3 to .4 20 to 30 .2 to .3 .2 to .3 10 to 20 .1 to .2 .1 to .2 5 to 10 0 to .1 0 to .1 0 to 5 Density. Urban agglomerations usually drive economic growth and poverty reduction. Somalia data shows high population density near large urban areas such as Mogadishu and Hargeisa. Such a pattern will continue to grow due to rapid urbanization and displacement. Given the urban nature of poverty, these northern regions still account for almost four in 10 of all the poor (Figure 4). Distance. Transportation plays a significant role in the spatial welfare gap in Somalia. The country's national transport infrastructure has suffered from a lack of investment and maintenance, with only 13% of its roads being paved. Transport prices for these routes vary widely and are among the highest in Africa, exceeding international benchmarks for developing countries (CEM, 2021). Access to markets may also affect the welfare of nomadic households (see Nomadic deep-dive chapter). Division. Somalia is one of the most vulnerable countries globally. The relative stability in the north is due to the greater homogeneity of the clans. In contrast, Southern Somalia is more densely populated and heterogeneous, with more communities competing for resources, leading to significant conflict (SCD 2023). Linkages between conflicts and poverty are shown in Box 7 and Figure 55. In addition, domestic market fragmentation due to illegal checkpoints and high transportation costs complicates logistics and dampens competitiveness (CEM, 2021). Climate risks. As noted earlier in the shocks deep-dive chapter, a larger portion of the population in central and southern Somalia is affected by drought, while a few northern districts are more prone to heat shocks. These shocks are linked to poverty patterns. A simple regression with poverty as the dependent variable and population density, Rural accessibility Index (RAI), and conflict as independent variables shows the expected relationships: greater population density and access to markets are associated with lower poverty, while greater conflict is associated with higher poverty (Table 19). ⁷⁴ “World Bank. 2009. World Development Report 2009: Reshaping Economic Geography. © World Bank. http://hdl.handle. net/10986/5991. ⁷⁵ Share within 2km of all season road. 40 SOMALIA POVERTY AND EQUITY ASSESSMENT CHAPTER 4: NOMADIC DEEP-DIVE 68. The movement of the nomadic population 69. The rest of this chapter will focus on what can played a key role in the stagnation of poverty be done to improve the welfare of the nomadic between 2017 and 2022. Between 2017 and population. The chapter will start by looking at the 2022, some of the nomadic population dropped characteristics of the nomadic population and their out and moved to urban areas, typically as IDPs, welfare. It will then look at the type of livestock demonstrated by the changes in the population owned, diversification, the use of inputs, output shares. The movement of these nomadic production, and commercialization across the households had a poverty-reducing impact on the nomadic consumption distribution to determine nomadic poverty rate, suggesting that these were whether there are differences between the non- poor households that had dropped out and moved poor and poor nomadic households. Further, given to urban areas. This movement also coincided with the movement of population from nomadic to an increase in inequality in nomadic areas, which urban areas and the likelihood that climatic shocks may have occurred as, during drought, richer will become more common, the chapter will also households often buy livestock at depressed prices look at urban IDPs and what can be done to better when the poorer households are forced to sell.⁷⁶ support their integration into urban areas. Box 9: The Nomadic Lifestyle Pastoral societies are found in environments Figure 81: Land Use Systems characterized by limited productive potential of the land, as well as scattered and highly variable precipitations. In such settings, nomadic livestock rearing offers better potential to exploit available resources than other productive systems (Figure 75).⁷⁷ This correlates with the Agropastoral concentration of the nomadic population in the Irrigated agropastoral northern regions (Figure 95). Mangroves Non-used forest A crucial element of Somali pastoral society is Oasis/frankincense its ability to spread risk based on principles of Pastoralism reciprocity and obligation among kin. Numerous Temporal waterbodies Kilometers mechanisms exist to provide livestock to kin- Urban area 0 155 310 620 mates in need, from outright gifts to interest-free loans of milking animals to credit paid-in livestock Source: World Bank 2023a. ⁷⁶ Aklilu and Catley 2009. ⁷⁷ Cossins 1985. 41 PART B: DEEP-DIVES - CHAPTER 4: NOMADIC DEEP-DIVE when the opportunity arises. Gifting of animals, especially camels, punctuates important life events, such as birth and marriage.⁷⁸ Attachment to nomadic pastoralism is driven by its cultural perception as “noble” and “pure”, in opposition to urban and agro-pastoralist communities which suffer broad political and social discrimination.⁷⁹ The political and cultural dominance of nomadic pastoralism dates back to the conquest of the Somali peninsula by northern clans.⁸⁰ While aspects of their hegemony were threatened by colonial rule, it was reasserted upon independence and strengthened during the Siad Barré era and the civil war, as the strongest nomadic clans struggled for political supremacy.⁸¹ What are the welfare conditions of the high rates of multidimensional poverty (91 the nomadic population? percent), with just under three quarters being both monetary and multidimensionally poor. There 70. The nomadic population suffers from the are also demographic differences, with nomadic highest monetary and non-monetary poverty. individuals having less education and being younger The nomadic population accounts for a relatively on average. small share of the population (11 percent), however, they display the highest poverty rate (78 percent). 71. However, there is a small group of nomadic The nomadic population also have the highest households who achieve higher levels of rates of extreme poverty and the poverty gap, consumption. This is demonstrated by the richest both of which increased between 2017 and 2022. nomadic quintile, which coincides with the non- Inequality was also the largest among nomadic poor nomadic households, who had consumption households and increased between 2017 and 2022. levels comparable to those of the richest rural In addition, they lag in literacy and enrollment quintile. The difference in consumption is also the rates, and have lower access to electricity and widest between the 4th and 5th quintiles among improved drinking water, which is reflected by nomadic households. Box 10: Drought and Nomadic Welfare The indicators presented in this chapter should be considered in the context of a prolonged drought. Somalia faced an unprecedented multi-season drought in 2022, which in turn resulted in the loss of livestock.⁸² Given the importance of livestock to nomadic households, this sub-population is likely to have been severely affected by the drought. The SIHBS 2022 data collection took place between May and July 2022, and therefore nomadic households had already likely been affected by the drought. For instance, in 2021 or 2022 70 percent of nomadic households reported a negative economic impact from the drought and 44 percent reported a negative economic impact from livestock death. Therefore, it ⁷⁸ Elmi 1989. ⁷⁹ Hill 2010. ⁸⁰ Lewis 1960. ⁸¹ Mukhtar 1996. ⁸² FEWS NET and FSNAU 2022. 42 SOMALIA POVERTY AND EQUITY ASSESSMENT is important that the data presented is considered with this context in mind. For instance, the average regional NDVI, a measure of vegetation, was far below the long-term average, while regions with a larger negative deviation from the long-term average NDVI had higher nomadic poverty rates (Figure 74 and Figure 75). Further, households with larger livestock holdings have been shown to be able to better smooth consumption during poor weather in pastoral zones in Western Africa, and so it is plausible there would be a similar effect in the Somali context.⁸³ Figure 82: Unweighted regional NDVI Figure 83: Correlation between Nomadic Poverty and NDVI Deviation84 3,000 4 Poverty Indicator with the national 2,800 2,600 3 poverty line NDVI 2,400 2 2,200 1 2,000 0 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 2000 -.25 -.2 -.15 -.1 -.05 0 Annual Average LTA % difference in NDVI from LTA Source: Authors’ estimates based on SIHBS 2022 and MODIS data. What enables these richer the median household having over 90 animals in nomadic households to achieve the richest quintile or around 14 TLUs and 3.8 TLU higher consumption? per capita (Figure 78). A stockless or near stockless household can be classified as one that has less than Larger herd sizes? 1 TLU per capita.⁸⁵ Across the nomadic population, 7 percent reported no livestock ownership, while 72. Most nomad households do not own enough 25 percent had 1 or less TLU per capita. Just under livestock to bring them out of poverty. There is little a quarter had a TLU per capita above 4.5, which difference in the number of livestock owned across can be considered a threshold for the “better-off” the bottom 60 percent of nomadic households, pastoralists (Figure 79).⁸⁶ As these stockless or with the median household owning over 50 animals. near-stockless households cannot produce food for The same applies to total tropical livestock units their consumption from livestock, they need cash (TLU) and TLU per capita. However, the number of earnings to survive. These cash earnings typically livestock increases for the top two quintiles, with take the form of aid or wage employment. ⁸³ Gascoigne, J. et al. 2024. ⁸⁴ The largest percentage deviation in the first 6 months of 2022 relative to the long-term average is used, at the district level. ⁸⁵ Little et al. 2008. ⁸⁶ Little et al. 2008 use this threshold in Northern Kenya. This is supported by similar sustainable herd sizes identified in other studies (Dahl and Hjort 1976; Lybbert et al. 2004; and Potkanski 2000). Nomadic households with a TLU per capita above 4.5 have the lowest poverty rate at 64 percent compared to 87 percent among those with 1 TLU per capita or less. 43 PART B: DEEP-DIVES - CHAPTER 4: NOMADIC DEEP-DIVE Figure 84: Median Livestock and TLU Ownership Figure 85: Nomadic Households by TLU per capita 100 16 100% 14 80 80% 12 60% 60 10 Animals TLU 8 40% 40 6 20% 4 20 0% 2 All Non-Poor Poor 0 0 Poorest 2nd 3rd 4th Richest All Poverty Livestock TLU (RHS) TLU p.c (RHS) None 1 or less 1 to 2 2 to 3 3 to 4.5 4.5 or more Source: Authors’ estimates based on SIHBS 2022. Box 11: Type of Pastoral Activities In line with McPeak and Little (2017), nomadic households can be classified into 4 groups based on their per capita TLU and per capita income. Just under half of nomadic households only own livestock and do not have other livelihoods (crops, wages, household enterprises, remittances, aid). A similar share has mixed livelihoods, owning both livestock and engaging in other livelihoods (Figure 80). Households with per capita TLU below the median and per capita income below the median are classified as “left out” as they have less access to herds and the cash economy. The second group is “moving from” as they have below the median per capita TLU but above the median per capita income, suggesting they are moving in a direction away from herd-based livelihoods and are occupying other areas of the economy. The third group, “staying with”, have below the median per capita income but above the median per capita TLU, as they have high levels of TLU but are not engaging with the cash economy as much as their peers. Lastly, the fourth group, “combining”, have above the median TLU per capita and income per capita, suggesting they are both involved in pastoralism and the cash economy (Table 6).⁸⁷ Just over a third of nomadic households are considered as combining, followed by a quarter as staying with, and around one-fifth as moving from or left out. However, the share in the staying with or combining categories increases across the consumption distribution, reaching over three-quarters among the richest compared to 44 percent among the poorest (Figure 81). Table 6: Classification of Pastoralist Type Per Capita Income Below median Above median Per capita TLU Below median 1) Left out 1) Moving from Above median 2) Staying with 2) Combining Source: McPeak and Little 2017. ⁸⁷ McPeak and Little 2017. 44 SOMALIA POVERTY AND EQUITY ASSESSMENT Figure 86: Type of Pastoralist Figure 87: Pastoralist Type based on per capita income and TLU 100% 100% 80% 80% 60% 40% 60% 20% 40% 0% 20% All Poor Poorest 2nd Non-Poor 3rd 4th Richest 0% All Poorest 2nd 3rd 4th Richest All Quintile All Quintile Livestock Only Mixed Non-livestock Only None Left out Moving from Staying with Combing Source: Authors’ estimates based on SIHBS 2022. Commercialization capita herd sizes (Figure 83). In fact, the differences in commercialization may also partly explain the 73. Richer nomadic households also appear increasing inequality in herd size as it often leads to better able to convert their livestock assets into the redistribution of livestock from smaller to larger consumption.⁸⁸ The poorest nomadic households, herds.⁹⁰ Milk production and herd accumulation is despite having a lower average value of livestock often a major production objective and therefore owned, have a larger ratio of livestock value to total pastoralists will organize their herds to meet this household consumption. For instance, on average objective. This may be supported by the fact that the poorest nomadic household has livestock which goat and sheep ownership is predominately female is valued 3.6 times larger than the total household animals, which favors milk production and herd consumption. In comparison, the average ratio growth, rather than animal trade (Figure 84).⁹¹ for the richest nomadic household is 1.6 (Figure 82). It is important to note, however, that nomadic Figure 88: Ratio of Livestock Value to Annual Consumption households often maximize milk production and 4.0 herd growth, which may in turn limit their potential 3.5 for the sale of livestock.⁸⁹ Most nomadic households 3.0 received revenue from the sale of animals, with 2.5 2.0 lower shares among households with below median 1.5 income (“Left out” and “Staying with”). The sale of 1.0 livestock output, such as milk, increases across the 0.5 0.0 bottom half of the consumption distribution, and is Poorest 2nd 3rd 4th Richest more common among households with larger per Nomadic capita herd sizes (“Staying with” and “Combining”). Average Median The sale of livestock is more common among the richer nomadic households, and those with larger per Source: Authors’ estimates based on SIHBS 2022. ⁸⁸ Livestock is valued according to the median selling price for rural, urban, and nomadic areas. ⁸⁹ Aklilu et al. 2013; Abdulahi 1990. ⁹⁰ Aklilu and Catley 2009. ⁹¹ McPeak and Little 2017. 45 PART B: DEEP-DIVES - CHAPTER 4: NOMADIC DEEP-DIVE Figure 89: Livestock Revenue Sources Figure 90: Average and Median Livestock Revenue 100% 1,600 1,400 80% 1,200 60% 1,000 USD 800 40% 600 20% 400 200 0% 0 Poorest 2nd 3rd 4th Richest Left out Poorest 2nd Moving from Staying with Combining 3rd 4th Richest Left out Moving from Staying with Combining Quintile Type Quintile Type Output Animal Slaughter Caring Average Median Source: Authors’ estimates based on SIHBS 2022. The importance of location? and 2022. During June 2022, while most districts were experiencing NDVI lower than the long- 74. Access to markets appears important for term average, there were a few exceptions with selling livestock for slaughter. There is a large minor positive deviations. For instance, NDVI in regional variation in the share of livestock-owning Xarardheere in Mudug was 21 percent below the households who sold at least one animal for long-term average, while Galdogob, also in Mudug, slaughter in the past 12 months, ranging from 52 was 6 percent above the long-term average. This percent in Bari to 8 percent in Bakool (Figure 85). highlights the climatic differences across regions The geographic location of the regions appears and even within regions (Figure 86). Given the important for accessing markets for slaughter. importance of vegetation for livestock grazing, For instance, the key export ports of Berbera and these variations likely impact nomadic households. Bossaso in Waqooyi Galbeed and Bari offer access Focus group discussions also highlighted the to Somalia’s most lucrative export markets of in negative impact climate change has had on the the Gulf Cooperation Council countries. In southern nomadic lifestyle.⁹³ Somalia, regions such as Gedo offer access to Kenya and Ethiopia through borders.⁹² Further, while the 76. A few regions had higher livestock revenue, nomadic poverty rates are extremely high, there even after controlling for herd size. Regression are some regions where the nomadic poverty rate is results show that relative to Awdal, nomadic lower relative to the nomadic average. For instance, households in Bari, Galgaduud, Gedo, Mudug, the nomadic poverty rate is below 70 percent in Sanaag, and Sool all had higher livestock revenue. Sanaag, Bari, and Galgaduud (Figure 94). However, livestock revenue is correlated with herd size, which was not constant across regions. Once this 75. Some regions suffered greater deviations is controlled for in the regression, only Galgaduud, from their long-term NDVI average. One rough Sanaag, and Sool had higher livestock revenue measure of the severity of the drought in June among nomadic households relative to those in 2022 is the degree to which NDVI differed from Awdal, all of which are located in the North of the its long-term average for June between 2000 country. In contrast, Bakool, Lower Juba, Middle ⁹² Hagmann and Stepputat 2016; Mahmoud 2010. ⁹³ The focus group discussion with nomadic individuals took place in Guriceel, Galguduud. 46 SOMALIA POVERTY AND EQUITY ASSESSMENT Shabelle, and Nugaal all had lower livestock revenue. reported lower revenue, this was no longer the case Similarly, while female-headed nomadic households once herd size was controlled for (Table 19). Figure 91: Share of Households with Livestock Selling an Figure 92: Percentage Deviation from the Long-Term Animal for Slaughter in the last 12 months NDVI in June 2022 % % .05 to .1 .6 to .75 0 to .05 .45 to .6 -.05 to 0 .3 to .45 -.1 to -.05 .15 to .3 -1.5 to -.1 0 to .15 -.2 to -.15 No data -.25 to -.2 Source: Authors’ estimates based on SIHBS 2022 and MODIS data. Box 12: Nomadic households and the Drought: Moving Up or Moving Out? Livestock ownership has declined most among the poorest nomadic households. In the context of severe drought, all nomadic quintiles reported a decline in the owned TLUs from 12 months before the survey. However, the decline was largest for households in the poorest quintile, for which the median declined by 42 percent of the initially owned TLUs. The “left out” and “moving from” households also experienced much larger declines in the median TLU despite having much lower initial TLU (Figure 87). The poorest quintile had the third-largest average number of TLUs per capita 12 months before the survey. Given that herd size is correlated with poverty, it is possible that these households lost many livestock, which resulted in a decline in consumption and hence being in the poorest quintile at the time of the survey (Figure 88). The reduction in TLU seems to be driven, at least in part, by higher mortality rates among poorer nomadic households for the main livestock types.⁹⁴ Further, the “Staying with” and “Combining” groups experienced lower median mortality rates. Finally, birth rates, which are often lower in drought periods, increase across the consumption distribution and are also larger for the “Staying with” and “Combining” groups for most animals.⁹⁵ ⁹⁴ The mortality rate is calculated as the number of livestock deaths over the past 12 months divided by the initial livestock ownership 12 months ago. ⁹⁵ Toulmin 1985; Otte et al. 2023. 47 PART B: DEEP-DIVES - CHAPTER 4: NOMADIC DEEP-DIVE Figure 93: Percentage Change in Median TLU from Figure 94: Change in Average TLU per capita Ownership96 12 months prior to the survey to date of interview 8 0% -5% -10% 6 -15% -20% -25% -30% 4 -35% -40% -45% 2 Poorest 2nd 3rd 4th Richest Left out Moving from Staying with Combining 0 12 months ago Survey Quintile Type Poorest 2nd 3rd 4th Richest This declining trend resulted in greater inequality in livestock ownership. Inequality in TLU increased among nomadic households, represented by an increase in the Gini for TLU from 0.51 to 0.54, coinciding with an increase in per capita household consumption inequality. The share of livestock owned by the poorest nomadic households decreased over the 12 months before the survey. What can the rest learn from the health can also positively impact productivity, richest nomadic households? as healthier livestock are likely to produce more output.⁹⁹ Based on previous work, a household needs 77. Supporting herd accumulation can have knock- at least 4.5 TLU per capita for mobility.¹⁰⁰ Around one- on effects in terms of resilience and productivity. quarter of nomadic households have the required per Having sufficient herd size allows greater mobility, capita TLU for mobility. Richer nomadic households which in turn can promote greater resilience. have larger total TLU and per capita TLU, which means Livestock that moves can potentially access a more they can more often benefit from increased mobility. diverse diet, which in turn improves their health. As The recent drought has exacerbated the differences a result, these animals are more resilient to climatic in TLU across the nomadic distribution. Therefore, shocks.⁹⁷ This is supported by higher mortality rates poorer nomadic households should be supported among sheep and goats among drought-affected in improving their ability to accumulate greater nomadic households and lower birth rates (Figure herd numbers and prevent livestock loss in the 89). Further, the increased mobility also enables these event of shocks. Land rights also play an important households to move away from areas suffering from role in enabling nomadic households to achieve drought. Greater herd sizes can also act as a buffer sufficient mobility.¹⁰¹ Further, focus group discussions to shocks, ensuring that households can restore their highlighted inter-clan conflict's negative impact on herd size after losses due to drought.⁹⁸ Improved nomadic mobility. ⁹⁶ This would not capture households who previously owned livestock 12 months ago but did not own any at the time of the survey. ⁹⁷ McPeak and Little 2017; Heritage Institute 2023. ⁹⁸ McPeak and Little 2017. ⁹⁹ Carter and Barrett 2013; Little et al. 2008; Abdulahi 1990. ¹⁰⁰ Focus groups discussions with nomads suggested that a herder would need at least 20 camels or 50 sheep or goats to have a viable herd size. Using this threshold also produces a larger share (69 percent) of nomadic households who meet this threshold. However, this does not account for household size. ¹⁰¹ Niamir-Fuller 2005; Homman, Rischkowsky and Steinbach 2004; Littleet al. 2008. 48 SOMALIA POVERTY AND EQUITY ASSESSMENT 78. Once a sufficient herd size is reached, to drought. A lack of access to water, fodder, and households can shift their focus to medicine can worsen the productivity of livestock, as commercialization. Households will often prioritize well as cause conflict due to the increased pressure herd accumulation over commercialization until on shared resources.¹⁰⁵ The unpredictable nature this point.¹⁰² Further, livestock also serves other of rain was stated as one factor that has made the social functions among the nomadic population, nomadic lifestyle harder.¹⁰⁶ Further, poor animal such as social insurance. The importance of this health can impact the viability of livestock exports, social insurance is highlighted by evidence from with serious implications for the entire Somali Turkana in Northern Kenya, which suggests that economy, as livestock is the country‘s main export.¹⁰⁷ herders fell into poverty not solely due to the loss of The share of households with expenditure on labor, animals but also due to the failure to establish social medicine, water, and fodder increases across the relations that provided support networks.¹⁰³ These consumption distribution, with water and medicine support networks can be established and developed being the most common. Some larger and wealthier by exchanging animal assets.¹⁰⁴ However, richer nomadic households can also use their influence households, i.e., those with larger TLU per capita, to gain better access to inputs.¹⁰⁸ Likewise, input more often sold animals for slaughter and more expenditure is most common among households often sold livestock output, which is facilitated by in the “Staying with” group, followed by the their larger herd size. “Combining” group (Figure 90). Improving access to these inputs can further help support resilience and 79. Improving access to key inputs can help prevent the need for distress sales during droughts improve livestock productivity and resilience when prices are low.¹⁰⁹ Figure 95: Mortality and Birth Rates by Drought Status Figure 96: Share of Livestock Owning Households with Expenditure on… 30% 100% 25% 80% 20% 15% 60% 10% 40% 5% 20% 0% 0% Mortality Birth Mortality Birth Mortality Birth Poorest 2nd 3rd 4th Richest Left out Moving from Staying with Combining Rates Rates Rates Rates Rates Rates Camels Sheep Goats No Drought Drought Labor Medicine Fodder Water Source: Authors’ estimates based on SIHBS 2022. ¹⁰² Carter and Barrett 2013; Little et al. 2008; Abdulahi 1990. ¹⁰³ Anderson and Broch-Due 2000. ¹⁰⁴ Little et al. 2008. ¹⁰⁵ World Bank 2023a. ¹⁰⁶ Based on a focus group discussion with nomadic individuals in Guriceel. ¹⁰⁷ World Bank and FAO 2018; World Bank 2021a. ¹⁰⁸ Aklilu and Catley 2009. ¹⁰⁹ Barrett, Bellemare, and Osterloh 2006. 49 PART C - POLICY RECOMMENDATIONS PART C: POLICY RECOMMENDATIONS 80. Somalia’s poverty rate remains high, with 81. Policy recommendations can be divided into two no poverty reduction in recent years. Economic areas: i) overarching economy-wide recommendations growth has failed to match population growth, and ii) sectoral-specific recommendations. Somalia with an annual average increase of 2 percent faces overarching constraints that cut across sectors. between 2019 and 2023, resulting in negative real Addressing these constraints can increase stability and GDP per capita growth.¹¹⁰ However, GDP growth is growth, which will also benefit many sectors of the estimated to be 3.7 percent in 2024 compared to economy and create the foundations for sustained 2.8 percent in 2023 as the economy recovers.¹¹¹ poverty reduction. However, there are also sectoral- The negative real GDP per capita growth coincided specific constraints related to the thematic focus with no change in the national poverty rate of this report that policies can address to promote between 2017 and 2022. With over half the Somali poverty reduction. It is also important that these population living in poverty, there is a pressing policy recommendations take into account Somalia’s need to ignite poverty reduction. Further, poverty salient features. For instance, Somalia is unusually increased in rural and nomadic areas between the urbanized for its income level, resulting from a history two years. Two key challenges that likely hinder of urban migration due to conflict and climate shocks, poverty reduction are the exposure to repeated with the bulk of the poor living in urban areas (Box climatic shocks and limited economic opportunities, 2). This movement towards urban areas will likely resulting in extremely low labor force participation continue, given the continued exposure to climatic rates, even for its income level, and especially shocks and the large share of near-stockless nomadic among women. This is reflected by the unclear households (Figure 114). This will further cement relationship between employment and poverty, the urban nature of poverty. Though the high level especially in urban areas. In addition, the labor of urbanization results from a series of shocks, it can market will come under increasing pressure due to be harnessed for poverty reduction and improved Somalia’s demographics. service delivery. ¹¹⁰ World Bank, 2024a. ¹¹¹ IMF 2024. 50 SOMALIA POVERTY AND EQUITY ASSESSMENT Table 7: Policy Recommendations Overarching recommendations Finding/Challenge Lack of economic growth and stability Potential Solutions Investment in basic infrastructure and services in border or transport corridors; improved governance measures to reduce multiple taxation; strengthen local community institutions; diversification of exports; continued development of the regulatory framework; continued development of social cohesion Sectoral recommendations Finding/Challenge Low levels of Low labor force High exposure to Livestock herds Marginalized groups human capital participation climate shocks not viable lagged behind Expand primary Boost demand for urban Improve management Adopt livestock Closing the gender gap enrollment low skilled workers of soil, water, and land insurance in secondary enrollment Potential Solutions Improve health Address women's Adopt climate- Improve access Improve access to services, esp. specific constraints to smart agricultural to key inputs services of IDPs family planning employment such as diversification risk of GBV and gender norms Diversify livelihoods Improve rangeland management Manage disaster risk Expand adaptive social protection Overarching need for greater economic growth reduce multiple taxation, and strengthening local and stability. community institutions can all improve cross-border 82. Stability and economic growth will be trade and potentially create jobs. Wholesale and important for sustained poverty reduction. The retail trade account for a quarter of all employment, limited economic integration can hinder economic and commerce activities are particularly large opportunities and worsen a country’s resilience. For among entrepreneurs. Product space analysis instance, the segmentation of domestic markets suggests there is scope for the diversification of increases costs for producers. This segmentation is exports from livestock to other products such as driven by transportation costs, which are largely due gums, resin, sesame, bananas, and fish, which again to poor infrastructure and multiple taxation points. has the potential to create additional employment In the short to medium term, investing in basic opportunities and may offer greater resilience infrastructure and services in border and transport among rural households.¹¹² This diversification corridors, improved governance measures to will be supported by improving infrastructure and ¹¹² Hansen et al. 2019. 51 PART C - POLICY RECOMMENDATIONS strengthening resilience to climatic shocks, especially Low labor force participation. through improved water management. In addition, 84. Utilize policies that can help increase labor the continued development of its regulatory demand, especially for low-skilled workers: The framework is needed to address constraints the bulk of the urban poor, and virtually all IDPs, have private sector faces.¹¹³ The continued development no formal education and are engaged in low-skill of social cohesion and trust will be important for employment with very low wages. The small share sustained economic growth. of higher quality employment, those that typically require education, are out of reach. Given this, policies Low levels of human capital. that positively shock labor demand for low-skilled 83. Human capital service delivery: In the medium labor are arguably the most effective way to increase term, sustained poverty reduction will require much earnings for the large pool of low-skilled labor. For higher levels of human capital, notably education. instance, policies promoting economic growth and Only 30 percent of the labor force has completed stability will increase labor demand indirectly. There is primary education, and enrollment remains also merit in exploring the feasibility of a large-scale exceptionally low (Figure 31). As most education urban public works program, which could act as a is fee-based, children from poor households are demand shock for low-skilled labor. Such programs largely excluded, with cost often cited as a reason are successful in other low-income countries.¹¹⁵ In for not attending or never attending school among addition, these public works could focus on improving poor households. This may reproduce poverty urban amenities and strengthening urban climate across generations. The fact that Somalia is highly resilience, such as creating drainage to counter urbanized in principle makes delivering public floods, planting trees to reduce urban temperatures, services more cost-effective as the population and other activities of a public nature with climate is concentrated in smaller areas. In addition, the benefits.¹¹⁶ Given the norms about social status and government must continue expanding the school low-wage work, only the ultra-poor would likely self- system to increase primary school enrollment, select into such programs, reducing the administrative whether through the public or private education cost of targeting such an intervention. system. If the latter, this can be subsidized for children from poor households. There should also 85. Female labor force participation. Policies be a focus on closing the gender gap in secondary that would increase overall labor demand must school enrollment. Expanding the education be complemented with interventions tailored system will also be a key factor in improving to women to help them gain equal access to medium to long-term productivity. Further, there employment opportunities. (i) Safety and GBV should be a focus on ensuring that skills needed in risk. The safety and risk of gender-based violence the economy, not just in existing sectors but also disproportionately affect women, hindering their in potential areas for growth, are provided by the mobility and economic opportunities. Evidence education system to reduce the skills mismatch from Bangladesh shows that women who feel safe between workers and jobs.¹¹⁴ are more likely to work, explore new opportunities, ¹¹³ World Bank 2021a. ¹¹⁴ Heritage Institute 2022. ¹¹⁵ Recent evidence from urban public works programs in low-income countries shows sizable benefits, increasing the welfare of the urban poor by 20 percent through direct benefits from participation in public works, indirect benefits from an increase in the economy-wide unskilled wage rate, and indirect benefits from improved urban amenities. Franklin et al. 2024. ¹¹⁶ Seetahul 2023. 52 SOMALIA POVERTY AND EQUITY ASSESSMENT and transition to higher-paying jobs in the service related agricultural activities.¹¹⁸ Increasing the sector.¹¹⁷ In Somalia, the labor force participation resilience of crop-related activities can also have regression shows that women who feel safe are 3.3 positive implications for food security.¹¹⁹ Other percentage points more likely to be economically adaptations include terracing, subsurface dams, active than women who feel unsafe after controlling and rainwater harvesting.¹²⁰ Adopting climate- for individual, and household characteristics. smart agricultural diversification, such as selecting Interventions such as adequate streetlights and more resilient crops or varieties, can help improve an enhanced law enforcement system can be resilience. Additional extension support, including adopted to make public spaces safer for women. (ii) agri-forecasts, climate-smart practices, and market Gender norms. Even after accounting for individual information, can further help with resilience and and household characteristics, female labor force productivity. Further, diversifying livelihoods into participation is still about 20 percentage points areas such as gums, resin, sesame, bananas, and lower than male. Gender norms may contribute to fish can promote new opportunities and offer this gap. Interventions targeting men's perceptions greater resilience among rural households. Such of women's work acceptability and promoting expansion will require the necessary infrastructure women's employment prospects have been to support the development of these activities effective in increasing women's job opportunities, outside of retail.¹²¹ especially in regions with large gender disparities like the Middle East, North Africa, and South Asia. 87. Disaster risk management and adaptive (See gender annex for more detail). social protection can help support rural and nomadic households in these areas. Those who Rural and nomadic households are highly exposed remain in rural or nomadic areas currently have to climate shocks. higher poverty rates, which have increased between 86. Further, the development of resilient rural 2017 and 2022, and are more vulnerable to drought livelihoods will be key. Managing key resources (Figure 74). Increased disaster risk management and such as soil and water will promote resilience integration into planning for key sectors can help against climatic shocks. While these decisions are improve preparedness for these climatic shocks. often private, public funding could still be used to Further, the continued use of Baxnaano to act as increase the uptake of such actions. For instance, an adaptive and scalable social safety net can help providing necessary digital and physical market buffer the poor from climatic shocks, as has been access infrastructure will play an important role. done in the past for drought and locusts, as well For water management, the immediate priority as potentially improving school enrollment, social is improving water availability through water cohesion, and mental health.¹²² Sustained financing infrastructure development. In addition, there will be important in line with fiscal sustainability. is scope for investment in irrigation systems, Similarly, improving the program's targeting to especially in areas with the potential for crop- minimize inclusion errors can help maximize the effectiveness of the limited available spending.¹²³ ¹¹⁷ Ahmed and Kotikula 2021. ¹¹⁸ Two areas show potential: a small area west of Hargeisa and a larger area between the Shabelle and Juba river valleys. Although, as only 13% of Somalia’s total land area is suitable for cultivation, this will not be an option for large employment gains. Giordano, Namara, and Bassini 2019. ¹¹⁹ Hansen et al. 2019. ¹²⁰ World Bank 2023a. ¹²¹ World Bank 2021a. ¹²² IMF 2023; d’Errico et al. 2020; Baird et al. 2013; Valli, Peterman, and Hidrobo 2019; de Milliano et al. 2021; and Kilburn et al. 2016. ¹²³ World Bank 2022b; Development Pathways 2022. 53 PART C - POLICY RECOMMENDATIONS Nomadic households often lack sufficient herd size. Cross-cutting themes 88. Specific policies can be introduced to help the 89. Policies and interventions should be resilience of livestock activities, which are key implemented with an understanding of local for nomadic households. Very few households are norms and beliefs. Failure to recognize how these engaged in crop production compared to livestock norms influence people's behaviors may result in keeping, which is especially common among the ineffective interventions. For example, FGDs with nomadic population. Maintaining a minimum herd nomadic individuals highlighted an unfavorable size is important as it enables households to support view of urban areas and a strong desire to remain their food consumption, as well as enables mobility in their traditional culture of livestock rearing. which can be beneficial for livestock health and Likewise, social norms are a barrier to female productivity. Restocking could be considered in the labor force participation, which can negatively instance of future droughts, targeting households impact overall economic growth. At the same whose livestock ownership is just below this time, members of high-ranking clans may avoid threshold, as interventions that do not lift pastoralist certain manual jobs. However, it has been shown households to the minimum viable herd size are that norms can change in Somalia. An impact likely to be unsuccessful.¹²⁴ Interventions that aim to evaluation of a program in Somalia indicates maintain herd size during shocks, such as veterinary that an intervention to change gender norms care, feeding, and water delivery, appear more cost- for young adolescents led to greater support for effective than interventions that seek to remove gender equality attitudes among girls and boys.¹²⁸ livestock from the system.¹²⁵ Further, increased Interventions encouraging people from different uptake of livestock insurance can help protect clans to interact and work together to ward goals livestock-owning households from livestock loss.¹²⁶ could also improve the level of trust in society and The improved water management mentioned above reduce friction in the labor markets. would also benefit livestock through improving their access to water, especially in times of shortages. 90. The theme of inclusivity needs to cut across Similarly, improved access to feed and fodder can policies. Marginalized groups still lag in multiple help protect livestock against drought and prevent dimensions of welfare. For instance, nomadic livestock loss.¹²⁷ The movement of livestock is an households or self-reported IDP households have important tool for improving livestock health and diet much higher poverty rates and more often work diversity and avoiding areas that may be experiencing in employment that offers low returns. Further, drought. Therefore, promoting effective rangeland women face unique challenges that often limit management and agreements on land user rights their engagement in the labor force. Thus, policies can help facilitate mobility among livestock-owning and interventions should ensure inclusion at the households, particularly the nomadic. planning and implementation stages. ¹²⁴ Little et al 2008; Santos and Barrett 2006. ¹²⁵ Little et al 2008; Morton et al. 2005; Catley 2007. ¹²⁶ World Bank 2022b; Hansen et al. 2019. ¹²⁷ World Bank 2023a. ¹²⁸ Brar et al. 2023. 54 SOMALIA POVERTY AND EQUITY ASSESSMENT REFERENCES Abdulahi A. M. 1990. Pastoral production systems in Africa: a study of nomadic household economy and livestock marketing in Central Somalia (Farming systems and resource economics in tropics Vol. 8.). Kiel: Wissenschafstverlag Vauk. Ahmed, Tanima; Kotikula, Aphichoke. 2021. Women’s Employment and Safety Perceptions: Evidence from Low-income Neighborhoods of Dhaka, Bangladesh. Washington, DC: World Bank. Ajefu, Joseph B., and Joseph O. Ogebe. 2021. "The effects of international remittances on expenditure patterns of the left-behind households in Sub-Saharan Africa." Review of Development Economics 25(1): 405-429. Aklilu, Yacob, and Andy Catley. 2009. "Livestock exports from the Horn of Africa: an analysis of benefits by pastoralist wealth group and policy implications." Medford (MA): Feinstein International Center, Tufts University. Aklilu, Yacob, Peter D. Little, Hussein A. Mahmoud, and John McPeak. 2013. "Market access and trade issues affecting the drylands in the Horn of Africa." Brief prepared by a Technical Consortium hosted by CGIAR in partnership with the FAO Investment Centre. Al-Ahmadi, Afrah, and Giuseppe Zampaglione. 2022. "From Protracted Humanitarian Relief to State-led Social Safety Net System: Somalia Baxnaano Program." Social Protection and Jobs Discussion Paper 2201, World Bank, Washington, DC. Ames, Brian, Ward Brown, Shanta Devarajan, and Alejandro Izquierdo. 2001. "Macroeconomic policy and poverty reduction." Pamphlet Series, International Monetary Fund, Washington, DC. An, Zidong, Tayeb Ghazi, Nathalie Gonzalez Prieto, and Aomar Ibourk. 2019. "Growth and jobs in developing economies: Trends and cycles." Open Economies Review 30: 875-893. Anderson, David M., and Vigdis Broch-Due. 2000. The poor are not us: poverty and pastoralism in Eastern Africa. Oxford: James Currey Publishers. Azizi, Soroosh. 2018. "The impacts of workers' remittances on human capital and labor supply in developing countries." Economic Modelling 75: 377-396. Baird, Sarah, Francisco H. G. Ferreira, Berk Özler, and Michael Woolcock. 2013. " Relative effectiveness of conditional and unconditional cash transfers for schooling outcomes in developing countries: a systematic review." Campbell systematic reviews 9(1): 1-124. Barrett, Christopher B., Marc F. Bellemare, and Sharon M. Osterloh. 2006. "Household-level livestock marketing behaviour among northern Kenyan and southern Ethiopian pastoralists." In Pastoral Livestock Marketing in Eastern Africa: Research and Policy Challenges, edited by John McPeak and Peter D. Little, 15-38. Rugby: Practical Action Publishing. Beegle, Kathleen, and Luc Christiaensen. 2019. Accelerating poverty reduction in Africa. Washington, DC: World Bank. Benhassine, Najy, Florencia Devoto, Esther Duflo, Pascaline Dupas, and Victor Pouliquen. 2015. "Turning a 55 REFERENCES Shove into a Nudge? A "Labeled Cash Transfer" for Education." American Economic Journal: Economic Policy 7(3): 86-125. Bolch, Kimberly, Luis F. Lopez-Calva, and Eduardo Ortiz-Juarez. 2023. "“When Life Gives You Lemons”: Using Cross-Sectional Surveys to Identify Chronic Poverty in the Absence of Panel Data." Review of Income and Wealth 69(2): 474-503. Bouoiyour, Jamal, and Amal Miftah. 2016. "The impact of remittances on children's human capital accumulation: Evidence from Morocco." Journal of International Development 28(2): 266-280. Brar, Rajdev, Niklas Buehren, Sreelakshmi Papineni, and Munshi Sulaiman. 2023. "Rebel with a Cause: Effects of a Gender Norms Intervention for Adolescents in Somalia." Policy Research Working Paper 10567, World Bank, Washington, DC. Bursztyn, Leonardo, Alessandra L. González, and David Yanagizawa-Drott. 2020. "Misperceived social norms: Women working outside the home in Saudi Arabia." American economic review 110(10): 2997-3029. Carter, Michael R., and Christopher B. Barrett. 2013. "The economics of poverty traps and persistent poverty: An asset-based approach." In Understanding and Reducing Persistent Poverty in Africa, edited by Christopher B. Barrett, Michael R. Carter and Peter D. Little, 12-33. Routledge. Catley, Andy. 2007 "Impact assessments of livelihoods-based drought interventions in Moyale and Dire Woredas." Briefing paper, Feinstein International Center, Tufts University, Medford (MA). Choudhry, Misbah Tanveer and Paul Elhorst. 2018. "Female labour force participation and economic development.", International Journal of Manpower 39(7): 896-912. Cossins, N. J. 1985. “The productivity and potential of pastoral systems.” ILCA Bulletin 21: 10-15. Dabalen, Andrew, Ambar Narayan, Jaime Saavedra-Chanduvi, and Alejandro Hoyos Suarez. 2014. Do African children have an equal chance? A human opportunity report for sub-Saharan Africa. Washington, DC: World Bank. Dahl, Gudrun, and Anders Hjort. 1976. Having herds: Pastoral herd growth and household economy. Stockholm: Department of Social Anthropology, University of Stockholm. Datt, Gaurav, and Martin Ravallion. 1992 "Growth and redistribution components of changes in poverty measures: A decomposition with applications to Brazil and India in the 1980s." Journal of development economics 38(2): 275-295. Davis, Benjamin, Stefania Di Giuseppe, and Alberto Zezza. 2017. "Are African households (not) leaving agriculture? Patterns of households’ income sources in rural Sub-Saharan Africa." Food policy 67: 153-174. de Milliano, Marlous, Clare Barrington, Gustavo Angeles, and Christiana Gbedemah. 2021 "Crowding-out or crowding-in? Effects of LEAP 1000 unconditional cash transfer program on household and community support among women in rural Ghana." World Development 143: 105466. d’Errico, Marco, Alessandra Garbero, Marco Letta, and Paul Winters. 2020. "Evaluating program impact on resilience: Evidence from Lesotho’s Child Grants Programme." The Journal of Development Studies 56(12): 2212-2234. Development Pathways. 2022. Targeting Evaluation of Somalia’s Shock-Responsive Safety Net for Human Capital Project. London: Development Pathways Ltd. 56 SOMALIA POVERTY AND EQUITY ASSESSMENT Doan, Miki Khanh, Ruth Hill, Stephane Hallegatte, Paul Andres Corral Rodas, Ben James Brunckhorst, Minh Nguyen, Samuel Freije-Rodriguez, and Esther G. Naikal. 2023. “Counting People Exposed to, Vulnerable to, or at High Risk From Climate Shocks—A Methodology.” Policy Research Working Paper 10619, World Bank, Washington, DC. Elmi, A. A. 1989. “Camel husbandry and management by Ceeldheer Pastoralists in Central Somalia.” Pastoral Development Network 27, Overseas Development Institute, London. Federal Government of Somalia. 2023. “Survey on Nomadic Movement into IDP Camps in Mogadishu, Kismayo, Beledweyne & Baidoa.” Somalia National Bureau of Statistics, Federal Republic of Somalia. Federal Government of Somalia and UNFPA. 2020. The Somali Health and Demographic Survey 2020. Somalia National Bureau of Statistics, Federal Republic of Somalia and United Nations Population Fund FEWS NET and FSNAU. 2022. “Somalia Food Security Outlook February to September 2022: Historic multi- season drought leads to Emergency (IPC Phase 4), with risk of further deterioration.” Famine Early Warning Systems Network and Food Security and Nutrition Analysis Unit. https://fews.net/east-africa/ somalia/food-security-outlook/february-2022 Fields, Gary S. 2012. "Poverty and Low Earnings in the Developing World." In The Oxford Handbook of the Economics of Poverty, edited by Philip N. Jefferson, 249-274. Oxford: Oxford University Press. Fox, Louise, and Thomas Pave Sohnesen. 2016. "Household enterprises and poverty reduction in sub- Saharan Africa." Development Policy Review 34(2): 197-221. Franklin, Simon, Clément Imbert, Girum Abebe, and Carolina Mejia-Mantilla. 2024. "Urban public works in spatial equilibrium: Experimental evidence from Ethiopia." American Economic Review 114(5): 1382-1414. Fratkin, Elliot, and Eric Abella Roth. 1990. "Drought and economic differentiation among Ariaal pastoralists of Kenya." Human Ecology 18: 385-402. Gascoigne, Jonn, Baquie, Sandra, Vinha, Katja Pauliina, Calcutt, Evie Isabel Neall, Kshirsagar, Varun Sridhar, Meenan, Conor, Hill, Ruth, Skoufias, Emmanuel. 2024. “The Welfare Cost of Drought in Sub-Saharan Africa.” Policy Research Working Paper 10683, World Bank, Washington, DC. Giordano, Mark F., Regassa Namara, and Elisabeth Bassini. 2019. The Impacts of Irrigation : A Review of Published Evidence. Washington, DC: World Bank. Gyimah-Brempong, Kwabena, and Elizabeth Asiedu. 2015. "Remittances and investment in education: Evidence from Ghana." The journal of international trade & economic development 24(2): 173-200. Hagmann, T., and Stepputat, F. 2016. Corridors of trade and power: Economy and state formation in Somali East Africa. Copenhagen: Danish Institute for International Studies. Hansen, James, Jon Hellin, Todd Rosenstock, Eleanor Fisher, Jill Cairns, Clare Stirling, Christine Lamanna, Jacob van Etten, Alison Rose, and Bruce Campbell. 2019. "Climate risk management and rural poverty reduction." Agricultural Systems 172: 28-46. Hardy, Morgan, and Gisella Kagy. "It’s getting crowded in here: experimental evidence of demand constraints in the gender profit gap." The Economic Journal 130, no. 631: 2272-2290. Heritage Institute. 2022. Youth unemployment and security in Somalia: Prioritizing Jobs for Achieving Stability. Mogadishu: The Heritage Institute for Policy Studies. 57 REFERENCES Heritage Institute. 2023. Climate Change and Conflict in the Horn: Challenges, Responses and New Mandates. Mogadishu: The Heritage Institute for Policy Studies. Hill, M. 2010. No redress: Somalia's forgotten minorities. London: Minority Rights Group International. Homann, Sabine, Barbara Rischkowsky, and Jörg Steinbach. 2004. "Herd mobility leads the way for sustainable pastoral development: the case of Borana rangelands, Southern Ethiopia." In Across disciplinary boundaries towards a sustainable life: psychodynamic reflection on human behavior, dedicated with eternal gratitude and in high esteem to Prof. Dr. Rainer Fuchs, edited by Elmar A. Stuhler and Shalini Misra, 183-196. Munich: Ranier Hampp Verlag. IMF. 2022. Somalia: Selected issues. Country Report 2022/376. Washington, DC: International Monetary Fund. IMF. 2023. Somalia Poverty Reduction Strategy Paper – Joint Staff Advisory Note. Country Report 2023/287. Washington, DC: International Monetary Fund. IMF. 2024. “IMF Staff Completes Staff-Level Agreement on the First Review of the Extended Credit Facility Arrangement for Somalia.” Press Release 24/77, March 8. Kilburn, Kelly, Harsha Thirumurthy, Carolyn Tucker Halpern, Audrey Pettifor, and Sudhanshu Handa. 2016. "Effects of a large-scale unconditional cash transfer program on mental health outcomes of young people in Kenya." Journal of Adolescent Health 58(2): 223-229. Klasen, Stephan. 2019. "What explains uneven female labor force participation levels and trends in developing countries?" The World Bank Research Observer 34(2): 161-197. Klasen, Stephan, Tu Thi Ngoc Le, Janneke Pieters, and Manuel Santos Silva. 2021. "What drives female labour force participation? Comparable micro-level evidence from eight developing and emerging economies." The Journal of Development Studies 57(3): 417-442. Lewis, I. M. 1960. “The Somali Conquest of the Horn of Africa.” The Journal of African History 1(02): 213-280. Little, Peter D., John McPeak, Christopher B. Barrett, and Patti Kristjanson. 2008. "Challenging orthodoxies: understanding poverty in pastoral areas of East Africa." Development and change 39(4): 587-611. Lopez-Calva, Luis F., and Carlos Rodríguez-Castelán. 2016. “Pro-Growth Equity: A Policy Framework for the Twin Goals.” Policy Research Working Paper 7897, World Bank, Washington, DC. Lowe, Matt, and Madeline McKelway. 2021. "Coupling labor supply decisions: An experiment in India." CESifo Working Paper 9446. Lybbert, Travis J., Christopher B. Barrett, Solomon Desta, and D. Layne Coppock. 2004. "Stochastic wealth dynamics and risk management among a poor population." The Economic Journal 114(498): 750-777. Mahmoud, H. A. 2010. “Livestock Trade in the Kenyan, Somali and Ethiopian Borderlands” Africa Programme Briefing Paper 2010/02, Chatham House/Royal Institute of International Affairs, London. McLean, Calum, Ludovico Carraro, Valentina Barca, and Laura Alfers. "Transfer values: how much is enough? Balancing social protection and humanitarian considerations." Social Protection Approaches to COVID- 19 Expert Advice Service (SPACE), DAI Global UK Ltd, United Kingdom. McPeak, John, and Peter D. Little. 2017. "Applying the concept of resilience to pastoralist household data." Pastoralism 7(1): 14. 58 SOMALIA POVERTY AND EQUITY ASSESSMENT Merotto, Dino Leonardo, Michael Weber, and Reyes Aterido. 2018 "Pathways to Better Jobs in IDA Countries: Findings from Jobs Diagnostics." Jobs Series 14, World Bank, Washington, DC. Morton, John, David Barton, Chris Collinson, and Brian Heath. 2005. Comparing drought mitigation interventions in the pastoral livestock sector. NRI (Natural Resources Institute) report, Greenwhich. Mukhtar, M. H. 1996. “The plight of the Agro-pastoral society of Somalia.” Review of African Political Economy 23(70): 543–553. Nagler, Paula, and Wim Naudé. 2017. "Non-farm entrepreneurship in rural sub-Saharan Africa: New empirical evidence." Food policy 67: 175-191. Niamir-Fuller, Maryam. 2005. "Managing mobility in African rangelands." In Collective action and property rights for sustainable rangeland management, edited by Esther Mwangi, 5-6. Consultative Group on International Agriculture, Washington, DC. Nübler, Laura, Karen Austrian, John A. Maluccio, and Jessie Pinchoff. 2021. "Rainfall shocks, cognitive development and educational attainment among adolescents in a drought-prone region in Kenya." Environment and Development Economics 26(5-6): 466-487. Nunez-Chaim, Gonzalo, and Utz Johann Pape. 2022. "Poverty and Violence: The Immediate Impact of Terrorist Attacks against Civilians in Somalia." Policy Research Working Paper 10169, World Bank, Washington, DC. Otte, Joachim, Yacob Aklilu, Dominik Wisser, Priti Rajagopalan and Zehra Zaidi. 2023. “Impact of the 2016/17 drought on Somali livestock keepers.” FAO Statistics Working Paper Series 23-37, Food and Agriculture Organization of the United Nations, Rome. Pape, Utz Johann and Philip Randolph Wollburg. 2019. “Impact of Drought on Poverty in Somalia.” Policy Research Working Paper 8698, World Bank, Washington, DC. Pennings, Steven Michael. 2022. A Gender Employment Gap Index (GEGI): A Simple Measure of the Economic Gains from Closing Gender Employment Gaps, with an Application to the Pacific Islands. Washington, DC: World Bank. Potkanski, Tomasz. 2000 "Mutual assistance among the Ngorongoro Maasai." In The Poor Are Not Us: Poverty and Pastoralism, edited by David M. Anderson and Vigdis Broch-Due, 199-217. Oxford: James Currey Publishers. Ravallion, Martin. 1994. "Measuring social welfare with and without poverty lines." The American economic review 84(2): 359-364. Ravallion, Martin, and Monika Huppi. 1991. "Measuring changes in poverty: A methodological case study of Indonesia during an adjustment period." The World Bank Economic Review 5(1): 57-82. Reliefweb. 2023. “Somalia: Floods - Oct 2023.” https://reliefweb.int/disaster/fl-2023-000190-som Rodrik, Dani. 2000. "Growth versus poverty reduction: a hollow debate." Finance and development 37(4): 8. Santos, P., and C. B. Barrett. 2006. "Heterogeneous wealth dynamics: On the roles of risk and ability.” Cornell University Working Paper. Schwartz, H. J., S. Shaabani, and D. Walther. 1991. Range Management Handbook of Kenya, Vol. II-1: Marsabit District. Ministry of Livestock Development, Republic of Kenya. 59 REFERENCES Seetahul, Suneha. 2023. Public Works Programs and Climate Change: What Can We Learn from the Literature and from Existing Programs? Washington, DC: World Bank. Somalia National Bureau of Statistics, 2023. Somalia Poverty Report. _____, 2024. Statistical Release: Gross Domestic Product (GDP) 2023. Stifel, David. 2010. "The rural non-farm economy, livelihood strategies and household welfare." African Journal of Agricultural and Resource Economics 4(1): 82-109. Toulmin, Camilla. 1985. "Livestock losses and post-drought rehabilitation in sub-Saharan Africa." Livestock Policy Unit Working Paper, International Livestock Centre for Africa, Addis Ababa. Valli, Elsa, Amber Peterman, and Melissa Hidrobo. 2019. "Economic transfers and social cohesion in a refugee-hosting setting." The Journal of Development Studies 55(sup1): 128-146. Wineman, Ayala, Nicole M. Mason, Justus Ochieng, and Lilian Kirimi. 2017. "Weather extremes and household welfare in rural Kenya." Food security 9(2): 281-300. World Bank. 2012. “Safety Nets: Public Works.” Africa Social Protection Policy Briefs, World Bank, Washington, DC. _____. 2019. Somali Poverty and Vulnerability Assessment. Findings from Wave 2 of the Somali High Frequency Survey. Washington, DC: World Bank. _____. 2021a. Somalia Country Economic Memorandum: Towards an Inclusive Jobs Agenda. Washington, DC: World Bank. _____. 2021b. Somalia Urbanization Review: Fostering Cities as Anchors of Development. Washington, DC: World Bank. _____. 2022a. Poverty and Shared Prosperity 2022: Correcting Course. Washington, DC: World Bank. _____. 2022b. Somalia Economic Update - Investing in Social Protection to Boost Resilience for Economic Growth. Washington, DC: World Bank. _____. 2023a. Somalia Climate Risk Review. Washington, DC: World Bank. _____. 2023b. Somalia - Systematic Country Diagnostic Update: Accelerating the Building of Inclusive Institutions for Resilience and Jobs. Washington, DC: World Bank. _____. 2024a. Macro Poverty Outlook for Somalia: April 2024. Macro Poverty Outlook, World Bank, Washington, DC. _____. 2024b. Poverty and Inequality Platform (version 20240326_2017). World Bank Group. Accessed June 2024. www.pip.worldbank.org. _____. 2024c. Somalia Economic Update, Eighth Edition: Integrating Climate Change with Somalia’s Development: The Case for Water. Washington, DC: World Bank. _____. Forthcoming (a). Are there differences in how displaced and host community women and men are faring in Somalia? Insights from the recent HFS and the Gender Dimensions of Forced Displacement research program. Washington, DC: World Bank. _____. Forthcoming (b). Gendered impact of the COVID-19 pandemic in Somalia: Results from the High Frequency Phone Survey of Households. Washington, DC: World Bank. 60 SOMALIA POVERTY AND EQUITY ASSESSMENT World Bank and FAO. 2018. Somalia: Rebuilding Resilient and Sustainable Agriculture. The World Bank and the Food and Agriculture Organization of the United Nations. Wu, Haoyu, Tom Bundervoet, Aziz Atamanov, and Pierella Paci. 2024. "The Growth Elasticity of Poverty: Is Africa Any Different?" Policy Research Working Paper 10690, World Bank, Washington, DC. Yoshida, Nobuo, Shinya Takamatsu, Shivapragasam Shivakumaran, Kexin Zhang, and Danielle Aron. 2022. “Poverty projections and profiling using a SWIFT-COVID19 package during the COVID-19 pandemic.” Paper prepared for the IARIW-TNBS Conference, “Measuring Income, Wealth and Well- being in Africa,” Arusha, November 12. 61 GENDER ANNEX GENDER ANNEX The gender gap in labor force participation rates (LFP) in Somalia is alarming. Women's LFP is persistently lower relative to men. Estimates from SIHBS 2022 indicate that Somalia has the lowest female labor market participation rates in Sub-Saharan Africa. Only 16 percent of women are engaged in paid labor compared to 41 percent of men, resulting in a gender gap of 26 percentage points. Compared to other low- income countries, Somalia has the lowest male LFP, the third lowest female LFP, and the fifth largest gender gap in LFP. However, it's noteworthy that the gender gap in LFP is relatively smaller in Somalia compared to similar low-income contexts such as Djibouti, Sudan, and Yemen. Regression analysis further supports these differences, with women significantly less likely to participate in the labor force than men, regardless of the area of residency (Table 14 and Table 15). Figure 97: Labor Force Status by Gender and Poverty Figure 98: Unemployment and Underemployment by Gender 100% 40% 80% 35% 30% 11% 60% 15% 19% 25% 40% 20% 9% 6% 12% 20% 15% 10% 23% 0% 20% 16% 17% 5% 14% 12% Female Non-Poor Poor Non-Poor Poor Male 0% Female Non-Poor Poor Non-Poor Poor Male All Male Female Employed Employed + Education Unemployed All Male Female Education Inactive Unemployed Underemployed Source: Authors’ estimates based on SIHBS 2022. Various socio-economic factors, including marital status, residence, IDP status, and household composition, shape the extent of gender inequality in labor force participation. The gender gap in LFP is exacerbated among the married population. While LFP rates are higher for married men than for women, women who are separated or divorced are much more likely to engage in paid work than men. LFP is also substantially higher for women living in urban areas and for IDP women. This is consistent with estimates from the HFS survey, which show that the share of adults working was higher in camps – 31 percent, compared to only 22 percent of women in the host community.¹²⁹ Further, the presence of another working member is negatively associated with women’s LFP but not with men’s. By contrast, larger household size is negatively associated with men’s LFP, but not women’s. For both men and women, age, higher education, household aid, and safety increase incentives for paid work, while remittance inflows reduce incentives for paid work. Estimates suggest that, like men, women with tertiary education have higher LFPs than those without education, consistent with estimates ¹²⁹ World Bank, forthcoming (a). 62 SOMALIA POVERTY AND EQUITY ASSESSMENT from other low-income countries.¹³⁰ There is no difference in LFP between women with primary or secondary education and those without education. Both men’s and women’s LFP rates are higher among the older population, households that received aid, and in safer areas. Lastly, as observed in other countries, remittance inflows are negatively associated with labor force participation regardless of gender. IDP women exhibit notably higher labor force participation rates compared to their counterparts outside of IDP households. Both men and women from IDP backgrounds are more actively engaged in the labor force compared to non-IDP individuals, but this is particularly evident among women, where the participation rate stands at 26 percent for IDPs versus 17 percent for non-IDPs. Additionally, IDPs face considerable unemployment rates. Women from IDP households are disproportionately represented in wage employment compared to non-IDP women, who are more frequently involved in household enterprises. Despite a similar proportion of wage employment between IDP women and men, a significant portion of IDP women are employed by other households, often within the services sector. Similarly, men from IDP backgrounds have a higher share of employment within other households compared to non-IDP men, although they also show a significant presence in the construction sector. These trends suggest that IDPs often find themselves compelled to occupy lower-skilled occupations due to comparatively lower levels of education and literacy. There are also large gender differences in employment sectors. Men are more likely to be employed in formal wage employment, especially among the poor. These formal wage jobs, including those in the government and private sector, typically offer more stable income, benefits, and job security. In contrast, female employment is more reliant on household enterprises. Women frequently engage in small-scale businesses, street vending, and home-based activities to generate income. This informal sector involvement could arise from limited access to formal employment opportunities due to factors such as gender discrimination, educational disparities, and cultural norms that prioritize men's participation in the formal workforce. There are also large gender differences in employment sectors. Women are more often employed in wholesale and retail trade, food and accommodation, and other services, while men more often work in construction, transport, and administrative services. Figure 99: Type of Employer by Gender and Poverty Figure 100: Type of Employment by Gender and Poverty 100% 100% 90% 90% 80% 80% 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Male Female Non-Poor Poor Non-Poor Poor Male Female Non-Poor Poor Non-Poor Poor All Male Female All Male Female Government Private Agriculture Private Non-Agriculture Wage HH Enterprise Owner Other Hhld NGO/Int Org. HH Enterprise Worker HH Agriculture Source: Authors’ estimates based on SIHBS 2022. ¹³⁰ Klasen et al. 2021. 63 GENDER ANNEX Women are more likely to be inactive. Over two-thirds of individuals who do not want to work are women. Regression analysis shows that women are more likely to be inactive, especially those who are married or those without any level of education. Conversely, women’s rates of inactivity are significantly lower among IDP households or households that received remittances (Table 16). Women, especially among the poor more often operate household enterprises, but male-owned household enterprises are twice as productive on average than female-owned household enterprises. Estimates suggest that 27% of the female working population are household enterprise owners compared to only 15% of men. However, female-owned enterprises are associated with lower revenue and profit per worker. Additionally, half of female-owned HHEs operate from the household, compared to male-owned household enterprises, which are more likely to operate from the market. Operating from the household may limit the number of accessible customers. Economic shocks¹³¹ are more likely to affect male-headed households compared to female-headed households. According to the 2021 Somalia Household Phone Survey (SHFPS),¹³² 64 % of male-headed households were affected by economic shocks, while the share of female-headed households affected by economic shocks was only 45 percent. This is likely driven by women being more involved in non-farm business activities, which are less prone to climate shocks. Notably, the same survey shows that female- headed households are also less susceptible to other shocks, including food price, health, natural disaster, and security shocks.¹³³ When it comes to coping strategies, male-headed households tend to rely more on assistance or loans from family and friends than female-headed households. About 54 percent of male- headed households report relying on assistance from family and friends as a coping strategy, while only 47 percent of female-headed households do the same. In contrast, female-headed households are more likely to rely on loans from financial institutions and NGO assistance. What works in other countries Obstacles hindering women's participation in Somalia's labor force encompass limited educational and skills training access, cultural expectations confining women to household roles, scant economic prospects, workplace gender bias, and security risks stemming from ongoing conflicts. Tackling these hurdles is pivotal for enabling women to actively seek and engage in employment. Although Somalia-specific data and research are scarce, insights from broader studies and experiences elsewhere can guide potential solutions. Various interventions aimed at husbands and extended families have effectively expanded women's job opportunities. In regions sharing gender disparities akin to Somalia, such as the Middle East and North Africa (MENA) and South Asia (SA), studies indicate that initiatives targeting men's perceptions of women's work acceptability or promoting women's employment prospects can boost their workforce participation.¹³⁴ ¹³¹ Economic shocks include (i) job loss, (ii) non-farm business closure, (iii) disruption of farming, livestock, and fishing activities, (iv) lack of availability of business/farming inputs, (v) increased price of farming/business inputs, and (vi) reduced price of farming/ business output. ¹³² World Bank, forthcoming (b). ¹³³ Food price shocks include (i) increase in price of major food items consumed. Natural disasters include (i) flooding, (ii) drought, and (iii) locust invasion. Security shocks include (i) theft/looting of cash and other property, and (ii) conflict or community violence. Health shocks include (i) illness, injury, or death of an income-earning member of household. ¹³⁴ Bursztyn, González, and Yanagizawa-Drott 2020; Lowe and McKelway 2021. 64 SOMALIA POVERTY AND EQUITY ASSESSMENT Empowering women by addressing demand-side barriers—like financial constraints, social norms, and childcare responsibilities—can also elevate female labor force participation. Examples include the World Bank's Liberia Economic Empowerment of Adolescent Girls and Young Women (EPAG) project and the Benin Youth Employment Project. Moreover, Somali women face challenges from climate shocks. Initiatives like the World Bank's efforts in the Democratic Republic of Congo, Ghana, and South Asia highlight the effectiveness of involving traditionally marginalized groups in climate action. These groups possess valuable traditional knowledge and practices, crucial for bolstering community resilience against climate-related disruptions. 65 ANNEX: CHAPTER 1 ANNEX: CHAPTER 1 WORKING-AGE AND HOUSEHOLD SIZE Table 8: Household Size, Working Age, and the Ratio of Working-Age to Members Av. Members Av. Working-Age Working Age / Members All All 6.7 3.0 45% Non-Poor 6.0 3.1 51% Poor 7.3 2.9 40% Nomadic All 6.0 2.6 43% Non-Poor 5.0 2.4 48% Poor 6.3 2.6 42% Rural All 6.2 2.6 42% Non-Poor 5.3 2.5 48% Poor 6.8 2.6 38% Urban All 7.0 3.3 47% Non-Poor 6.4 3.3 52% Poor 7.9 3.3 41% Nomadic Poorest 7.0 2.7 39% 2nd 6.3 2.5 39% 3rd 6.3 2.5 41% 4th 5.6 2.8 49% Richest 5.0 2.4 47% Rural Poorest 7.8 2.9 37% 2nd 6.6 2.4 36% 3rd 6.3 2.5 39% 4th 6.4 2.8 44% Richest 4.7 2.4 51% Urban Poorest 8.2 3.2 39% 2nd 7.8 3.4 43% 3rd 7.3 3.4 46% 4th 6.7 3.5 52% Richest 5.6 3.1 55% Source: Authors’ estimates based on SIHBS 2022. 66 SOMALIA POVERTY AND EQUITY ASSESSMENT Access to Electricity Figure 101: International Comparison in Access to Electricity 70 60 50 % of Popluation 40 30 20 10 0 Guinea-Bissau Liberia Burundi South Sudan Sudan Ethiopia Mali Rwanda Uganda Madagascar Mozambique Sierra Leone Congo, Dem. Rep. Burkina Faso Niger Central African Republic Malawi Gambia, The Somalia Togo Eritrea Chad Source: Authors’ estimates based on SIHBS 2022. MULTIDIMENSIONAL POVERTY DEFINITION Multi-dimensional poverty is defined in this analysis as follows: Table 9: Multidimensional Poverty Definition Dimension Indicator Definition Weight Food Insecurity Deprived if any member were hungry but did not eat because there was not enough 1/3 money or other resources for food or went without eating for a whole day because of a lack of money or other resources. Housing Basic Services Deprived if no access to electricity. 1/18 Characteristics Deprived if no access to improved water in the dry season. 1/36 Deprived if no access to improved water in the wet season. 1/36 Deprived if no access to improved sanitation. 1/18 Quality Deprived if floor is of poor quality 1/18 Deprived if roof is of poor quality 1/18 Deprived if cooking fuel is of poor quality 1/18 Education Deprived if the head never enroll in school 1/3 Source: Bolch, Lopez-Calva, and Ortiz-Juarez 2023. 67 ANNEX: CHAPTER 1 HUMAN OPPORTUNITY INDEX A child’s background often acts as a determinant to their access to an opportunity i.e., a good or service that should be universally available within society. These background factors may include gender of the household head, the education level of the household head, a family’s wealth status, ethnicity, or geographical location. These factors are referred to as circumstances. The idea is that circumstances should never determine whether a child has access to an opportunity. The Human Opportunity Index (HOI) unpacks existing inequalities by looking at the coverage rate of a particular opportunity accounting for distributional disparities amongst circumstance groups - clusters of individuals with the same set of circumstances. In other words, the HOI measures how circumstances influence a child’s access to different opportunities. The HOI methodology uses the dissimilarity index (D-Index) to measure inequality in access to an opportunity. It explores how a set of circumstances result in disproportionate access to an opportunity. The D-Index ranges between 0 and 1, where 0 indicates no inequality, and 1 indicates that the entire access to an opportunity is limited to a specific circumstance group e.g., males, children with educated parents and those living in urban areas. The methodology further decomposes the contribution of each circumstance through a Shapley decomposition which estimates the marginal contribution of each circumstance to inequality. Since the HOI is a function of a set of given circumstances, the Shapely decomposition is useful for understanding how each of the circumstances contributes to the inequality of opportunities. The formula of the human opportunity index is given as HOI=(1-D) × C where D is the inequality index and C is the coverage rate Dabalen et al. (2014) explored access to different opportunities in education, basic infrastructure services, health, and access to a bundle of basic goods and services in 20 Sub-Saharan Africa countries using DHS data. They found mixed results across countries and in some instances within a country in opportunities related to access to school attendance and those related to the quality of education (starting school on time and finishing primary school). Table 10: Definition and reference groups for various opportunities Opportunity Definition Reference Group Primary school attendance Child is currently enrolled in primary school 6 – 13 years Secondary school attendance Child is currently enrolled in secondary school 14 – 17 years Access to electricity Child lives in a household with access to electricity 6 - 18 years Improved source of drinking water Child lives in a household with access to improved source of drinking water. 6 - 18 years Improved source of sanitation Child lives in a household with access to improved source of sanitation. 6 - 18 years Source: Dabalen et al. 2014. 68 0 20 40 60 80 Secondary NER Coverage Primary NER HOI Currently attending school Improved sanitation Access to electricity SOMALIA POVERTY AND EQUITY ASSESSMENT Source: Authors’ estimates based on SIHBS 2022. Improved drinking water (dry) Improved drinking water (rainy) Figure 102: Coverage and Human Opportunity Index, 2022 69 0 5 10 15 20 25 30 Secondary NER Primary NER Currently attending school Figure 103: D-Index, 2022 Access to electricity Improved sanitation Improved drinking water (dry) Improved drinking water (rainy) ANNEX: CHAPTER 2 ANNEX: CHAPTER 2 LABOR FORCE PARTICIPATION REGRESSIONS Table 11: Labor Force Participation Regression, 2022135 All Men Women Rural Urban Nomadic Gender Male 0.000 0.000 0.000 0.000 (.) (.) (.) (.) Female -0.231*** -0.299*** -0.220*** -0.140*** (0.015) (0.039) (0.013) (0.030) Age 0.033*** 0.037*** 0.031*** 0.035*** 0.036*** 0.007 (0.002) (0.004) (0.003) (0.003) (0.003) (0.005) Age Squared -0.000*** -0.000*** -0.000*** -0.000*** -0.000*** -0.000 (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Household Size -0.005*** -0.010*** 0.000 -0.001 -0.006** 0.003 (0.002) (0.003) (0.002) (0.004) (0.003) (0.003) Marital Status Married 0.000 0.000 0.000 0.000 0.000 0.000 (.) (.) (.) (.) (.) (.) Divorced or widowed 0.024 -0.184*** 0.120*** 0.042* 0.022 -0.024 (0.018) (0.028) (0.020) (0.022) (0.026) (0.017) Never Married -0.099*** -0.322*** 0.061*** -0.117*** -0.091*** -0.095*** (0.015) (0.027) (0.015) (0.014) (0.020) (0.022) Education None 0.000 0.000 0.000 0.000 0.000 0.000 (.) (.) (.) (.) (.) (.) Incomplete Primary 0.026* 0.008 0.018 0.070** 0.019 -0.065 (0.013) (0.020) (0.011) (0.024) (0.016) (0.050) Complete Primary 0.007 -0.010 -0.020 0.056** -0.000 -0.041 (0.016) (0.019) (0.015) (0.025) (0.019) (0.070) Complete Secondary 0.048** 0.033 0.004 0.067 0.042** 0.136** (0.021) (0.025) (0.017) (0.078) (0.019) (0.063) Complete Tertiary 0.249*** 0.226*** 0.163*** 0.281** 0.236*** (0.027) (0.039) (0.039) (0.098) (0.028) Dependency 0.003 0.018** 0.002 -0.010 0.003 0.001 Ratio (0.004) (0.008) (0.005) (0.011) (0.005) (0.010) Rural 0.000 0.000 0.000 Residency (.) (.) (.) Urban 0.005 -0.013 0.023 (0.013) (0.021) (0.015) ¹³⁵ The results are from an OLS regression with the dependent variable being a dummy variable for labor force participation. 70 SOMALIA POVERTY AND EQUITY ASSESSMENT All Men Women Rural Urban Nomadic Nomadic -0.181*** -0.277*** -0.066** (0.034) (0.042) (0.029) No 0.000 0.000 0.000 0.000 0.000 0.000 IDP (.) (.) (.) (.) (.) (.) Yes 0.039 0.007 0.070** 0.080*** 0.029 0.001 (0.025) (0.026) (0.030) (0.024) (0.033) (0.028) No 0.000 0.000 0.000 0.000 0.000 0.000 Household (.) (.) (.) (.) (.) (.) Remittance Yes -0.118*** -0.135*** -0.073*** -0.143*** -0.130*** -0.026 (0.017) (0.018) (0.018) (0.041) (0.018) (0.035) No 0.000 0.000 0.000 0.000 0.000 0.000 (.) (.) (.) (.) (.) (.) Household Aid Yes 0.042*** 0.045*** 0.034** 0.002 0.047** 0.043* (0.013) (0.014) (0.015) (0.018) (0.017) (0.022) Other Working No 0.000 0.000 0.000 0.000 0.000 0.000 Member (.) (.) (.) (.) (.) (.) Yes -0.078*** -0.016 -0.037* -0.116*** -0.108*** 0.225*** (0.024) (0.023) (0.021) (0.031) (0.023) (0.037) Child in No 0.000 0.000 0.000 0.000 0.000 0.000 Household (.) (.) (.) (.) (.) (.) Yes 0.027* 0.018 0.002 0.058* 0.020 0.022 No (0.014) (0.016) (0.012) (0.031) (0.017) (0.024) Secondary City No 0.000 0.000 0.000 0.000 0.000 0.000 (.) (.) (.) (.) (.) (.) Yes -0.057* -0.097** -0.019 0.005 -0.056** -0.076 (0.031) (0.042) (0.023) (0.025) (0.023) (0.058) Poverty Non-Poor 0.000 0.000 0.000 0.000 0.000 0.000 (.) (.) (.) (.) (.) (.) Poor -0.007 -0.014 -0.013 -0.031 -0.013 0.025 (0.010) (0.016) (0.008) (0.022) (0.011) (0.033) Dummy Region Region Region Region Region Region Observations 21204 9774 11430 4878 14023 2303 R-squared 0.249 0.339 0.095 0.288 0.263 0.270 Adjusted 0.247 0.336 0.093 0.283 0.262 0.260 R-squared AIC 19872.612 9845.168 8501.543 4576.214 13224.993 1043.542 Source: Authors’ estimates based on SIHBS 2022. 71 ANNEX: CHAPTER 2 Table 12: Labor Force Participation Regression including perception of safety, 2022 All Men Women Rural Urban Nomadic Gender Male 0.000 0.000 0.000 0.000 (.) (.) (.) (.) Female -0.272*** -0.354*** -0.261*** -0.159*** (0.019) (0.042) (0.020) (0.033) Age 0.037*** 0.035*** 0.034*** 0.035*** 0.042*** 0.008 (0.003) (0.003) (0.003) (0.005) (0.003) (0.005) Age Squared -0.000*** -0.000*** -0.000*** -0.000*** -0.000*** -0.000 (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Household Size -0.007*** -0.014*** 0.000 -0.002 -0.008** 0.001 (0.002) (0.004) (0.002) (0.004) (0.003) (0.003) Marital Status Married 0.000 0.000 0.000 0.000 0.000 0.000 (.) (.) (.) (.) (.) (.) Divorced or widowed 0.038** -0.187*** 0.121*** 0.063** 0.040 -0.026 (0.018) (0.029) (0.020) (0.023) (0.026) (0.016) Never Married -0.112*** -0.320*** 0.055*** -0.135*** -0.098*** -0.109*** (0.016) (0.029) (0.018) (0.016) (0.022) (0.026) Education None 0.000 0.000 0.000 0.000 0.000 0.000 (.) (.) (.) (.) (.) (.) Incomplete Primary 0.044*** 0.044* 0.035** 0.098*** 0.023 -0.013 (0.015) (0.022) (0.013) (0.032) (0.015) (0.047) Complete Primary 0.025 0.018 -0.012 0.074** 0.014 -0.070 (0.019) (0.020) (0.019) (0.030) (0.023) (0.076) Complete Secondary 0.060** 0.044 0.022 0.069 0.055** 0.141** (0.023) (0.028) (0.019) (0.094) (0.022) (0.063) Complete Tertiary 0.246*** 0.226*** 0.176*** 0.287** 0.236*** (0.028) (0.042) (0.042) (0.100) (0.029) Dependency 0.002 0.021** 0.000 -0.011 0.001 0.006 Ratio (0.005) (0.008) (0.005) (0.011) (0.006) (0.013) Residency Rural 0.000 0.000 0.000 (.) (.) (.) Urban 0.006 -0.018 0.029* (0.014) (0.021) (0.016) Nomadic -0.209*** -0.333*** -0.078** (0.038) (0.045) (0.032) IDP No 0.000 0.000 0.000 0.000 0.000 0.000 (.) (.) (.) (.) (.) (.) Yes 0.046* 0.012 0.075** 0.103*** 0.032 0.012 (0.024) (0.024) (0.030) (0.033) (0.031) (0.022) Household No 0.000 0.000 0.000 0.000 0.000 0.000 Remittance 72 SOMALIA POVERTY AND EQUITY ASSESSMENT All Men Women Rural Urban Nomadic (.) (.) (.) (.) (.) (.) Yes -0.131*** -0.155*** -0.084*** -0.147*** -0.146*** -0.007 (0.019) (0.019) (0.022) (0.045) (0.020) (0.038) Household Aid No 0.000 0.000 0.000 0.000 0.000 0.000 (.) (.) (.) (.) (.) (.) Yes 0.036** 0.033** 0.029* -0.001 0.044** 0.033 (0.014) (0.016) (0.016) (0.020) (0.017) (0.028) Other Working Member No 0.000 0.000 0.000 0.000 0.000 0.000 (.) (.) (.) (.) (.) (.) Yes -0.078*** -0.008 -0.045* -0.103*** -0.114*** 0.254*** (0.024) (0.026) (0.023) (0.031) (0.023) (0.057) Child in Household No 0.000 0.000 0.000 0.000 0.000 0.000 (.) (.) (.) (.) (.) (.) Yes 0.025 0.018 0.003 0.060* 0.019 -0.006 (0.015) (0.020) (0.012) (0.029) (0.020) (0.018) Feel Safe in Public No 0.000 0.000 0.000 0.000 0.000 0.000 Spaces (.) (.) (.) (.) (.) (.) Yes 0.033** 0.031* 0.033* 0.000 0.048** 0.011 (0.013) (0.017) (0.018) (0.023) (0.017) (0.037) Dummy Region Region Region Region Region Region Observations 17761 8120 9641 4129 11678 1954 R-squared 0.244 0.300 0.089 0.284 0.252 0.286 Adjusted R-squared 0.242 0.297 0.085 0.278 0.250 0.275 AIC 18246.684 8931.167 8211.040 4193.103 12172.506 973.667 Source: Authors estimates based on SIHBS 2022. 73 ANNEX: CHAPTER 2 Characteristics of “Better Jobs” Table 13: Characteristics of Better jobs Formal Non-Formal Armed Other Individual Characteristics Residency Rural 25% 13% 21% Urban 72% 87% 79% Nomadic 3% 0% 0% Sex Male 76% 97% 78% Female 24% 3% 22% Age 15-24 16% 1% 7% 25-34 32% 25% 45% 35-44 30% 37% 28% 45-54 16% 13% 11% 55-64 6% 24% 7% Education None 61% 44% 27% Incomplete Primary 9% 11% 6% Primary 9% 23% 11% Secondary 11% 21% 21% Tertiary 9% 0% 35% Urban Quintile Poorest 16% 2% 5% 2nd 13% 24% 13% 3rd 15% 12% 11% 4th 13% 36% 25% Richest 15% 13% 25% Employment Employer Government 10% 82% 33% Private Agriculture 10% 0% 7% Private Non- 54% 12% 35% Agriculture Household 23% 0% 1% NGO 2% 4% 15% Int Org 1% 2% 9% Sector Agriculture 7% 0% 1% Mining & Utilities 2% 0% 1% Manufacturing 3% 0% 2% Construction 15% 13% 3% Trade 17% 0% 5% Transport 15% 6% 7% Food 4% 0% 2% ICT, Finance 3% 0% 8% Admin 9% 81% 24% Social 10% 0% 35% Other Services 17% 0% 13% 74 SOMALIA POVERTY AND EQUITY ASSESSMENT Formal Non-Formal Armed Other Employment Earnings ($) Av. Month 321.48 256.72 2.34 Av. Hour 2.79 1.14 300.00 Med Month 240.00 200.00 1.88 Med Hour 1.25 0.87 Source: Authors’ estimates based on SIHBS 2022. Household Enterprises Table 14: Marginal Effects from Probit Regression on Household Having an Enterprise Coefficient SE P Score Head Gender Female 0.031 0.009 0.001 Head Education Incomplete Primary 0.053 0.025 0.037 Complete Primary 0.031 0.017 0.064 Complete Secondary & Post-Secondary 0.052 0.024 0.031 Completed University 0.040 0.034 0.241 Location Urban -0.012 0.015 0.437 Nomadic -0.126 0.014 0.000 IDP Non-IDP 0.080 0.017 0.000 Income Aid -0.002 0.016 0.911 Domestic Remittances -0.101 0.007 0.000 International Remittance -0.070 0.009 0.000 Composition Household Size 0.008 0.002 0.000 Dependency Ratio -0.004 0.005 0.381 Wage Earner -0.081 0.013 0.000 Finance Member has bank account -0.066 0.018 0.000 Source: Authors’ estimates based on SIHBS 2022. Given the presence of a household enterprise is an endogenous regressor in the household welfare equation, the estimation of the welfare impact of having a household enterprise utilizes a maximum likelihood estimator with an endogenous regressor under the Recursive Bivariate Regression (RBR) model. This approach will improve the causal inference on the effect of a household enterprise on a households welfare. The RBR jointly determines equations as a system of two equations that allows the error terms to be correlated, and the household enterprise variable is an endogenous regressor in the equation determining household welfare. This will enable the estimation of the average welfare effect of having a household enterprise. The choice of population density as an exclusion restriction is validated by running regressions to determine its relevance in explaining variations in household enterprise ownership and whether or not it has a direct effect on household welfare. As shown in table 15, population density has a significantly negative association with owning a household enterprise. This is likely as in areas that have greater population 75 ANNEX: CHAPTER 2 density, have greater employment opportunities and therefore there is less reliance on household enterprises. Table 16 further explores whether population density has a statistically significant direct effect on household welfare beyond its effect through ownership of a household enterprise, for which there is none. Table 15: Determinants of Having an Enterprise and its impact on household welfare Has Enterprise Log Consumption Per Capita Population Density -0.0017* (0.0010) Has Enterprise 0.6306*** (0.0598) Female Head 0.1245*** 0.0173 (0.0424) (0.0136) Head Education: None - - - - Head Education: Incomplete Primary 0.2719*** 0.1063*** (0.0668) (0.0232) Head Education: Primary 0.1363* 0.1843*** (0.0735) (0.0246) Head Education: Secondary 0.2709*** 0.3128*** (0.0776) (0.0270) Head Education: Tertiary 0.0666 0.5248*** (0.1072) (0.0361) Head Age 0.0054*** -0.0000 (0.0014) (0.0005) Area: Rural - - - - Area: Urban 0.0144 0.3230*** (0.0471) (0.0157) Area: Nomadic -0.9906*** -0.3254*** (0.1015) (0.0233) Non-IDP 0.4429*** 0.2869*** (0.0772) (0.0233) Aid 0.0361 -0.0043 (0.0443) (0.0145) Domestic Remittance Value -0.0004*** 0.0000** (0.0001) (0.0000) International Remittance Value -0.0001*** 0.0001*** (0.0000) (0.0000) Household Size 0.0644*** -0.0949*** (0.0076) (0.0028) Wage Earner -0.3462*** -0.0027 (0.0465) (0.0145) Walking time to nearest city -0.0001** -0.0001*** (0.0001) (0.0000) 76 SOMALIA POVERTY AND EQUITY ASSESSMENT Has Enterprise Log Consumption Per Capita Constant 17.4024 6.4063*** (12.0476) (0.0633) Observations 7,156 7,156 Source: Authors’ estimates based on SIHBS 2022. Table 16: Exploring the direct effect of household enterprise on household welfare (1) pcerl_coef_popd VARIABLES l_pcer has_enterprise 0.190*** (0.0260) Pop Density -1.30e-06 (5.54e-06) city_wlk -0.000287 (0.000203) city_mtr 0.00274 (0.00180) RECODE of pop_wp (pop_wp) 0.0582** (0.0229) rai 0.00324* (0.00169) RECODE of hh1_17 (How many kilometers is this house from the -0.0718*** nearest all-season (0.0150) Constant 6.293*** (0.0987) Observations 7,156 R-squared 0.043 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Regression estimation Source: Authors’ estimates based on SIHBS 2022. 77 ANNEX: CHAPTER 2 Table 17: Household Enterprise Regressions Profit Profit Per Worker Revenue Per Worker Poverty Non-Poor 0.000 0.000 0.000 (.) (.) (.) Poor -146.306* -78.706* -0.433*** (82.09) (42.60) (0.12) Household Size 1-3 HH members 0.000 0.000 0.000 (.) (.) (.) 4-6 HH members 382.659 170.307* -0.001 (248.44) (91.58) (0.33) 7-9 HH members 487.927* 199.406** 0.108 (257.17) (95.28) (0.32) 10+ HH members 377.234 166.748* 0.067 (269.54) (98.47) (0.33) Location Rural/IDP 0.000 0.000 0.000 (.) (.) (.) Urban -71.677* -38.107* -0.363** (42.25) (21.84) (0.15) Nomadic -30.602 -89.374 -0.016 (136.53) (56.62) (0.44) Enterprise Size Single 0.000 0.000 0.000 (.) (.) (.) 2 to 4 -114.206 -120.407** -0.814*** (96.71) (49.95) (0.11) 5 or more -36.362 -125.794*** -2.954*** (130.54) (32.39) (0.32) Operating Inside Home 0.000 0.000 0.000 Location (.) (.) (.) Outside Home -48.015 -10.047 0.045 (51.83) (28.51) (0.18) Marketplace -122.046 -28.685 0.388** (86.86) (39.54) (0.16) Other -71.095 -33.502 0.166 (69.72) (35.66) (0.18) license_main=0 0.000 0.000 0.000 Licensed (.) (.) (.) license_main=1 25.054 48.922 0.763*** (109.71) (57.91) (0.17) Male 0.000 0.000 0.000 Owner Gender (.) (.) (.) Female -205.594** -89.008** -0.328** (87.68) (43.75) (0.13) None 0.000 0.000 0.000 78 SOMALIA POVERTY AND EQUITY ASSESSMENT Profit Profit Per Worker Revenue Per Worker Owner Education (.) (.) (.) Incomplete Primary -100.925* -57.408* -0.199 (52.19) (29.98) (0.16) Completed Primary -97.562 -35.158 -0.158 (87.11) (32.61) (0.17) Completed Secondary -82.243 -29.154 -0.010 (172.07) (87.02) (0.22) Completed Tertiary 530.120* 155.410 1.086*** Region, Sector, Type, Region, Sector, Type, Region, Sector, Type, Controls Owner Age Owner Age Owner Age 901 901 803 Observations 0.086 0.084 0.420 R-squared 0.043 0.042 0.389 Adjusted R-squared Source: Authors’ estimates based on SIHBS 2022. Potential Impact of Urban Public Works To demonstrate the potential impact of an urban public works program, a rough back of the envelope calculation can be made to assess the potential impact on poverty. It is assumed that the poorest 20 percent of urban households self-select into such a program, and it is limited to one member per household. The number of days worked per week is restricted to 3, resulting in 156 days per year. A daily wage rate of $2 to $5 is calculated and the additional income is added to total household consumption. Poverty is then recalculated based on the per capita consumption with the additional earnings from the public works program. The wage cost of the public works would depend on the daily wage rate, ranging from 0.7 to 1.8 percent of GDP and 8 to 20 percent of government expenditure in 2022.¹³⁶ With the assumption that wage costs account for 80 percent of total costs, this cost increases from 0.9 to 2.3 percent of GDP and 10 to 26 percent of government expenditure in 2022.¹³⁷ While there is no change in the poverty rate, the poverty gap decreases both nationally and in urban areas. Given that IDPs are concentrated in the poorest urban quintile, these households tend to be quite far from the poverty line. As a result, a public works program that targets the poorest urban households does not reduce the poverty headcount, either nationally or in urban areas. However, at the national level the poverty gap is reduced by over a percentage point if the daily wage is $4, while in urban areas a daily wage of $3 reduces the urban poverty gap by 1.5 percentage points. The $5 daily wage decreases the urban poverty gap by 2.5 percentage points. It should be noted, this only considers the direct impact of such a program, and as shown in other countries, there are indirect benefits from the increase in low-skilled wage rates, and other potential benefits may arise from increased domestic spending for household enterprises. ¹³⁶ This only considers the payment of wages associated with such a program. There would also be an administrative cost. GDP in 2022 is assumed to be $10.42bn (https://data.worldbank.org/country/somalia) and government expenditure is $918.7mn (https://www.unicef.org/esa/media/11721/file/Somalia%20National%20Brief.pdf). ¹³⁷ Examples from Ethiopia and Liberia restricted wage costs to 80% of program costs (World Bank 2012). 79 ANNEX: CHAPTER 2 Figure 104: Impact on the Poverty Gap 25.0% 20.0% 15 .0% 10 .0% 5 .0% 0 .0% Before $2 Wage $3 Wage $4 Wage $5 Wage National Urban Source: Authors’ estimates based on SIHBS 2022. 80 SOMALIA POVERTY AND EQUITY ASSESSMENT ANNEX: CHAPTER 3 Table 18: Self-Reported Exposure to Shocks138 Any Shock Food Price Increase Drought Livestock Death All All 75% 53% 47% 14% Poverty Non-Poor 73% 55% 39% 8% Poor 78% 51% 56% 21% Nomadic All 96% 54% 87% 51% Non-Poor 93% 54% 85% 40% Poor 96% 54% 87% 55% Rural All 77% 52% 53% 19% Non-Poor 75% 58% 46% 13% Poor 78% 49% 57% 23% Urban All 70% 54% 37% 5% Non-Poor 70% 55% 32% 4% Poor 71% 52% 43% 7% Source: Authors’ estimates based on the SIHBS 2022. Table 19: Poverty and 3 D’s Regression VARIABLES poor_ub Pop Density -4.14e-05*** (1.08e-05) Share of population within 2km of All-Weather Road -0.0149*** (0.00377) Share of population within 5km of conflict (2018-2022) 0.00831*** (0.00247) Constant 0.148* (0.0863) Observations 7,156 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Probit estimation with SVY settings ¹³⁸ Defined as experiencing a severe negative economic impact. 81 ANNEX: CHAPTER 4 ANNEX: CHAPTER 4 Figure 105: Nomadic Poverty Rates by Region Figure 106: Nomadic Population Share in each Region % % .95 to 1 .35 to .4 .9 to .95 .3 to .35 .85 to .9 .2 to .25 .8 to .85 .15 to .2 .7 to .8 .1 to .15 .6 to .7 .05 to .1 0 to .6 0 to .05 No data No data Source: Authors’ estimates based on SIHBS 2022. LIVESTOCK REVENUE REGRESSION Table 20: Livestock Revenue Regression total_lstck_earn total_lstck_earn Male 0.000 0.000 (.) (.) Female -153.184** 3.099 (60.205) (50.537) None 0.000 0.000 (.) (.) Incomplete Primary -212.428 -218.495* (148.115) (122.751) Completed Primary -25.175 2.783 (241.205) (199.904) Completed Secondary & Post-Sec -55.316 -80.435 (456.736) (378.523) 82 SOMALIA POVERTY AND EQUITY ASSESSMENT total_lstck_earn total_lstck_earn Awdal 0.000 0.000 (.) (.) Bakool -57.029 -514.228** (310.393) (258.310) Bari 288.420* 5.346 (154.559) (128.915) Bay 14.175 -218.786 (307.655) (255.250) Galgaduud 748.402*** 224.422* (155.239) (131.443) Gedo 399.883** 35.272 (163.495) (136.787) Hiraan 238.632 200.719 (160.578) (133.093) Lower Juba -142.018 -564.402** (327.096) (271.949) Lower Shabelle 129.411 162.864 (158.892) (131.694) Waqooyi Galbeed -141.985 -93.415 (155.951) (129.269) Middle Shabelle -152.676 -448.884* (313.747) (260.463) Mudug 354.402** 110.626 (154.511) (128.663) Nugaal 211.322 -221.784* (156.502) (131.597) Sanaag 380.816** 363.862** (174.468) (144.593) Sool 792.617*** 727.114*** (159.490) (132.221) Togdheer 168.717 158.115 (150.195) (124.476) Yes 0.000 0.000 (.) (.) No 189.744* 239.719*** (107.975) (89.521) (max) ndvianom=0 0.000 0.000 (.) (.) (max) ndvianom=1 -51.076 312.072 (260.215) (216.460) tlu_tot 27.496*** (1.413) Observations 850 850 83 ANNEX: CHAPTER 4 total_lstck_earn total_lstck_earn R-squared 0.126 0.400 Adjusted R-squared 0.104 0.385 AIC 13761.220 13442.872 Standard errors in parantheses. Weights are applied in all estimations. ="* p<0.10 ** p<0.05 *** p<0.01" Source: Authors’ estimates based on SIHBS 2022. 84