© 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 Some rights reserved 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. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http:// creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: World Bank (2024). Expanding Development Approaches to Refugees and Their Hosts in Ethiopia. Washington DC. © World Bank. Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by the World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. Photos: © World Bank TABLE OF CONTENTS Abbreviations........................................................................................................................................................................................................................................................................................ i Acknowledgements....................................................................................................................................................................................................................................................................... ii Executive Summary......................................................................................................................................................................................................................................................................... vi 1. Introduction................................................................................................................................................................................................................................................................................ 1 1.1 How can we achieve better development outcomes for all?.................................................................................................................................................... 5 1.2 How does Socio-Economic Survey of Refugees in Ethiopia (SESRE) contribute to the debate on policies?........................................ 6 2. Sociodemographic Profile ........................................................................................................................................................................................................................................... 9 2.1 Demographic characteristics .............................................................................................................................................................................................................................. 9 2.2 Education ........................................................................................................................................................................................................................................................................... 12 2.3 Health and nutrition................................................................................................................................................................................................................................................... 17 2.4 Living conditions........................................................................................................................................................................................................................................................... 21 3. Jobs and Livelihoods ........................................................................................................................................................................................................................................................ 24 3.1 Labor market outcomes of in-camp refugees and their hosts................................................................................................................................................. 27 3.2 Labor market outcomes of OCP refugees and their hosts........................................................................................................................................................... 34 3.3 Refugee youth................................................................................................................................................................................................................................................................. 37 4. Refugees’ Aspirations........................................................................................................................................................................................................................................................ 39 5. Welfare and Equity.............................................................................................................................................................................................................................................................. 42 5.1 Welfare dimensions..................................................................................................................................................................................................................................................... 42 5.1.1 Monetary poverty and inequality .............................................................................................................................................................................................. 42 5.1.2 Expenditure patterns ......................................................................................................................................................................................................................... 44 5.1.3 Multidimensional poverty .............................................................................................................................................................................................................. 46 5.1.4 Food security........................................................................................................................................................................................................................................... 47 5.1.5 Shocks and coping strategies........................................................................................................................................................................................................ 48 5.2 Determinants of welfare.......................................................................................................................................................................................................................................... 49 5.3 Cost of basic needs for refugees ...................................................................................................................................................................................................................... 53 6. Markets and Opportunities ....................................................................................................................................................................................................................................... 56 6.1 Spatial disparities in refugees labor market access and outcomes....................................................................................................................................... 57 6.2 Effects of local factors on refugees’ labor market outcomes...................................................................................................................................................... 61 7. Social Cohesion....................................................................................................................................................................................................................................................................... 63 7.1 Attitudes between refugees and hosts....................................................................................................................................................................................................... 64 7.2 Social interactions ....................................................................................................................................................................................................................................................... 68 8. Policy Recommendations ............................................................................................................................................................................................................................................ 72 References .............................................................................................................................................................................................................................. 77 Annexes .................................................................................................................................................................................................................................. 82 Annex A: Description of Refugees by Country of Origin............................................................................................................................................... 83 Annex B: Refugee Policies in Ethiopia.............................................................................................................................................................................. 85 Annex C: Survey Design and Methodology..................................................................................................................................................................... 92 Annex D: Descriptive Statistics and Regression Results.............................................................................................................................................. 97 Annex E: Robustness Checks of Refugees’ Consumption.......................................................................................................................................... 124 Annex F: Comparison of Results from Skills Profile Survey and SESRE.................................................................................................................. 128 LIST OF FIGURES Figure ES.1: Desired location in three years ����������������������������������������������������� iv Figure 3.7: Female work type........................................................................... 30 Figure ES.2: Expected location in three years ��������������������������������������������������� iv Figure 3.8: Female occupations........................................................................ 30 Figure ES.3: Poverty incidence............................................................................ v Figure 3.9: Refugee work location.................................................................... 30 Figure ES.4: Food insecurity scale...................................................................... v Figure 3.10: Hours per week.............................................................................. 31 Figure ES.5: Refugee employment and proximity to resource hubs .................. vi Figure 3.11: Hourly earnings.............................................................................. 31 Figure ES.6: Host response to “Refugees are good people” ............................ vii Figure 3.12: Share employed by age – camp refugees ������������������������������������ 31 Figure ES.7: Host response to “Would you feel comfortable having a Figure 3.13: Share employed by age – camp hosts ����������������������������������������� 31 refugee as a neighbor?” ��������������������������������������������������������������� vii Figure 3.14: Household owns crops................................................................... 32 Figure 1.1: Refugees and asylum seekers in Ethiopia by country of origin, 1984-2023.................................................................................... 1 Figure 3.15: Household owns livestock ������������������������������������������������������������� 32 Figure 2.1: Refugees by survey domain ���������������������������������������������������������� 10 Figure 3.16: Total value of livestock................................................................... 32 Figure 2.2: Country of birth.............................................................................. 10 Figure 3.17: Value per tropical livestock unit ����������������������������������������������������� 32 Figure 2.3: Refugees arrival in Ethiopia (15 years and above) ..................... 10 Figure 3.18: Household has non-farm business ������������������������������������������������ 33 Figure 2.4: Age structure.................................................................................. 11 Figure 3.19: Value of productive assets among households with non-farm business......................................................................... 33 Figure 2.5: Gender composition........................................................................ 11 Figure 3.20: Household primary income source ������������������������������������������������ 33 Figure 2.6: Marital status (18 years and above) ��������������������������������������������� 11 Figure 3.21: Household primary income source ������������������������������������������������ 35 Figure 2.7: Education level (18 years and above) ������������������������������������������ 13 Figure 3.22: Work status by gender ................................................................... 36 Figure 2.8: Youth (15 to 24) education level �������������������������������������������������� 13 Figure 3.23: Work type by gender....................................................................... 36 Figure 2.9: Refugees’ education outside of Ethiopia (18 years and above) .... 15 Figure 3.24: Occupation...................................................................................... 36 Figure 2.10: Children currently attending school ����������������������������������������������� 15 Figure 3.25: Occupation among completed secondary or more ....................... 36 Figure 2.11: Gross Enrollment Rate (GER) �������������������������������������������������������� 16 Figure 3.26: Refugee occupation concentration ������������������������������������������������ 37 Figure 2.12: Net Enrollment Rate (NER) ������������������������������������������������������������ 16 Figure 3.27: Youth work status........................................................................... 38 Figure 2.13: Share of children and youth above school age in education ....... 16 Figure 3.28: Male youth work status.................................................................. 38 Figure 2.14: Reasons for not currently attending school ���������������������������������� 17 Figure 3.29: Female youth work status �������������������������������������������������������������� 38 Figure 2.15: Faced any health problem ������������������������������������������������������������� 18 Figure 4.1: Desired location in three years ����������������������������������������������������� 39 Figure 2.16: Received medical assistance when faced with health problem .... 18 Figure 4.2: Expected location in three years ��������������������������������������������������� 39 Figure 2.17: Use of the national healthcare system when faced with health problems.............................................................................. 19 Figure 4.3: Locus of control............................................................................. 41 Figure 2.18: Child nutritional indicators ������������������������������������������������������������� 21 Figure 4.4: Locus of control by type of control ����������������������������������������������� 41 Figure 2.19: Presence of any disability ������������������������������������������������������������� 21 Figure 5.1: Poverty incidence........................................................................... 43 Figure 2.20: Dwelling type.................................................................................. 22 Figure 5.2: Income inequality, Gini index ���������������������������������������������������������� 43 Figure 2.21: Housing quality............................................................................... 22 Figure 5.3: Expenditure components (in birr) ������������������������������������������������� 44 Figure 2.22: Access to drinking water and hygiene .......................................... 23 Figure 5.4: Shares of food expenditure ����������������������������������������������������������� 44 Figure 2.23: Access to toilet facility and waste disposal ������������������������������������ 23 Figure 5.5: Food and non-food expenditures shares by sources ................... 45 Figure 2.24: Source of lighting........................................................................... 23 Figure 5.6: Food expenditure shares by food groups ������������������������������������ 46 Figure 3.1: Top 3 difficulties with being a refugee ������������������������������������������ 25 Figure 5.7: Multidimensional poverty incidence, severity, and vulnerability ... 46 Figure 3.2: Work status..................................................................................... 28 Figure 5.8: Perceived changes in household living standards .......................... 47 Figure 3.3: Work type........................................................................................ 28 Figure 5.9: Food insecurity scale for refugees and hosts 48 Figure 3.4: Occupation...................................................................................... 28 Figure 5.10: Dietary diversity and food consumption status ............................ 48 Figure 3.5: Work status by gender.................................................................... 29 Figure 5.11: Type of shocks experienced ���������������������������������������������������������� 49 Figure 3.6: In-camp refugee share employed by age ������������������������������������� 29 Figure 5.12: Shock coping strategies ���������������������������������������������������������������� 49 Figure 5.13: Household composition by quintiles ����������������������������������������� 50 Figure 7.16: Share of refugees who agree they are “culturally similar to hosts”....................................................................... 70 Figure 5.14: Demographic characteristics by quintile ���������������������������������� 50 Figure 7.17: Share or refugees involved in a community representative body 70 Figure 5.15: Poverty headcount rate for in-camp refugees and their hosts, by domain ............................................................... 50 Figure 7.18: Share or refugees engaged in a community representative body by demographic group ���������������������������������������������������� 70 Figure 5.16: Poverty incidence decreases with education of the household head ......................................................................... 51 Figure 7.19: Discrimination and harassment ������������������������������������������������ 71 Figure 5.17: Household wealth indicators by expenditure quintiles ............. 52 Figure 7.20: Discrimination and harassment by demographic group ....... 71 Figure 5.18: Labor market outcomes by expenditure quintiles .................. 52 Figure D.1: Age group by gender................................................................. 100 Figure 5.19: Poverty rates and employment for refugees and hosts ........... 53 Figure D.2: Refugees’ education document ����������������������������������������������� 100 Figure 5.20: Share of consumption provided in-kind or for free Figure D.3: Share of school-age children in education per household ..... 100 by consumption per capita quintiles among in-camp refugees 54 Figure D.4: School-age children currently attending school by gender .... 101 Figure 5.21: Poverty incidence at consumption and pre-assistance consumption levels .................................................................... 54 Figure D.5: Reasons for not currently attending school by gender ........... 101 Figure 5.22: Costs of basic needs per refugee per year under Figure D.6: Average annual household education expenditure (in ETB) ... 101 different scenarios .................................................................... 54 Figure D.7: Type of health institutions......................................................... 102 Figure 6.1: Refugee incidence....................................................................... 58 Figure D.8: Problems faced in health institutions ��������������������������������������� 102 Figure 6.2: Labor force participation rate by proximity to resource hub, Figure D.9: Stunting by gender of children ������������������������������������������������� 102 market accessibility.................................................................... 59 Figure D.10: Childbirth in health institutions (children under five years) 103 Figure 6.3: Refugees’ labor market outcomes �������������������������������������������� 59 Figure D.11: No birth evidence available (children under five years) ......... 103 Figure 6.4: Sectoral employment.................................................................. 60 Figure D.12: Average annual per capita health expenditure ....................... 103 Figure 6.5: The share of refugee youth who are NEET ������������������������������� 60 Figure D.13: Types of disability...................................................................... 104 Figure 6.6: Local labor supply effect of refugee’s odds of employment 61 Figure D.14: Rent expenditure (Refugees and hosts in Addis Ababa) ........ 104 Figure 6.7: Local unemployment level matters to obtain jobs ................... 61 Figure D.15: Hand washing facility................................................................. 104 Figure 6.8: Distance to the nearest city and the chance of obtaining a job for refugees ������������������������������������������������������ 62 Figure D.16: Top 3 difficulties with being a refugee by survey domains ..... 108 Figure 6.9: Employment and proximity to resource hubs .......................... 62 Figure D.17: Work status by survey domains ������������������������������������������������ 108 Figure 7.1: Host response to “Refugees are good people” ....................... 64 Figure D.18: Type of work by survey domains ���������������������������������������������� 108 Figure 7.2: Host response to “Would you feel comfortable having a refugee Figure D.19: Occupation by survey domains ������������������������������������������������� 108 as a neighbor?”.......................................................................... 64 Figure D.20: Work location by survey domains ��������������������������������������������� 108 Figure 7.3: Host response to “Refugees are good people” by gender ..... 65 Figure D.21: Hours per week by survey domains ����������������������������������������� 108 Figure 7.4: Host response to “Would you feel comfortable having a refugee as a neighbor?” by gender ��������������������������������������� 65 Figure D.22: Hourly earnings by survey domains ����������������������������������������� 109 Figure 7.5: Share of hosts who agree refugees should have access to... 66 Figure D.23: Household owns crops ............................................................. 109 Figure 7.6: Host beliefs about refugee impact in Ethiopia ......................... 66 Figure D.24: Household owns livestock.......................................................... 109 Figure 7.7: Negative experiences due to refugees �������������������������������������� 66 Figure D.25: Total value of livestock ............................................................. 109 Figure 7.8: Positive experience due to refugees ����������������������������������������� 66 Figure D.26: Value per tropical livestock unit ������������������������������������������������ 109 Figure 7.9: Are most Ethiopians/refugees in Ethiopia trustworthy? .......... 67 Figure D.27: Household has non-farm business ����������������������������������������� 109 Figure 7.10: Host attitudes index.................................................................... 67 Figure D.28: Value of productive assets in households with business ........ 110 Figure 7.11: Host attitudes index by gender ������������������������������������������������� 67 Figure D.29: Primary source of income pre-post migration by survey domains......................................................................... 110 Figure 7.12: Share with family or friends in Ethiopia ������������������������������������� 68 Figure D.30: Youth work status by survey domains �������������������������������������� 110 Figure 7.13: Share with friends in Ethiopia by demographic group ............. 68 Figure D.31: In-camp refugee locations by ecological Zone ........................ 115 Figure 7.14: Share of refugees who think interactions with hosts are “easy to do” ............................................................................ 69 Figure D.32: Refugee’s labor market performance ��������������������������������������� 115 Figure 7.15: Who do refugees rely on in times of need �������������������������������� 69 Figure D.33: Economic sector........................................................................ 115 LIST OF TABLES Table 2.1: Household characteristics............................................................................................................................................................................ 11 Table 3.1: Labor force statistics...................................................................................................................................................................................... 28 Table 3.2: Labor force statistics...................................................................................................................................................................................... 35 Table 3.3: Youth labor force statistics (age 15-24)........................................................................................................................................................ 38 Table B.1: Pledges made at 2016 UN Leaders’ Summit and progress........................................................................................................................ 88 Table B.2: GRF pledges and implementation progress................................................................................................................................................ 90 Table C.1: The distribution of sampled and surveyed households by domains....................................................................................................... 94 Table D.1: Demographic characteristics by survey domains...................................................................................................................................... 97 Table D.2: Education outcomes by survey domains.................................................................................................................................................... 98 Table D.3: Health outcomes by survey domains.......................................................................................................................................................... 99 Table D.4: Living conditions by survey domains........................................................................................................................................................... 99 Table D.5: Labor force statistics by survey domains.................................................................................................................................................... 105 Table D.6: Determinants of refugee-host earnings gap............................................................................................................................................... 105 Table D.7: Determinants of employment outcomes.................................................................................................................................................... 106 Table D.8: Refugee Household Reliance on NGOs/Donations .................................................................................................................................... 107 Table D.9: Determinants of refugee-host earnings gap............................................................................................................................................... 107 Table D.10: Refugee intention to migrate abroad........................................................................................................................................................... 111 Table D.11: Poverty headcount rate by subgroups......................................................................................................................................................... 112 Table D.12: Determinants of welfare (total expenditure per capita)............................................................................................................................ 113 Table D.13: Determinants of welfare for in-camp refugees........................................................................................................................................... 114 Table D.14: Variables used to estimate employment outcomes.................................................................................................................................. 116 Table D.15: Factors determining the odds of obtaining a job for refugees: logit model............................................................................................. 117 Table D.16: Proximity and market accessibility effects on engagement in agriculture activity: logit model........................................................... 119 Table D.17: Proximity and market accessibility effects on engagement in service sector: logit model................................................................... 120 Table D.18: Regression analysis of host and refugee attitudes..................................................................................................................................... 121 Table D.19: Regression analysis of social integration outcomes.................................................................................................................................. 122 Table D.20: Regression analysis of social integration and labor market outcomes.................................................................................................... 123 Table E.1: Food aid data/information received from UNHCR..................................................................................................................................... 124 Table E.2: Food aid and consumption comparisons................................................................................................................................................... 125 Table E.3: Aggregate food expenditures........................................................................................................................................................................ 126 Table E.4: Food aid data/information received from WFP.......................................................................................................................................... 126 Table E.5: Food quantity and expenditure comparisons............................................................................................................................................ 127 Table E.6: Aggregate food expenditures........................................................................................................................................................................ 128 Table F.1: Results on common indicators from SPS 2017 and SESRE 2023.............................................................................................................. 129 LIST OF BOXES Box 1.1: Comparison of SPS 2017 and SESRE 2023.................................................................................................................................................. 7 Box 2.1: Education system for refugees in Ethiopia.................................................................................................................................................. 13 Box 2.2: Refugees under the Out-of-Camp Policy (OCP).......................................................................................................................................... 14 Box 3.1: Eritrean refugee sample in the SESRE.......................................................................................................................................................... 26 Box 3.2: OCP Refugees under the Amnesty Program................................................................................................................................................ 27 Box 3.3: Refugee Vocational Training and Cooperatives.......................................................................................................................................... 34 Box 5.1: Consumption aggregation and poverty measurement............................................................................................................................. 43 Box 5.2: Disparity between refugee ration aid and reported consumption quantities......................................................................................... 45 Box 5.3: MPI methodology........................................................................................................................................................................................... 47 Box 5.4: Estimation of the cost of basic needs for refugees..................................................................................................................................... 55 Box 6.1: Measurement of proximity and market access index in Ethiopia............................................................................................................. 58 Box 7.1: Socio-political tensions in the Gambella Region........................................................................................................................................ 65 Box B.1: Employment pathways of refugees............................................................................................................................................................. 87 ABBREVIATIONS ARRA Administration for Refugee and LFPR Labor Force Participation Rate Returnee Affairs LFS Labor Force Survey CBHI Community Based Health Insurance LoC Locus of Control CPI Consumer Price Index LSMS Living Standards Measurement Study CRRF Comprehensive Refugee Response MoE Ministry of Education Framework MoLS Ministry of Labor and Skills CSB Corn Soy Blend MoLSA Ministry of Labor and Social Affairs DICAC Development and Inter-Church Aid Commission MoR Ministry of Revenue EA Enumeration Area MoTRI Ministry of Trade and Regional Integration EMIS Education Management Information System MoU Memorandum of Understanding EOP Economic Opportunities Program MPI Multidimensional Poverty Index ESDP Education Sector Development NEET Not in Employment, Education or Programme Training ESS Ethiopian Statistical Service NER Net Enrollment Rate EUAA European Union Agency for Asylum NGO Non-governmental Organization FDRE Federal Democratic Republic of Ethiopia OAU Organization of African Unity GCR Global Compact on Refugees OCP Out-of-Camp Policy GDP Gross Domestic Product PPP Purchasing Power Parity GER Gross Enrollment Rate RRS Refugees and Returnees Service GIZ German Agency for International SESRE Socio-economic Study of Cooperation Refugees in Ethiopia GoE Government of Ethiopia TVET Technical and Vocational Education and Training GRF Global Refugee Forum UNHCR The United Nations Refugee Agency HoWStat Household Welfare Statistics Survey UNICEF The United Nations International IOM International Organization for Migration Children’s Emergency Fund IPCC Intergovernmental Panel on Climate UPSNJP Urban Safety Net and Jobs Project Change WFP World Food Programme ISCO International Standard Classification of Occupations WHO World Health Organization JDC Joint Data Center ACKNOWLEDGEMENTS The report titled "Expanding development approaches to refugees and their hosts in Ethiopia" was prepared by a team of the Poverty and Equity Global Practice at the World Bank led by Christina Wieser (Senior Economist, World Bank), including Wondimagegn Mesfin Tesfaye (Economist, World Bank), Fikirte Girmachew (Consultant, World Bank), Jeremey Aaron Lebow (Young Professional, World Bank), and Manex Bule Yonis (Economist, World Bank), under the adept guidance of Pierella Paci (Practice Manager, World Bank). The report is an output of a collaboration effort between the World Bank, the Ethiopian Statistical Service (ESS), the Ethiopia Refugees and Returnees Service (RRS), and UNHCR with generous financial support from the World Bank and UNHCR Joint Data Center on Forced Displacement (JDC). Our appreciation extends to the UNHCR and RRS colleagues for their unwavering support. We offer our special thanks to Mulualem Desta (Deputy Director General, RRS) and Ashenafi Demeke (Education Team leader, RRS) for their critical guidance and support throughout the survey design, implementation, and report development phases. Moreover, we would like to thank colleagues from RRS who provided invaluable comments on the various draft stages of this report: Bruhtesfa Mulugeta, Zewdu Bedada, Daniel Adefires, Anteneh Mekasha, Anteneh Gorfu, Yewulsew Nigussie, Fantaw Kabtamu, Biruk Kebede, Dr. Goitom Ademnur, Dr. Tagay Kelil, Daniel Darcha. Additionally, we acknowledge the Household Expenditure and Welfare Statistics core team at ESS for their exceptional work on the data collection, from the survey's inception to the report's culmination: Amare Legesse, Alemayehu Teferi, Efrem Belachew, Salah Yusuf, Seid Jemal, Hagos Haile, Zenaselase Siyum, Tsigab Halefom, Yirga Nigussie, Kassu Gebeyehu, Zemecha Abdella, Mengistu Abebe, Aklilu Fikre, and Sisay Guta. Our gratitude goes to Leslie Velez (Assistant Representative (Protection, UNHCR) for her unwavering support and encouragement throughout the whole SESRE process. We are particularly thankful to UNHCR colleagues Yonas Lemma, Yonatan Assefa, Michel Uwamahoro, Millicent Lusigi, and Mekdes Aschalew for enabling access to administrative refugee data and their support during the sampling and data collection stages. Moreover, our thanks go to the following UNHCR colleagues for their invaluable comments on the various draft stages of this report: Emily Lugano, Annick-Laure Tchuendem, Jed Fix, Theresa Beltramo, Alessio Baldaccini, Anna Gaunt, Asaad Kadhum, Benoit d’Ansembourg, Berhanu Geneti, Campbell Macknight, Daniel Gebrekidan, Florah Bukania, Florence Nimoh, Johannes Abate, Joyce Wahome, Katie Ogwang, Nada Omeira, Nathalie Bussien, Robert Nyambaka and Yukta Kumar. Our thanks also go to our JDC colleagues Felix Schmieding (Senior Statistician, UNHCR) and Harriet Kasidi Mugera (Senior Data Scientist, World Bank) for their tireless support and advice during the preparation of SESRE. The report benefited from the insights of peer reviewers: Leslie Velez, Mulualem Desta, Takaaki Masaki (Senior Economist, World Bank), and Precious Zikhali (Senior Economist, World Bank). We are indebted to Nistha Sinha (Senior Economist, World Bank) for her constructive comments that significantly enhanced the report. We also recognize Aldo Morri for his excellent editorial support. Finally, we extend our deepest gratitude to the survey respondents for their willingness to share their experiences which has been instrumental in deepening our understanding of the challenges and needs faced by both refugees and host households. i EXECUTIVE SUMMARY Introduction The Socio-Economic Survey of Refugees in Ethiopia (SESRE) plays a crucial role in informing E thiopia, with its long history of hosting refugees, is grappling with the complex challenges of accommodating close to 1 million refugees policy decisions by providing comprehensive data on the socioeconomic dimensions of refugees and host communities. By highlighting and asylum seekers. These come primarily from socioeconomic interactions and outcomes, SESRE neighboring countries like South Sudan, Somalia, aims to guide development interventions and Eritrea, and Sudan housed in camps in mostly rural facilitate refugee integration. The survey covers areas spread around the country near border areas. various aspects, including demographic profiles, livelihoods, welfare patterns, and social cohesion, While Ethiopia has adopted progressive refugee offering valuable insights for policymakers and policies, including the Comprehensive Refugee humanitarian actors. Response Framework (CRRF), challenges persist in translating these policies into tangible SESRE is a separate but integrated survey socioeconomic outcomes for refugees. Despite alongside the Ethiopian Household Welfare Ethiopia’s efforts to shift from a camp-based Statistics Survey (HoWStat),1 the national approach to a more inclusive model promoting household survey to measure poverty and other self-reliance and integration, refugees live largely socio-economic outcomes. Like most national in camps, are reliant on humanitarian aid, and face poverty surveys, HoWStat excludes displaced barriers to accessing employment and education. populations—Internally Displaced People (IDPs) or The country’s new Refugee Proclamation grants refugees—including in Ethiopia. To have up-to-date refugees the right to basic services, work, and information on the socio-economic outcomes and freedom of movement, but implementation delays poverty levels of refugees and to allow comparison hinder their realization. to Ethiopian host communities, the SESRE applied the same questionnaire and data collection methods To address these challenges and achieve better as the HoWStat, with some modifications. The World development outcomes for both refugees and host Bank, Ethiopia’s RRS, Ethiopia’s Statistical Service, communities, a shift towards supporting refugees’ and UNHCR collaborated to implement SESRE and self-reliance and economic integration is essential. was the first of its kind. This involves enabling refugees to move toward economic opportunities, facilitating their access to This report uses data from the SESRE extensively the labor market through self-employment, wage- to analyze the Ethiopian refugee situation and employment, and special projects, and integrating to devise policy directions. The SESRE covers refugee children into the education system. Though three types of groups: (i) refugees in camps; (ii) refugees in Ethiopia still face significant barriers to refugees out-of-camps in Addis Ababa; and (iii) host accessing employment and education, hampering communities; all of which require a distinct sampling their long-term integration and exacerbating their procedure. The sampling frame for refugee camps vulnerability, initiatives are on the way to improve is based on UNHCR’s proGRES database. SESRE is socioeconomic outcomes. a representative survey of the refugee population 1 Formerly the Household Consumption and Expenditure Survey and Welfare Monitoring Survey. ii Executive Summary of Eritrean, South Sudanese, and Somali origin and primary school enrollment rates are similar living in camps in Ethiopia, refugees living in Addis between refugees and hosts, but secondary school Ababa, and their respective host communities. Host enrollment rates are much lower among refugees. communities are defined as Ethiopian non-displaced Inadequate school infrastructure, the need to households living enumeration areas adjacent to support family income, and families unwilling to send the refugee camps. SESRE data was collected from children to school are some main reasons for the low November 2022 to January 2023, from a nationally secondary school enrollment. Refugee children are representative sample of 3,452. The following also much more likely to not attend education at the represents a summary of findings stemming from appropriate age. Providing sufficient, appropriate, the SESRE data and associated statistical regression and sustainable support from all responsible actors work using this data. can overcome some of these challenges. Sociodemographic Health: The prevalence of illness and getting medical assistance are similar between refugees and Ethiopia is a second home for close to one million hosts, with child nutritional problems of stunting, refugees who predominantly originate from South underweight, and wasting challenging for both Sudan, Somalia, and Eritrea. Around 88 percent of refugee and host children. refugees live in camps, and the rest reside in urban areas under the Out-of-Camp Policy (OCP) regime. Basic infrastructure: Refugees and hosts have similar Refugees fled from their country mainly due to access to WASH facilities and access to electricity. conflict and violence. After they arrive in Ethiopia, However, housing conditions are worse for in-camp refugees stay, on average, 15 years. refugees, who mainly live in shelters, whereas OCP refugees live in rented housing of better quality. Refugees and hosts share similar demographic characteristics regarding age and gender. However, Jobs and Livelihoods in-camp refugees have a higher share of children and youth, with a significantly higher number of In-camp refugees mainly rely on humanitarian second-generation born in Ethiopia compared to aid as they have low employment rates and few OCP refugees, the majority being within a working opportunities to generate income. Labor market age group. outcomes show high inactivity and unemployment rates for in-camp refugees. If refugees earn income, The refugee policy granted refugees the right to they are less likely than hosts to earn from agriculture, access basic services, including primary education livestock, and non-farm business. Given low and healthcare services in camps and secondary education, employed refugees tend to work in low- education and health services under the national skill jobs, though there is a disparity in the occupation system. types among refugees by country of origin: Eritrean refugees work in crafts and related trades, while Education: Educational attainment is low among South Sudanese refugees are engaged in elementary refugees and hosts, but the majority of refugees have occupations, and Somali refugees work in a mix of no education or attend below primary education. services, sales, and skilled agriculture. Besides low This is worse for in-camp refugees. OCP refugees employment, refugees’ ownership of assets such as (Eritreans in Addis Ababa) have better education agricultural land, livestock, and productive assets is before they arrive in Ethiopia. School attendance lower than that of hosts. iii Executive Summary Working outside of camps helps improves refugees’ household incomes. On average, in-camp refugee livelihoods. A significant proportion of in-camp women and men are equally likely to be employed refugees work outside camps despite not having work (around 25 percent), while among hosts, men are permits, earning more than those working inside twice as likely to be employed (62 percent compared camps. For employed in-camp refugees, hourly and to 37 percent for women). Like men, refugee monthly earnings are lower than for hosts. However, women are more likely than host counterparts to having lower wages is not associated with education be self-employed and less likely to be in high-skill level or experience. Having higher educational occupations. attainment and experience increases the likelihood of employment and income levels for hosts, not for Refugee Aspirations refugees. Refugees receive returns from education Despite low resettlement rates, most refugees and experience only when working outside camps. unrealistically aspire to go to a Western country in Likewise, reliance on assistance for in-camp refugees the next three years (Figure ES.1). Even when asked declines when a household member works outside where they would realistically be in the next three the camp. years, one-third of refugees believe that they will live in a Western country (Figure ES.2). The intention Similar to in-camp refugees, OCP refugees heavily to migrate abroad is higher for youth. Refugees also rely on remittances. Selection criteria for OCP perceive they have less control over their lives than refugees allow refugees to get OCP permits based on hosts, a result driven by South Sudanese refugees. self-reliance or support from others. Hence, Eritrean These intentions to migrate combined with low “locus refugees in Addis Ababa also had better educational of control” (LOC) may limit refugees’ investment in attainment, indicating that they were relatively improving their livelihoods or to integrate. well-off before displacement and continued having support from their families after displacement. Welfare and Equity Few OCP refugees work but if they work, they In-camp refugees are poorer than their hosts. While face occupational downgrading regardless of monetary poverty appears to be high in refugee- demographic characteristics, and all refugees are concentrated areas, it is more prevalent among less likely to be employed in high-skill jobs than in-camp refugees than their hosts or OCP refugees hosts despite completing secondary education. (Figure ES.3). Welfare varies significantly over the different groups of refugees in Ethiopia, with Eritrean Female refugees have high employment rates—as refugees having the lowest poverty incidence and high as men’s—and their high work participation South Sudanese refugees the highest. Although rate makes a critical contribution to refugee poverty incidence is higher for refugees, the high Figure ES.1: Desired location in three years Figure ES.2: Expected location in three years 100 100 80 80 60 Percent 60 Percent 40 40 20 20 0 0 Eritrean Somali South OCP All Eritrean Somali South OCP All (camps) Sudanese Refugees (camps) Sudanese Refugees Ethiopian refugee camp Ethiopian city Country of birth Ethiopian refugee camp Ethiopian city Country of birth Other African country Western country Other African country Western country Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. iv Executive Summary Figure ES.3: Poverty incidence Figure ES.4: Food insecurity scale 8.1 84% 7.0 Poverty headcount rate (%) 75% 3.4 4.0 32% 2.9 25% 18% 2.1 7% In Camp Addis Ababa Total In camp Addis Ababa Total Hosts Refugees Hosts Refugees Source: World Bank staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. poverty rates among hosts imply that refugee host self-sufficiency and food security of in-camp refugees communities are themselves severely resource- and host communities by improving their livelihood constrained; this calls for the urgent need of place- opportunities. based developmental investment in the area to benefit both refugees and host communities. For in-camp refugees, consumption (expenditures) tends to increase with certain characteristics. Besides losses refugees have endured, welfare and These include higher education, access to mobile economic disparities between refugees and host phones, owing a non-farm business, possessing a communities in Ethiopia are due to limited access bank account, and being closer to a market town or refugees have to livelihood opportunities and legal and Woreda capitals. Education (of the household restrictions on their employment. Legal restrictions head) and employment strongly correlate with (i.e., not having work permits) and location often higher household expenditures, providing evidence prevent refugees from working, which limits their again that improved access to education and labor ability to generate income and improve their markets would reduce poverty among refugees. economic situation. As a result, many refugees rely on food aid and have limited access to necessities Policies that limit formal employment and mobility such as housing and electricity. of in-camp refugees contribute to their economic exclusion. We analyzed economic aid needed for Multidimensional poverty tends to be high among each refugee under three scenarios: (i) no economic refugees. Low living standards and low education opportunities, (ii) current level of integration, and primarily drive multidimensional poverty. Standard (iii) full integration. Under the hypothetical scenario of living indicators, low-quality cooking fuel, of “no economic opportunities”—where refugees inadequate housing and low asset ownership, are not allowed to work but solely depend on contribute half to non-monetary poverty. Moreover, aid or assistance—the annual cost of basic needs deprivation in education and child malnutrition per refugee would be approximately US$378. The also contribute most to multidimensional poverty “current” level of economic integration scenario, among refugees. where refugees can find opportunities to earn money or work—assuming that assistance is the Refugee households tend to have worse food gap between the consumption of refugees to the security than hosts. In-camp refugees have less poverty line— reduces the cost by 44 percent to an diverse diets, suffer food insecurity, and have low annual US$221 per person. Further, the cost of basic consumption status compared to hosts (Figure ES.4). needs could decrease to an estimated US$78 per Broadly, there is a need to enhance the economic year if the country adopts a “full inclusion” scenario v Executive Summary where in-camp refugees have equal opportunities The local labor market structure affects the as hosts. The results show that refugee integration possibility of refugees finding jobs. Naturally, the has considerable potential to save money, creating better the local labor market, the easier for refugees an “economic-inclusion dividend” that could be to find employment. High local unemployment allocated to other interventions. reduces refugees’ job prospects, regardless of the gender of the refugee. The structure of sectoral Markets and Opportunities employment in the local market also affects the Ethiopia’s 24 refugee camps2 are spatially odds of refugee employment; the higher the share dispersed, and location matters significantly in of employment in the trade and services, the better terms of refugee’s ability to work. About 88 percent the likelihood of employment for refugees. Refugees of refugees in Ethiopia remain in camps (based on are more likely to work where most land is used for SESRE data). The different camp areas have different non-agriculture (built-up and shops). Overall, results geographic, social, and economic contexts, and are indicate the importance of agglomeration effects, as in different ecological zones, with different ethnic refugees perform better in labor markets with urban and language linkages between the refugees and characteristics. local host communities. Refugees overall have lower employment rates and incomes and are more likely Proximity to resource hubs and connectivity help to engage in the informal sector than their hosts, but refugees to work, regardless of gender. Refugees spatial disparity in labor market access and outcomes in well-connected areas have better prospects of among refugees exists. The local labor market being employed. The gender gap persists at any structure, proximity to resource hubs (Zone capitals, level of market access but is more pronounced with Woreda cities), and market connectivity significantly decreased accessibility. Refugees are also more likely explain the differences in refugee labor market to work in agriculture in areas with poor market outcomes, highlighting the importance of refugees’ access, while more connectivity encourages service locations in terms of providing opportunities for self- sector work. reliance (Figure ES.5). Social Cohesion Figure ES.5: Refugee employment and proximity to resource hubs Hosts display a generally positive attitude towards 0.7 refugees (Figure ES.6 and Figure ES.7). Cultural and 0.6 Probability of being employed 0.5 linguistic proximity and perception of improvement 0.4 of local infrastructure are related to hosts’ positive 0.3 attitude and trust towards refugees. Positive 0.2 attitudes are stronger among Somali refugees and 0.1 hosts and weaker between South Sudanese refugees 0.0 and hosts. Even though both Somali and South Nearest to Nearest to Nearest to Remote zone Woreda border Sudanese refugees are culturally similar to their Male Female hosts, the socio-political tension in the Gambella Source: World Bank Staff based on SESRE 2023. Note: Predicted marginal probabilities of being employed based on proximity region weakens host attitudes toward South to resource, tabulated by gender. Sudanese refugees of Nuer ethnicity. 2 Although there are approximately 30 refugee camps in Ethiopia, this report refers to the 24 camps included in SESRE. vi Executive Summary Hosts tend to support refugees’ right to work Refugee Forum to improve the socio and economic and move to locations with better economic opportunities for refugees through an agenda to opportunities. Hosts support increasing refugees’ transform camps to human settlements as well as economic opportunities in Ethiopia, but some for inclusion into national services for education, perceive that refugees increase insecurity and are including secondary education as well as health taking their land. Hosts’ perceptions of adverse (UNHCR, 2024). The recommendations below effects from refugees are low, but they indicate the advance these commitments backed by the findings impact on economic competition, price increases, in this survey. deforestation, and security issues. Key policy recommendations stemming from this Trust between refugees and hosts is similar, with analysis are: refugees being more trusting. Refugees are more likely to trust a host if they are culturally similar. Promote refugee self-reliance: Cultural proximity and positive perceptions of ◆ Enable mobility for refugees to access areas with economic benefits improve the co-existence of higher economic opportunities. refugees and hosts. Still, additional effort is required ◆ Facilitate labor market access for refugees by to improve the social integration of refugees for easing restrictions and providing work permits. enhanced economic integration. ◆ Integrate refugee children into national Policy Recommendations education system to improve their long-term prospects. Addressing challenges refugees face in Ethiopia ◆ Strengthen inclusive healthcare systems to requires a concerted effort to promote their self- address the health needs of refugees. reliance, economic integration, and access to education and health. By leveraging data from Focus on place-based interventions: initiatives like SESRE and adopting a comprehensive ◆ Invest in refugee hosting areas to benefit both approach that considers the needs of both refugees refugees and host communities. and host communities, Ethiopia can maximize the benefits from hosting refugees while minimizing ◆ Direct additional educational resources to associated costs. The Government of Ethiopia districts hosting refugees to support integration. has committed to a significant shift in its refugee ◆ Expand access to social safety nets for vulnerable management policies and most recently in its refugees and hosts. pledges and commitments made at the 2023 Global Figure ES.6: Host response to “Refugees are good people” Figure ES.7: Host response to “Would you feel comfortable having a refugee as a neighbor?” 100 100 90 90 80 80 70 70 Percent 60 60 Percent 50 50 40 40 30 30 20 20 10 10 0 0 Eritrean Somali South Sudanese Addis Ababa All Hosts Eritrean Somali South Sudanese Addis Ababa All Hosts Strongly agree Agree Disagree Strongly disagree Comfortable Neutral Not comfortable Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. vii Executive Summary Continue implementation of progressive policies: ◆ Address challenges in accessing business ◆ Implement concrete actions to fulfill licenses for refugee self-employment, including Government pledges and proclamations to access to finance. move away from encampment toward mobility based on economic opportunities. Improve cooperation and coordination ◆ Harmonize national and sub-national laws to ◆ Invest in and accelerate inclusive approaches support the full implementation of refugee to economic opportunities and self-reliance to protection. support the GoE in implementing the Refugee ◆ Coordinate efforts among stakeholders to track Proclamation of 2019. progress and share best practices. ◆ Define better coordination and engage line ◆ Redesign the out-of-camp policy (OCP) ministries to achieve better outcomes for to encourage mobility to realize greater refugees and their hosts. socioeconomic opportunities for refugees ◆ Improve the coverage, accuracy, reliability, while accelerating and automating issuance quality, and comparability of data to provide the of work authorizations to enable sustainable analytical underpinning for policy decisions. improvements in refugees’ lives.   viii 1. Introduction C onflict, political unrest, environmental disruption, and economic instability has forcibly displaced millions of people globally extreme poverty and achieve the Sustainable Development Goals (SDGs). (Ferris, 2010; Black, 2001). Over the last decade, Ethiopia has a long history of hosting refugees and the number of forcibly displaced persons has has one of the largest refugee populations in Africa. continuously increased. In mid-2023, there were The refugee situation in Ethiopia is characterized 36.4 million refugees worldwide (UNHCR, 2023d). by both complex humanitarian emergencies and As development reduces global poverty, extreme protracted refugee status. Forced displacement is poverty is increasingly concentrated among a pressing issue in the country, a result of conflict, vulnerable groups; refugees are among these drought, flood, economic instability, and political vulnerable groups (World Bank, 2017). Therefore, instability in neighboring countries (Martin, 2010; the plight of the forcibly displaced poses significant UNHCR, 2020d; IPCC, 2019). As of year-end 2023, challenges to broad development efforts to eradicate more than 922,000 refugees and asylum seekers Figure 1.1: Refugees and asylum seekers in Ethiopia by country of origin, 1984-2023 1,000,000 800,000 Population 600,000 400,000 200,000 0 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 Years Sudan Somalia South Sudan Eritrea Kenya Other countries Source: UNHCR Refugee Data Finder 2024. 1 Introduction were seeking refuge in Ethiopia, with the majority (CRRF), a global framework for a comprehensive originating from South Sudan (420,000), Somalia response to refugee situations. After endorsing the (280,000), Eritrea (170,000), and Sudan (49,000). Global Compact on Refugees (GCR) in 2018, Ethiopia Ethiopia is a signatory to the 1951 UN Convention continued its commitment by adopting a new, on the Status of Refugees and its 1967 Protocol, progressive Refugee Proclamation in January 2019.5 with an obligation to protect refugees and asylum seekers. Most refugees (92 percent) are living in Despite these groundbreaking legal and policy approximately 30 camps and sites located in Afar, actions, many refugees in Ethiopia remain poor Amhara, Benishangul-Gumuz, Gambella, Somali, and depend heavily on humanitarian aid. This and Tigray regions, with an increasing number of report highlights that Ethiopia’s progressive policy refugees living in the capital city of Addis Ababa framework has not yet translated into tangible (70,000) (UNHCR, 2023e). The camps are in different socioeconomic outcomes for refugees. Refugees locations with ethnic and language linkages are mainly living in camps, and few refugees benefit between the refugees and the host community. They from the progressive policy framework. Delays in are spatially dispersed, have different geographic, implementing policies makes it difficult for refugees social, and economic contexts, and are in different to encourage mobility to access better economic ecological zones. For example, about 38 percent opportunities or to access land or finance, or get a of refugees live in drought-prone lowland and work permit to work in the labor market outside of pastoralist areas, whereas 60 percent of refugees are refugee camps. in humid reliable lowland areas. The GoE has a clear long-term vision to address Ethiopia made significant progress articulating the refugee situation in Ethiopia through gradual a more progressive and comprehensive refugee transformation of the existing refugee response response.3 The Government of Ethiopia (GoE) made model into a more comprehensive approach. Until a groundbreaking shift in its refugee policies over recently, Ethiopia’s refugee response model has the last few years, especially since 2016, shifting focused on protecting and assisting them in camps, its refugee policies from an encampment approach where services are delivered through parallel systems toward greater socio-economic inclusion. The GoE typically financed externally. In many refugee- has adopted several national laws and policies to hosting areas in Ethiopia, except for a few areas such protect refugees and ensure respect for their rights. as Addis Ababa, refugees and host communities The Refugee Proclamation—the primary legal share cross-border cultural and economic framework that governs refugee protection and connections, common ties of kinship, language, management in Ethiopia—was enacted in 20044. In and ethnicity, and relatively fluid attachments to 2017, Ethiopia became the first country to fully adopt national identity (see Annex A for more details). The the Comprehensive Refugee Response Framework refugee camps and sites in the country span a broad 3 Annex B summarizes the evolution of refugee policies in Ethiopia. 4 In 2004, the country enacted its first national refugee proclamation that granted restricted rights to refugees. The Refugee Proclamation #409/2004 was not comprehensive enough to improve protection and assistance, promote sustainable solutions for refugees, and support host communities per international standards. The previous refugee law did not reflect the recent policy commitments of the Government and did not confer legal standing to their implementation. For a long period of time, it has had a limitation, particularly in terms of the various privileges that are newly accorded by the revised refugee law to both asylum seekers and refugees, whether as equal to that of foreigners residing in the country or the same as Ethiopian nationals. These include rights to access services, work, move freely, and locally integrate. 5 To complement the Proclamation three directives came into effect on 30 December 2019 namely: Directive to Determine the Conditions for Movement and Residence of Refugees Outside of Camps, Directive No.01/2019; Directive to Determine the Procedure for Refugees Right to Work, Directive No. 02/2019; and Refugees and Returnees Grievances and Appeals Handling Directive, Directive 03/2019. These secondary legislations will have huge contribution to the proper interpretation and implementation of the country’s refugee law. 2 Introduction range of protractedness6 and some camps are in well and brings benefits to host communities similar locations with few economic opportunities, making to those of voluntary migrants. A recent review of the refugees dependent on humanitarian assistance for literature on refugee effects on host communities years and often decades. As one of the champions showed that most studies find a positive or non- of the CRRF, GoE aims to enhance the self-reliance significant effect of forced displacement on hosts’ and resilience of refugees and host communities and employment, wages, and household well-being. prepare them for durable solutions by supporting This finding is contrary to popular perceptions their socio-economic integration and strengthening (Verme and Schuettler, 2021). In some exceptional their contribution to the country’s socio-economic cases, a refugee influx creates challenges for host development.7 communities, typically related to exacerbating existing imbalances, negative outcomes for specific Across the world, including Ethiopia, refugees groups who directly compete with refugees in tend to be poorer than most host populations. In the labor market, and overburdening public Uganda, for example, 46 percent of refugees lived infrastructure or services (Hanafi et al., 2021). in poverty, compared to 17 percent of hosts, in 2018 (World Bank, 2019). In Kalobeyei Settlement Deteriorating economic conditions and soaring in Turkana County in Kenya, more than half of inflation rates exacerbate the already challenging refugees are poor (58 percent), higher than the conditions for refugees and host communities. national poverty rate of 37 percent, lower than the The country is grappling with a complex array of poverty rate in Turkana County but comparable to emergencies, with increasing needs for solutions for the average poverty rate of the 15 poorest counties refugees. These challenges were intensified by an in Kenya (UNHCR and World Bank Group, 2020). economic downturn marked by persistent inflation, About 72 percent of registered Venezuelans in Brazil which has consistently outpaced the Sub-Saharan live in extreme poverty, compared to 48 percent of African regional average over the past decade. The Brazilians (Shamsuddin et al., 2021). Similarly, this persistent inflation is driven by various factors, report finds that poverty rates in Ethiopia’s refugees including supply-demand imbalances, unrest, high in camps are much higher than for host communities. global commodity prices, and relaxed monetary and fiscal policies. In 2022, the average inflation Yet, refugee inflows can significantly affect host rate stood at 34 percent. The surge in prices has communities. Governments have been preoccupied rendered necessities unaffordable for many, hitting with whether the arrival of large numbers of people marginalized populations the hardest, including in specific locations creates risks or opportunities for refugees. The economic turmoil is further fueled by decades. Experience has shown that opportunities ongoing conflict and instability, adding complexity typically result if the influx of refugees is managed and uncertainty. 6 The oldest camp has operated for 31 years (Kebribeyah refugee camp in the Somali region), the newest (Alemwach in the Amhara Region) was established only in as a response to the war in Tigray. 7 The Ethiopian Refugees and Returnees Service (RRS, formerly the Administration for Refugee and Returnee Affairs (ARRA)) is responsible for implementing this long-term strategy and coordinating country-level refugee assistance and protection programs. ARRA was first established in 1992 as the main government department responsible for refugee affairs, housed within the former National Intelligence and Security Services (NISS). The former ARRA was elevated to an agency level in accordance with Proclamation No. 1097/2018, which defines the Powers and Duties of the Executive Organs of the Government and established the Agency for Refugees and Returnees Affairs (ARRA) under the Ministry of Peace in 2018. In 2021, a new government announced through the Definition of Powers and Duties of the Executive Organs Proclamation No. 1263/2021, during which it reestablished ARRA as Refugees and Returnees Service (RRS). The RRS became one of the executive organs accountable to the NISS which is accountable to the Prime Minister’s Office and oversees the Immigration and Citizenship Service other than RRS. 3 Introduction Refugees in Ethiopia have been severely affected and flood events on food security is particularly by ongoing conflict and unrest across Ethiopia. acute among refugee-hosting communities, notably The country has dealt with multiple crises, including in the Somali and Afar regions. The drought has rampant inflation, a devastating war, and frequent destroyed agricultural production, leading to severe droughts and floods. SESRE data collection was food shortages for refugees and host communities. carried out between November 2022 and January In the Somali region, the Fafan and Siti Zones have 2023, marked by drought, inflation, and insecurity, been hit hard by drought events. Amhara region— posing significant threats to the livelihoods of the most severely affected North Gondar Zone— refugees and host communities in an already has also suffered from below-average rainfall, fragile economy. The conflict in Northern Ethiopia causing crop failures, livestock deaths, and continued until 2022 and disrupted the socio- worsening food insecurity. Similarly, the drought economic conditions in the North Gondar Zone, has impacted around 250,000 people across five home to over 20,000 refugees. The November 2022 districts in the Central Gondar Zone, resulting in peace agreement between the federal government a significant decrease in water availability with and Tigrayan authorities offered a glimmer of hope dire consequences for the health and nutrition of for ending the war in Northern Ethiopia. Still, tensions the population. Overall, the effects of the drought in other parts of Ethiopia continued. Additionally, have been particularly pronounced in regions that following the outbreak of fighting in the Amhara host refugees. region in 2023, refugees in the Alemwach camp in the Amhara region faced attacks by unidentified armed Challenges related to hosting refugees can groups. Moreover, food assistance for refugees has be overcome through national development been unstable due to funding shortfalls, resulting strategies that keep both host communities and in reduced food rations for hundreds of thousands refugees in mind. Host communities have their of refugees for several months in 2022 and 2023. own development priorities and needs. Supporting Insecurity in the Gambella and Benishangul-Gumuz them in managing these new circumstances to regions not only undermined refugee and host facilitate their poverty reduction can create a more community livelihoods but also heightened tensions accepting environment for refugees. In Ethiopia, between them. The intensification of conflict in among the major refugee-hosting regions, four— Western Oromia further disrupted humanitarian Afar, Benishangul-Gumuz, Gambella, and Somali— operations in Eastern Benishangul-Gumuz, blocking are designated as “emerging regions,” and the transport of relief and commercial supplies and Tigray is considered post-conflict. These regions affecting 76,000 refugees. are the least developed regions in the country, characterized by harsh weather conditions, poor The concerning economic outlook in Ethiopia infrastructure, low administrative capacity, high is exacerbated by the prolonged and recurrent poverty, and poor development outcomes. The arid drought the country has seen in decades, affecting environment in the Afar and Somali regions, and the vast swathes of the southern and eastern regions, small and scattered nomadic populations, make including refugee-hosting areas. This prolonged it more challenging to provide services. Focusing drought has heightened vulnerabilities, leading to on development interventions that benefit both widespread food insecurity and increased exposure communities can foster improved socioeconomic to diseases, including multiple outbreaks of outcomes and social cohesion. waterborne diseases. The impact of these drought 4 Introduction 1.1 How can we achieve better development development approaches centering around mobility outcomes for all? within the host country enable refugees to move The GoE has recently bolstered development- where economic opportunities are highest and oriented initiatives to complement humanitarian allow them to contribute to the local economy more interventions and improve the lives of refugees productively. and host communities. Yet, most refugee responses in the country remain humanitarian-focused. While refugees in Ethiopia face challenges accessing Better integration and increased attention to the labor market, the GoE has taken positive steps easing the pressure on host communities can to support their economic integration and self- further support refugees and their hosts. But how reliance. Restrictions on the right to work affect can this be achieved? Development approaches refugees in many countries. Only 75 out of the 145 are most successful when they focus on building signatories to the Refugee Convention grant the self-reliance—including offering refugees secure right to work (Zetter and Ruaudel, 2016). Even terms of stay, mobility to access better economic countries that grant access to the formal labor opportunities, and access to the labor market— market in the same way as nationals—Burkina and supporting refugees’ pursuit of economic Faso, Cameroon, Democratic Republic of the opportunities while simultaneously supporting Congo, Djibouti, Mauritania, Niger, and Rwanda— refugee-hosting communities (Betts et al., 2014; often restrict access in practice by requiring certain Clements et al., 2016; Krause and Schmidt 2020). identification documents, the country employers are This not only can improve refugees’ outcomes, it reluctant to hire refugees (World Bank Group, 2021), can also reduce the burden on host communities by or restrictions exist, such as wait periods, limited reaping economic benefits from refugees’ presence. mobility, owning property, or accessing finance. The path of self-reliance includes, at a minimum: In other countries—Burundi, Chad, Uganda, and (i) encouraging mobility within the host country to Ethiopia—access to the labor market is limited by access better economic opportunities; (ii) enabling regulations, such as requiring work permits, caping and incentivizing labor market participation; and (iii) the percentage of foreign workers, or restricting providing access to education for refugee children. work to certain sectors of employment (World Bank Group, 2021). In addition to wage employment, Many refugees do not enjoy mobility in their host self-employment can be an important avenue, but territories, including choice of residence.8 Globally, in many countries, access to self-employment is one-third of refugees are prevented from moving restricted for refugees, including in Ethiopia where freely (UNHCR, 2022e). About 27 percent of the refugees require business licenses. refugee population is contained in camps, leading to a situation in which they cannot be self-reliant Overall, while the GoE has made progress in and thereby improve their economic opportunities creating a legal framework for refugees to obtain to reduce their dependence on support from their work permits, refugees still face significant hosts and the international community (World challenges accessing employment, contributing to Bank, 2017). In Ethiopia, 88 percent of refugees their overall vulnerability and lack of self-reliance. live in camps. Denying refugees mobility to settle As this report shows, few refugees in Ethiopia where they would like comes at a cost, as the work, and those who do work mostly inside camps. choice of location within the host country matters Research indicates that extended periods of forced for refugees’ labor market outcomes. Therefore, unemployment negatively affect refugees’ longer- 8 As granted under Article 26 of the 1951 Geneva Refugee Convention. 5 Introduction term labor market participation (Hainmueller et articulating more progressive and comprehensive al., 2016; Hvidtfeldt et al., 2018; Brell et al., 2020), refugee responses. In 2023, the GoE included thereby also hampering the host society through new pledges and commitments made in the 2023 larger expenditure on assistance and forgone taxes Global Refugee Forum, including a significant shift (Marbach et al., 2018; Fasani et al., 2022). Enabling in its refugee management policies. This includes refugees’ labor market participation outside of improving the socio and economic opportunities refugee camps as early as possible is key to achieving for refugees through an agenda to transform camps their integration (Fasani et al., 2022; Slotwinski et al., into human settlements and including refugees in 2019) by limiting long-term scarring effects, such as national services for education, including secondary long-term unemployment or inactivity. education and health (UNHCR, 2024). Systematically collecting high-quality data on refugees and their Integrating refugee children into education hosts in one survey is pertinent to inform the GoE’s soon after arrival avoids loss of valuable years of roadmap and programs to address the development education and harm to human capital accumulation needs of refugees and hosts. The national household that hinders future prospects. Globally in 2019, survey of Ethiopia–Household Welfare Statistics almost half of all refugee children were out of school Survey (HoWStat)—excludes the majority of (UNHCR, 2020d). Of those attending school, most do displaced populations (Internally Displaced People not make it past basic education; gross enrolment in [IDPs] or refugees) from its sample of households. primary education stood at 77 percent in 2019. Yet, Thus, we have limited in-depth information on the the contrast between primary and secondary school socio-economic outcomes—including on poverty— enrolment remains stark, and only 31 percent of for refugees across all camps in Ethiopia to compare refugee children were enrolled in secondary school, with Ethiopian hosts. much below the global average of secondary school enrolment. A recent study of refugee children in SESRE collected data from November 2022 to Kakuma refugee camp in Kenya, for example, found January 2023, from a nationally representative that literacy and learning outcomes for refugee sample of 3,452 refugee households and their children were significantly lower than in immediate hosts. The SESRE covers all currently operating host community or the rest of Kenya (Piper et al., refugee camps of major refugee groups: Eritreans, 2020). This report shows that education outcomes South Sudanese, and Somalis, as well as the out-of- for children in Ethiopia are low across all population camp refugees of Addis Ababa and their respective groups and ages but particularly for refugee host communities. The survey was aligned with children. COVID-19 exacerbated this situation for the HoWStat methodology, allowing comparability many refugee children (Wieser, 2020). between refugees and their host communities. The World Bank, Ethiopia’s RRS, Ethiopia’s Statistical 1.2 How does Socio-Economic Survey of Service, and UNHCR collaborated to implement Refugees in Ethiopia (SESRE) contribute to SESRE10 and was the first of its kind, building on the debate on policies? the “Skills Profile Survey 2017, A Refugee and Host The Socio-Economic Survey of Refugees in Community Survey” conducted in Ethiopia in 2017. Ethiopia (SESRE) is a representative survey of The SPS 2017 was conducted in refugee camps the refugee population in Ethiopia and their host and host communities in four regions in Ethiopia. communities, the first of its kind.9 Ethiopia made The survey was used to draw a profile for skills significant progress over the past few years in and potential opportunities for refugees and host 9 See Annex C for detailed information on survey design and methodology. 10 Financial support was provided by the World Bank and UNHCR Joint Data Center on Forced Displacement. 6 Introduction communities to design a better mix of approaches methodology, results based on the SPS cannot be that could help the government in designing directly compared with those of SESRE. Box 1.1 livelihood opportunities for these communities. summarizes the similarities and differences between Due to differences in scope, sampling design, and SPS 2017 and SESRE 2023. Box 1.1: Comparison of SPS 2017 and SESRE 2023 The Skills Profile Survey (SPS), conducted in 2017, is a household survey focused on collecting data on refugees from South Sudan, Somali, Eritrea, and Sudan living in camps in Ethiopia, as well as from host communities. The sample frame for the survey was derived from the list of all refugee camps, sites, and locations provided by UNHCR-Ethiopia as of January 2017, covering the four main regions that host refugees: Tigray, Afar, Gambella, Benishangul-Gumuz, and Somali. The SPS specifically excludes refugee households living out of camp, thereby making it representative of the refugee population residing in camps in Ethiopia. In contrast, SESRE, carried out in 2023, expanded its data collection to include out-of-camp refugees living in Addis Ababa. The SPS and SESRE both utilized stratified sampling designs but with different methodologies and definitions of the host households. The SPS employed a multi-stage stratified random sampling approach, dividing refugee camps into EAs of 150 by 150 meters using GIS technology. The number of EAs selected from each camp was proportional to the size of the camp, ensuring all camps in the sample frame were surveyed. In this way, all the camps in the sample frame were selected in the sample and were surveyed. For host households, areas within 5-kilometer radius of the camps were divided into EAs of 300 by 300 meters, with only residential EAs as per Open Street Maps included in the sample frame. SESRE, on the other hand, used a stratified, two-stage cluster sample design. Initially, camps were divided into EAs, and pseudo EAs were created from the proGRES database by grouping 150-200 households consecutively. EAs and households within those EAs were then selected. For host households, EAs adjacent to refugee camps were used as the sampling frame. While the definition of host households differed between SPS and SESRE, both surveys shared similarities in the selection of EAs and the random sampling of households within those EAs. In SPS, all households within the selected EAs for host community sampling were listed, and 12 households were randomly chosen and surveyed per EA. SESRE also selected 12 refugee and host households per EA, treating EAs as the Primary Sampling Unit and households as the Secondary Sampling Unit. (i) The distinct sampling designs and objectives of the two surveys render challenges in comparing findings from the two surveys. Moreover, the two surveys are not comparable in other ways, including: (ii) Surveyed population: SESRE includes out-of-camp refugees living in Addis Ababa, while SPS does not. (iii) Survey scope: The methodology to sample and definitions of host communities varied between the two surveys. (iv) Survey design: SESRE’s questionnaire aimed at comparability with the national poverty survey, while SPS aimed at comparability across countries. This rendered differences in the contents of the surveys, where the SESRE employed the same survey instrument as the national poverty survey, while the SPS used an instrument specific to the survey. (v) Differences in consumption: SESRE includes a full consumption module while SPS relied on the Rapid Consumption methodology. Moreover, there are differences in the recall period for food consumption data collection. While SPS used the past 7 days recall period, SESRE collects food consumption data through two visits—last 3 days and past 4 days—that leads to considerable difference in consumption aggregates. 7 Introduction Box 1.1: Comparison of SPS 2017 and SESRE 2023 (vi) Differences in poverty estimation: The poverty measurement methodology is distinct for each survey with SESRE applying the same methodology as HoWStat. Despite the difference in methodology used in the Skills Profile Survey (SPS) and SESRE, we find similar patterns in some of the indicators common in both surveys among in camp refugees such as demographic composition, primary and secondary net enrollments, housing condition, access to basic infrastructures, employment and attitude of hosts toward refugees. Moreover, both surveys find that in camp refugees in Ethiopia are much poorer than host community households and poverty rates are heterogenous across refugee groups: South Sudanese refugees have the highest incidence of poverty, while Eritrean refugees have the lowest poverty incidence amongst the refugees. However, different poverty estimation methodologies and different poverty lines are used. Likewise, both surveys indicate that refugees are more food insecure than the host community. Annex F, Table F.1 shows a summary of these findings. Source: Pape, U. J., Petrini, B., and Iqbal, S. A. (2018). Informing Durable Solutions by Micro-Data: A Skills Survey for Refugees in Ethiopia. This report uses SESRE data to describe the of refugees and their hosts, covering labor market socioeconomic dimensions of refugees and their outcomes for refugees inside and outside of camps host communities. By highlighting socioeconomic and those of hosts living in the vicinity of refugees, outcomes of refugees and hosts in Ethiopia, we aim as well as a subsection on labor market outcomes of to provide analytical underpinnings for development youth. Chapter 4 dives deeper into refugees’ future interventions in Ethiopia. The analysis focuses on aspirations and their feeling of personal control over aspects such as economic activity, livelihoods, their lives. Chapter 5 describes the welfare situation welfare patterns, as well as on social dynamics and of refugees and their hosts by: (i) understanding longer-term socioeconomic viability of refugee host different dimensions of welfare, such as monetary areas. Focusing on socioeconomic interaction, social poverty, inequality, multidimensional poverty, inclusion, and relations among refugees and between food security, and shocks; and (ii) understanding refugees and their host communities, this work aims determinants of welfare and estimating the cost to inform policies and operations (humanitarian to meet basic needs through a combination of actors, development partners, and government) to assistance and some economic inclusion of refugees facilitate refugee integration and their lives, along into national systems. Chapter 6 aims to understand with hosting communities. how the location of camps determines labor market outcomes, highlighting the importance of refugees’ This report’s eight chapters aim to location as part of the development strategy for comprehensively provide an overview of SESRE refugees in Ethiopia. Chapter 7 looks at social results, with the final chapter highlighting cohesion by showcasing attitudes between refugees policy implications. This Chapter 1 introduces the and hosts and the level of social integration of refugee situation in Ethiopia. Chapter 2 presents refugees. Chapter 8 highlights policy directions the sociodemographic profile of refugees and based on the conclusions of the report to maximize their hosts, including demographic characteristics, the benefits of hosting refugees while minimizing education, health, and living conditions. Chapter 3 the costs. provides an in-depth profile of jobs and livelihoods 8 Sociodemographic Profile 2. Sociodemographic Profile T his chapter presents results sociodemographic outcomes of refugees and hosts. It provides the context for refugees and on respectively (Figure 2.1). More than 30 percent of refugees in camps are born in Ethiopia (Figure 2.2). Somali refugees have a higher share of refugees their hosts, covering demographic characteristics, born in Ethiopia (38 percent). According to UNHCR human capital, living conditions, and displacement estimates (2023)12, around 1.9 million children were experience, which are crucial to understand refugees’ born as refugees between 2018 and 2022 globally. context in Ethiopia. These results are presented across the eight domains: Eritrean, Somali, South Overall, Ethiopia’s refugee situation is protracted; Sudanese, and refugees in Addis Ababa and their refugees have been in Ethiopia for an average hosts, as well as broad categories between in-camp of about 14 years. Refugees differ by country of and out-of-camp refugees and hosts. origin. For example, Eritrean refugees have been in Ethiopia for average of slightly more than 16 years, 2.1 Demographic characteristics Somalis for just under 16 years, and South Sudanese In the SESRE sample, most refugees are from South for roughly 15 years. On the other hand, refugees Sudan, accounting for 53 percent of all refugees.11 in Addis Ababa arrived nine years ago, on average South Sudanese refugees reside in camps in Gambella (Figure 2.3a). Globally, the number of refugees in and Benishangul-Gumuz regions. Somali refugees protracted situations—at least 25,000 refugees from living in camps in the Somali region constitute 30 the same country who lived in exile for more than five percent of the refugee sample. Eritrean refugees consecutive years—increased over time, accounting who reside in camps in the Amhara and Afar regions for 40 percent of all refugees in 2021 (World Bank, and Addis Ababa under the Out-of-Camp Policy 2023). In Ethiopia, 95 percent of all refugees live in a (OCP) constitute 5 and 12 percent of the sample, protracted situation, according to SESRE data. 11 According to UNHCR mid-2022 statistics, South Sundanese, Somali and Eritrean refugees constitute 46, 29 and 18 percent of the total refugee populations in Ethiopia. (https://www.unhcr.org/refugee-statistics/download/?url=2bxU2f) https://www.unhcr.org/refugee-statistics/insights/explainers/children-born-into-refugee-life.html 12 9 Sociodemographic Profile Figure 2.1: Refugees by survey domain Figure 2.2: Country of birth 100 90 11% 12% 31% 38% 80 32% 5% 70 60 Percent 50 53% 88% 40 68% 62% 63% 30% 30 20 10 0 Eritrean Somali South Addis Sudanese Refugees South Sudanese Somali Eritrean Addis Refugees Country of origin Ethiopia Other Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. Refugees fled from their country of birth mainly under age 15. This is slightly higher than the 47 due to conflict and violence. Almost all South percent of hosts under age 15. In Addis Ababa, Sudanese refugees, or 99 percent, left their country however, refugees are less likely to be children, because of conflict and violence, as did 91 percent of with roughly 70 percent of refugees and 59 percent Somalis, and 73 percent of Eritrean refugees. On the of hosts between the ages of 15 and 44. other hand, OCP13 refugees came to Ethiopia with the hope of going to a Western country (35 percent), There is no major difference between hosts and to escape from conflict and violence (28 percent), refugees in gender composition. The share of and for social and economic reasons (22 percent) female hosts and refugees is higher for both in- such as education, health problem, marriage or camp and OCP refugees and their hosts. Moreover, family reunification, and employment (Figure 2.3b). the gap between the percentage of females and Eritrean refugees live in Addis Ababa under OCP, males is larger among hosts and refugees in Addis which requires refugees to cover their cost of living Ababa compared to in-camp counterparts. Across without support from the international community. refugees, South Sudanese (54 percent) and the OCP (55 percent) have a higher share of female refugees Age structure is similar between hosts and (Annex D, Table D.1). The proportion of married refugees, but we see differences in age structure individuals aged 18 and above is higher among for in-camp and OCP refugees. While the majority refugees than hosts, except for refugees in Addis of in-camp refugees are children (below age 15) Ababa. Hosts have a relatively higher percentage of and youth (ages 15 to 24), most OCP refugees married individuals. In Addis Ababa, 60 percent of are between the ages of 15 and 44 (Figure 2.4). In refugees are unmarried. camps, about 53 percent of refugees are children Figure 2.3: Refugees arrival in Ethiopia (15 years and above) a. Years since arrival b. Reasons for leaving the country of birth All 14.1 Addis refugees South Sudanese Addis refugees 8.8 Somali South Sudanese 14.9 Eritrean Somali 15.7 0 20 40 60 80 100 Percent Eritrean 16.2 Conflict and violence Natural/man-made disaster Personal threat and political persecution Social and economic reasons Hope to go to a Western country Other reasons Source: World Bank Staff based on SESRE 2023. 13 For details on OCP refugees, please see Box 2.2. 10 Sociodemographic Profile Figure 2.4: Age structure >=65 >=65 45 to 64 45 to 64 25 to 44 25 to 44 15 to 24 15 to 24 <15 <15 -60 -40 -20 0 20 40 60 -60 -40 -20 0 20 40 60 Percent Percent In camp refugees In camp hosts Addis refugees Addis hosts Source: World Bank Staff based on SESRE 2023. Figure 2.5: Gender composition Figure 2.6: Marital status (18 years and above) Refugees 47% 53% Refugees 33% 50% 17% Total Total Hosts 47% 53% Hosts 30% 59% 11% Refugees 45% 55% Refugees 60% 28% 12% Ababa Ababa Addis Addis Hosts 45% 55% Hosts 36% 54% 9% Refugees 48% 52% Refugees 27% 55% 18% In camp In camp Hosts 49% 51% Hosts 22% 66% 12% 0 20 40 60 80 100 0 20 40 60 80 100 Percent Percent Male Female Never Married Married Other Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. Refugees have larger households, younger heads, refugees have the smallest average household and a higher proportion of female heads than hosts. size and the lowest dependency ratio. Refugee Household size is higher among in-camp refugees households are more likely to be headed by women. compared to hosts. Across in-camp refugees, South The share is exceptionally high for South Sudanese Sudanese refugees have the highest number of refugees, where 84 percent are female-headed household members, averaging roughly seven (Annex D, Table D.1). This reflects a large share of members per household. The dependency ratio— women (71 percent) aged 25 to 44 among South the ratio of dependents of those under age 15 and Sudanese refugees (Annex D, Figure D.1). Refugees above age 64—to working members in a household, have younger household heads than hosts, except is also higher for in-camp refugees relative to hosts for Somali refugees, with the youngest household and highest among South Sudanese refugees. OCP heads in Addis Ababa. Table 2.1: Household characteristics In camp Addis Ababa Total Hosts Refugees Hosts Refugees Hosts Refugees Household size 5.2 6.2 3.5 2.7 4.2 5.4 Dependency ratio 1.1 1.5 0.5 0.4 0.7 1.3 Female-headed 44% 73% 45% 58% 44% 69% Head’s age 42.3 39.6 42.0 30.9 42.1 37.6 Source: World Bank Staff based on SESRE 2023. 11 Sociodemographic Profile 2.2 Education about 333,000 Venezuelan children were enrolled in government schools in 2020 (about 3.4 percent of Integrating refugee children into educational the total student population in Colombia) (UNHCR, programs soon after arrival14 avoids the loss of 2021a). In Turkey, the government is supporting valuable years of education and human capital the transition of Syrian refugee children into the accumulation that can hinder future prospects. national school system, redirecting resources to Integrating refugees into national education systems locations with high concentrations of refugees can improve future outcomes for refugee children (Abu-Ghaida and Silva, 2020), resulting in nearly and their hosts (UNHCR, 2020; Piper et al., 2020; 80 percent of Syrian primary school-aged refugee Abu-Ghaida and Silva, 2020; Crawford et al., 2015; children being enrolled in education programs by Bilgili et al., 2019). Investment in human capital 2020/2021 (UNHCR, 2021). In addition to benefiting development can enable refugees to contribute to refugee children, high school enrolment and local economies to benefit refugees and hosts alike, learning outcomes increased for local Turkish and it can contribute to the recovery of countries students (Tumen, 2019; 2021). of origin and the hosting communities. Despite the positive externalities of integrating refugees into public-school systems, a large influx of refugee Educational attainment is low among both hosts children can exacerbate existing inefficiencies. and refugees, especially in camps. About 73 percent Where there are large inflows, the national of in-camp adult refugees and 59 percent of adult education system might require additional human hosts (aged 18 and above) either did not attend and financial resources to integrate newly arrived school or did not complete primary education. South children. Increasing the supply and improving the Sudanese refugees have the worst educational quality of schools in affected areas, supported by attainment. Refugees in Addis Ababa have relatively external assistance and financing, can avoid tension better educational attainment compared to their between refugees and host community populations hosts, with a higher percentage of refugees in over competition for access to education.15 Support Addis Ababa completing primary (45 percent) and is particularly needed in remote areas—where secondary (27 percent) education (Figure 2.7). refugees are often hosted—where educational Although more educated, even youth’s (age 15 to 24) services are already strained for local children (Abu- educational attainment is low, with large differences Ghaida and Silva, 2020). by survey domains. Many youth refugees and hosts have not completed primary education (Figure An inclusive education system can benefit both 2.8). While the percentage of youth who completed refugees and native children. Research has shown primary education is close to 50 percent for hosts, it is that an inclusive education system has positive only 35 percent of in-camp refugee youth. Moreover, externalities for host community children (Abu- there are large differences by location, with only 22 Ghaida and Silva, 2020). A Rwanda study showed percent of Eritrean refugee youths in camps having that local Rwandan children’s school attendance is completed primary school compared to 37 percent higher among communities within a 10-kilometer of in-camp Somali youth. On the other hand, refugee radius of a refugee camp. Other countries have youth in Addis Ababa have similar levels of education also integrated refugee children into the national compared to their hosts (Annex D, Table D.2). education system (Bilgili et al., 2019). In Colombia, 14 In an emergency setting, it is recommend to provide refugees with access to educational programmes within the first 3 months of arrival in the hosting country. 15 Moreover, access to education services is often tied to having identification documents. Enabling refugees to receive identification documents is critical, not only for accessing education but also other services such as health or financial services. 12 Sociodemographic Profile Box 2.1: Education system for refugees in Ethiopia According to Ethiopia’s Refugee Proclamation 2019, refugees are granted access to pre-primary and primary education in the same way as nationals. Whereas, secondary education, higher education, technical and vocational education and training, and adult and non-adult formal education is provided with available resources. At all levels, refugees receive the same education as nationals in terms of curriculum, access to national examinations, and accredited certificates. Yet, the administration of primary schools in camp settings follows a parallel system. For refugees, the RRS administers primary education for in-camp refugees in partnership with UNHCR16. Primary education is also provided in partnership with an NGO (Plan International) in refugee camps in Gambella, Benishangul- Gumuz, and Amhara regions. For hosts, meanwhile, management of primary public schools falls under the responsibility of the Ministry of Education (MoE) and Regional Education Bureaus. Primary education is delivered by a combination of national and refugee incentive teachers.17 It is important to note that all national teachers are qualified, but some refugee incentive teachers receive training to build their capacity. Moreover, there are not enough female teachers in camps, with the majority of teachers being male. Refugee schools also face high turnover of teachers as they take on better-paid jobs (UNHCR, 2017). Secondary education, on the other hand, is provided for refugees in camp-based refugee schools and in government administered public schools, with support from UNHCR’s NGO partner (Development and Inter-Church Aid Commission, DICAC). This relies on qualified national teachers exclusively. UNHCR and partners are working towards progressive transfer of secondary school administration from DICAC to Regional Education Bureaus. The MoE is responsible for managing refugee and national education for higher education. However, enrollment of refugees in higher institutions is low due to low absorption capacity of higher education institutions (UNHCR, 2020a). At the same time, the Ministry of Skills and Labor (MoLS) is responsible for managing Technical and Vocational Education and Training (TVET). OCP refugees receive education at the same levels as nationals through the national education system. Refugee education data is integrated into the MoE’s Education Management Information System (EMIS) (UNHCR, 2020), and a separate chapter on refugee education is included in the annual education statistics report of the MoE. The GoE included expanding primary and secondary education for refugees in the national five-year Education Sector Development Programme VI (ESDP), covering 2020 to 2025, but little progress has been made. Figure 2.7: Education level (18 years and above) Figure 2.8: Youth (15 to 24) education level 100 100 80 80 60 60 Percent Percent 40 40 20 20 0 0 Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees In camp Addis Ababa Total In camp Addis Ababa Total No education Incomplete primary Complete primary No education Incomplete primary Complete primary Complete secondary Complete post-secondary Complete secondary Complete post-secondary Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. 16 In January 2024, RRS handed over the management of refugee primary education to DICAC, Plan International, and EDUKANS in a transition process to eventually transfer responsibility of refugee primary education to the Ministry of Education and the Regional Education Bureaus. 17 Incentive teachers are refugees who teach in return for a small stipend. They may be qualified teachers but at a minimum received training. 13 Sociodemographic Profile OCP refugees attended education outside of highlighting a systematic difference in the selection Ethiopia and have relatively better educational of OCP refugees. attainment. One reason why OCP refugees have higher educational attainment compared to other The percentage of primary school-age children refugees (especially other Eritreans) could be related attending primary education is similar for hosts to being relatively better off in their countries of and refugees. A more liberal education policy origin (qualification for OCP requires having some towards refugee education in Ethiopia increased resources other than international aid); thus, they the likelihood of attending school (World Bank, are more likely to have received education at home. 2023c). In-camp refugees and hosts are less likely A much higher proportion of adult refugees in Addis to receive education than OCP refugees and hosts; Ababa (89 percent) attended education outside in camps, 63 and 61 percent of refugee and host of Ethiopia than refugees in camps (49 percent). children attend school, while 80 and 82 percent of Despite high school attendance rates outside OCP refugee and host children attend school (Figure of Ethiopia, only 23 percent of adult refugees in 2.10a). On the other hand, refugee children and Addis Ababa have education documents, and only youth are half as likely to go to secondary school 15 percent of those with education documents (22 percent) compared to hosts (44 percent) (Figure were able to verify the documents through the 2.10b). There is no difference in primary education responsible Ethiopian authority (Annex D, Figure attendance between boys and girls in refugee and D.2). Concerning educational level, most in-camp host communities, but refugee and host secondary refugees had education below primary level. In school-age girls are less likely to attend secondary contrast, most OCP refugees completed primary school than boys (Annex D, Figure D.4). and secondary education before moving to Ethiopia, Box 2.2: Refugees under the Out-of-Camp Policy (OCP) The GoE introduced the Out-of-Camp Policy (OCP) in 2010 to give refugees opportunities to live in Addis Ababa and other non-camp locations of their choice (RRS, 2017). In 2019, the RRS introduced a directive for implementing the OCP, enabling refugees to establish residence outside the camp to broaden employment opportunities and achieve self-reliance. Refugees who live for more than one month in a camp can apply for a regular OCP residency permit. To be eligible for OCP residency, a refugee should be able to prove they can cover the cost of living or provide a sponsor and receive a work permit. OCP residency permit rules exempt refugees with special conditions (orphaned children, with medical issues, single mothers, elderly, and with urgent overseas travel). Refugees who are no longer beneficiaries of the urban assistance program can also get the permit if they meet the requirements of the OCP residency permit (RRS, 2019). Arrival before and after November 2020 The Tigray region of Ethiopia used to host Eritrean refugees in four camps before the outbreak of conflict between the regional and the Federal government in November 2020.18 Consequently, refugees in the region fled to neighboring Afar and Amhara regions and Addis Ababa.19 Hence, the RRS granted out-of-camp residency for those refugees who arrived in Addis Ababa due to the conflict. As of April 2023, Eritrean refugees relocated from Tigray to Addis Ababa constitute 36 percent of the total Eritrean population in Addis Ababa (UNHCR, 2023a). OCP refugees who arrived before and after November 2020 have similar sociodemographic characteristics except age, education, and child health outcomes. Refugees after November 2020 are younger and less educated compared to refugees before November 2020. Moreover, child health problems in terms of nutritional indicators, underweight, stunting, and wasting are higher among refugees moved from camps relative to refugees who were in Addis Ababa for a longer time. https://www.unrefugees.org/news/ethiopias-tigray-refugee-crisis-explained/ 18 19 https://www.hrw.org/news/2021/09/16/ethiopia-­eritrean-­refugees-­targeted-­tigray 14 Sociodemographic Profile Figure 2.9: Refugees’ education outside of Ethiopia (18 years and above) a. Attended education b. Education level 100 100 90 80 80 70 60 Percent 60 50 Percent 40 40 30 20 20 10 0 Eritrean Somali South Addis 0 Sudanese refugees Eritrean Somali South Addis Incomplete primary Complete primary Sudanese refugees Complete secondary Complete post-secondary Source: World Bank Staff based on SESRE 2023. Figure 2.10: Children currently attending school a. All children b. School-age children 90 100 80 90 70 80 60 70 60 50 Percent Percent 50 40 40 30 30 20 20 10 10 0 0 Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees In camp Addis Ababa Total In camp Addis Ababa Total Attending school Primary school Secondary school Primary school (7 to 14 years) Secondary school (15 to 18 years) Source: World Bank Staff based on SESRE 2023. Primary and secondary school enrollment rates based refugees, primary NER is higher among South vary between hosts and refugees; refugee primary Sudanese refugees (80 percent) compared to Eritrean education rates are higher than secondary (65 percent) and Somali (62 percent) refugees. enrollment. The Primary Gross Enrollment Rate (GER)20—which shows the share of children going Despite similar enrollment rates in primary to school—is similar between hosts (98 percent) compared to their hosts, refugees struggle to and refugees (97 percent). For South Sudanese attend secondary education. Secondary GER22 and and Somali refugees, primary GER are higher than NER23 for refugees are almost half those of hosts. For their respective hosts. However, the primary Net example, only 23 percent of secondary school-aged Enrollment Rate (NER)21—which shows whether refugee children and youth attend secondary school. children attend education at the right age—is This share is higher (41 percent) for hosts. Refugee higher for hosts (76 percent) compared to refugees Secondary NER also is very high compared to the (70 percent), though the gap is small. Across camp- national 5 percent in 2021/22 (MoE, 2022). By country 20 The Primary Gross Enrollment Rate (GER) is the ratio of the number of children enrolled in primary school irrespective of age to the number of children of primary school age (age 7 to 14). 21 The Primary Net Enrollment Rate (NER) is the ratio of the number of children of primary school age enrolled in primary school to the number of children of primary school age (age 7 to 14). 22 The secondary Gross Enrollment Rate (GER) is the ratio of the number of children enrolled in secondary school irrespective of age to the number of children of secondary school age (age 15 to 18). 23 Secondary Net Enrollment Rate (NER) is the ratio of the number of children of secondary school age enrolled in secondary school to the number of children of secondary school age (age 15 to 18). 15 Sociodemographic Profile Figure 2.11: Gross Enrollment Rate (GER) Figure 2.12: Net Enrollment Rate (NER) 120 120 100 100 80 80 Percent Percent 60 60 40 40 20 20 0 0 In camp Addis Total In camp Addis Total In camp Addis Total In camp Addis Total Ababa Ababa Ababa Ababa Primary GER Secondary GER Primary NER Secondary NER Hosts Refugees Hosts Refugees Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. of origin, Somali refugees have higher secondary 27 and 18 percent of refugees (mainly South Sudanese NER (32 percent) compared to South Sudanese (17 and Somali refugees) and hosts aged 19 to 24 are still percent) and Eritrean (14 percent) refugees. In Addis in secondary schools (Figure 2.13). Ababa, 83 percent of primary-aged refugee children attend primary school, but only 31 percent of Reasons for not attending school differ among secondary-aged refugee children attend secondary children of primary and secondary school age. school (Annex D, Figure D.3). Following the revision Most children who do not attend primary school do of the Refugee Proclamation in 2019, the Ministry so because their families think they are too young of Education (MoE) and RRS have tried to provide or unwilling to send them to school. This is similar primary education to refugee children in the same for hosts and refugees. For secondary school-age circumstances as nationals. However, transition children, the reasons for not attending school differ from primary to secondary school remains very low between refugees and hosts and across refugee among refugees due to, among other things, limited domains. Reasons related to need to work is higher available schools, lack of adequate infrastructure, among hosts (47 percent) compared to refugees (17 and examination bottlenecks24 (UNHCR, 2020a). percent), whereas family unwillingness is higher for refugees (33 percent) than hosts (16 percent). Refugee children face challenges in attending Being unable to attend school due to need to education at the appropriate age. This is also work is higher among host boys than girls, while demonstrated by the fact that many children and Figure 2.13: Share of children and youth above school age youth participate in primary or secondary education in education despite being much older. Among youth aged 15 to 70 18, 61 percent of refugees (driven by South Sudanese 60 refugees) and 31 percent of hosts attend primary 50 education. In South Sudan, 70 percent of children are Percent 40 30 out of school (UNICEF, 2021). Hence, limited access 20 to education in the country of origin contributes to 10 late school entry among South Sudanese refugee 0 Hosts Refugees Hosts Refugees Hosts Refugees children. In Somali refugee camps, parents want their In camp Addis Ababa Total children to attend Quranic schools before attending Primary school (15 to 18 years) Secondary school (19 to 24 years) primary schools, which could contribute to delaying Source: World Bank Staff based on SESRE 2023. school attendance. Regarding secondary education, 24 Grade 12 national examinations for refugees and host communities are administered in nearby government public universities—a long distance for refugees based in remote locations—impacting the performance of the refugee students. 16 Sociodemographic Profile family unwillingness is higher among refugee girls 2.3 Health and nutrition compared to boys (Annex D, Figure D.5). Refugee Experiences before, during, and after displacement parents are less likely to send their children to can have stark consequences on the health of secondary school in part because the opportunity refugees. Before displacement, refugees often live cost of schooling becomes higher since children in countries experiencing economic turmoil and going to school cannot help support the family humanitarian crises. During their flight, refugees (UNHCR, 2020a). Across refugees, Somali refugee face harsh and uncertain conditions. Refugees children of secondary school age do not attend often struggle to integrate and feel accepted when school due to family unwillingness (52 percent). On they arrive at their new destinations. All of this can the other hand, sickness/injury or natural or human- severely affect their physical and psychological caused disasters, such as drought or violence, hinder wellbeing. Yet, many refugees face barriers to children from going to school among South Sudanese accessing health services they need—including and Eritrean refugees (Figure 2.14). Studies show accessing health providers and getting medicines or that refugee children also face challenges to attend medical supplies—due to distance, safety, language, school due to a lack of academic records, mental health issues and (Thomas, 2016), and language policy, or financial constraints. Good health is an barriers (Reddick and Chopra, 2021). essential requirement to rebuild refugees’ lives after displacement. Refugees, like any other population, Given that the international community provides have varied health-related issues, including education in camp settings, refugees spend less on noncommunicable and communicable diseases and education compared to hosts. The average annual trauma from injuries and violence. Research shows expenditure on education per school-age child that conflict inflicts extensive psychological harm among in-camp refugees is much lower than among on many refugees, particularly youth and children, hosts (Annex D, Figure D.6). On the other hand, which often remain unaddressed (Simpson, 2018; refugees in Addis Ababa spend less on education Bosqui and Marshoud, 2018; Dong, 2018). Refugee than their hosts, likely reflecting a much lower share women are specifically vulnerable to sexual of refugee children who attend private education. and other forms of gender-based violence and Almost half of all host children, and 27 percent of require specialized care and access to sexual and refugee children, attend private schools. reproductive healthcare. Figure 2.14: Reasons for not currently attending school South Refugees 15 to 18 years Sudanese Hosts Refugees Somali Hosts Refugees Eritrean Hosts South Refugees Sudanese 8 to 14 years Hosts Refugees Somali Hosts Eritrean Refugees Hosts 0 10 20 30 40 50 60 70 80 90 100 Percent Need to work Unable to cover education expenses (fee and materials) School too far Too young Marriage or pregnancy Family not willing Sickness/injury or natural or human calamites Negative perception towards the benefit of education Other Source: World Bank Staff based on SESRE 2023. 17 Sociodemographic Profile Access to essential health services when and where Receiving health care refugees need them is crucial for allowing them to No significant difference exists between hosts restart their lives. Refugees need access to treatment and refugees regarding prevalence of illness and and preventive care during health emergencies, receiving medical assistance. The proportion the importance of which manifested during the of refugees facing health problems in Ethiopia is COVID-19 pandemic. Aligned with the GCR, refugees slightly higher for in-camp refugees (19 percent) should be able to access essential health services compared to OCP refugees (14 percent). Yet, through the national health systems of the host significant differences exist across survey domains. A countries at affordable costs and sufficient quality. much larger proportion of South Sudanese refugees Ethiopia provides access to healthcare for refugees (25 percent) and their hosts (29 percent) faced health through health centers in camps, with referrals to problems in the two months before their survey services outside of camps for complicated cases interview compared to any other group. Of the South or for secondary and tertiary healthcare. UNHCR, Sudanese refugees and hosts that had health issues, in partnership with the RRS and other operational 65 percent of refugees and 59 percent of hosts were partners, provides primary healthcare services for ill due to malaria. Somali refugees and their hosts in-camp refugees. For refugees in settlement sites,25 have the lowest share of illness (Annex D, Table healthcare service is provided by government health D.3). Of the ill, most hosts and refugees received the centers, through Regional Health Bureaus, and necessary treatment in health institutions. However, through UNHCR partners. In the case of medical in-camp refugees (91 percent) are more likely to get conditions that cannot be addressed by treatment treatment than OCP refugees (71 percent). received in health centers, refugees are referred to nearby zonal and regional hospitals for secondary Refugees access medical services in health care and to hospitals in Addis Ababa for tertiary institutions located inside and outside of camps. care. Refugees with complicated health problems, Most in-camp refugees get medical assistance in or needing long-term regular checkups, are granted health centers and health institutions implemented OCP residency. They receive assistance through by RRS or NGOs within and outside camps. Refugees the urban assistance program, which covers their in Addis Ababa—who, due to their OCP, have medical expenses. Other OCP refugees can access to access healthcare without support from the public or private health services but must pay for international community—get medical services their own medical costs. from private sources (58 percent) and government Figure 2.15: Faced any health problem Figure 2.16: Received medical assistance when faced with health problem 20 100 80 15 60 Percent Percent 10 40 5 20 0 0 Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees In camp Addis Ababa Total In camp Addis Ababa Total Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. 25 Currently, there is only one refugee settlement site in Ethiopia: Alemwach in the Amhara region. 18 Sociodemographic Profile health institutions (40 percent) (health centers and working with the GoE tow include OCP refugees hospitals) (Annex D, Figure D.7). Both refugees and in the Community Based Health Insurance (CBHI) hosts face problems concerning service delivery Scheme. Of the total refugee households living of health institutions, which is higher for in-camp under OCP, 10 percent moved to Addis Ababa to refugees and hosts compared to OCP refugees and access basic social services, such as education and their hosts. Problems mainly relate to shortage or health. Thus, some refugees from the above may unavailability of medicines and long wait times to get receive medical services free of charge. Based on services in camps (Annex D, Figure D.8). Refugees use SESRE data, most OCP refugees (58 percent) rely health facilities outside of the camp, but 66 percent of primarily on private healthcare services, while 40 Eritrean refugees26 and 41 percent of Somali refugees percent access healthcare through the national have better usage of out-of-camp healthcare services system (Annex D, Figure D.7). This can help explain compared to South Sudanese (5 percent). Overall, in- why OCP refugees’ average annual per capita camp refugees with chronic27 illnesses tend to receive expenditure on health is almost twice that of hosts. treatment in health institutions located outside of In-camp refugees have access to healthcare through camps compared to those with non-chronic diseases the international community, so their out-of-pocket (Figure 2.17b). Refugees receive follow-up treatment spending on health is thus very low and much lower for tuberculosis and antiretroviral therapy (ART) than hosts’ out-of-pocket health expenditures. The for HIV/AIDS in camp health facilities and also get difference in per capita health expenditure is large treatment for common illnesses such as asthma, between Eritrean and South Sudanese refugees and diabetes, hypertension, epilepsy, and mental issues. their hosts but low among Somali refugees and their hosts (Annex D, Figure D.12). OCP refugees spend more on health compared to hosts. OCP refugees can no longer rely on Child Nutrition and Health Outcomes international aid sources for healthcare, and do not Child nutrition and health represent a significant have health insurance. However, OCP refugees who challenge for both hosts and refugees. The get out-of-camp residency permits due to health nutritional status of children under age five is based problems receive medical services free of charge on anthropometry measures; that is, stunting, under the urban assistance program. UNHCR is underweight, and wasting. A child is identified as Figure 2.17: Use of the national healthcare system when faced with health problems a. Overall b. By type of illness 70 70 60 60 50 50 40 40 Percent Percent 30 30 20 20 10 10 0 0 Eritrean Somali South Sudanese All in camp Eritrean Somali South Sudanese All in camp Chronic illness Non-chronic illness Source: World Bank Staff based on SESRE 2023. 26 Driven by Eritrean refugees in Alemwach camp who do not have a health facility inside the refugee site and get medical services from government health facilities outside the camp. 27 Chronic illness: tuberculosis, hepatitis-B, asthma, uric acid, blood pressure, diabetes, HIV/AIDS, kidney problem, epilepsy, cancer, mental illness 19 Sociodemographic Profile “stunted”, “underweight”, or “wasted” if height- and Somali refugees compared to their hosts. At the for-age, weight-for-age, and weight-for-height same time, it is lower among South Sudanese and “z-scores28” are more than two standard deviations OCP refugees compared to hosts. In-camp refugee below the 2006 World Health Organization (WHO) and host children also suffer from wasting. The Child Growth Standard medians for these measures. percentage of wasted in-camp refugee children (14 percent) is higher compared to OCP refugee children Child stunting is a major child health problem for (6 percent). Somali and OCP refugees have a higher both hosts and refugees, but stunting rates are proportion of wasted children than their hosts, largest for refugee children. Stunting is the impaired whereas the wasting rate is lower among South growth and development children experience Sudanese refugees compared to their hosts (Annex from poor nutrition, inadequate maternal health, D, Table D.3). and repeated infection during the critical first 1,000 days of a child’s life. Stunting has long- Refugees in camps have better access to health lasting consequences such as impaired cognitive institutions for child delivery than hosts. Access development, health issues, increased mortality, to healthcare during childbirth is crucial for the and reduced earning potential in adulthood. SESRE health of mothers and newborns. In camps, 87 results show high stunting rates for both hosts and percent of refugee mothers give birth to children in refugees, with regional differences. The prevalence health institutions (health centers and hospitals), of stunting is higher among Eritrean refugees (52 while 75 percent of host mothers give birth in health percent) and their hosts (43 percent), followed by institutions. As a result, 91 percent of births among Somali refugees (47 percent) and their hosts (37 in-camp refugees are assisted by skilled health percent). Children in the South Sudanese domain personnel, while the rate is only 77 percent among show the lowest stunting rates, yet 26 percent of hosts. For OCP refugees and hosts, more than 90 refugees and their host children are too short for percent of children are born in health institutions, their age. Children in Addis Ababa have lower shares and all births are attended by professional or trained of stunting, but stunting rates are still high for OCP health workers (Annex D, Figure D.10). refugees (27 percent) and their hosts (24 percent) (Annex D, Table D.3). Overall, stunting rates are A significant share of both in-camp refugees higher for boys than girls among both refugees and and hosts have no registration of births for their hosts (Annex D, Figure D.9), which is also confirmed children. Birth registration provides legal identity by studies using the demographic and health survey for children and ensures protection and access in Ethiopia (Tasic et al., 2020; Gebreegziabher and to essential services such as health, education, Regassa, 2019; Gebru et al., 2019). and justice (UNICEF, 2019). Yet, 51 percent of refugees and 56 percent of host children have no Being underweight is another large challenge birth evidence (either vaccination card or birth related to nutrition among children under age certificate). Availability of birth evidence is better five in Ethiopia. In-camp refugees and hosts have a for Somali refugees (67 percent) compared to other higher percentage of underweight children (25 and refugees and hosts (38 percent). In Addis Ababa, 29 percent) compared to OCP refugees and hosts birth documentation is available for over 90 percent (2 and 11 percent). Across refugees, the proportion of refugee and host children born there (Annex D, of underweight children is higher among Eritrean Figure D.11). 28 z-scores are calculated as (X-m)/SD, where X is child height, weight or age, m and SD are the mean and standard deviation value of the distribution corresponding the reference population (2006 WHO Child Growth Standards). 20 Sociodemographic Profile Figure 2.18: Child nutritional indicators 40 35 30 25 Percent 20 15 10 5 0 In camp Addis Ababa Total In camp Addis Ababa Total In camp Addis Ababa Total Stunted Underweight Wasted Hosts Refugees Source: World Bank Staff based on SESRE 2023. Disability 2.4 Living conditions Disability rates are similar for hosts and refugees. Housing differs drastically between refugees and The proportion of individuals with a disability29 is their hosts, with in-camp refugees living in UN or comparable between refugees and hosts, except NGO shelters, and refugees in Addis Ababa residing for Eritrean refugees. The percentage of persons in rented houses. Most Eritrean refugee households with disabilities is higher among Eritrean refugees (61 percent) live in temporary30 shelters provided by (8 percent) compared to hosts (5 percent). At the UN or NGOs, whereas 50 percent of Somali refugee the national level, and estimated 9 percent of households and 71 percent of South Sudanese the population lives with at least one disability, refugee households live in UN or NGO-provided according to 2016 national survey results (UNICEF, permanent31 shelters. In Addis Ababa, 97 percent 2018). Disability is more prevalent among elderly of refugee households live in rented houses. SESRE refugees and hosts who above age 60 compared to data also show that OCP refugees pay higher rents adults and children. Refugees and hosts mainly face than hosts; on average, refugees pay roughly ETB disabilities related to seeing, walking, or climbing 31,600 per year per adult equivalent, while hosts pay steps (Annex D, Figure D.13). slightly less than half of that (ETB 18,700) (Annex D, Figure D.14). Refugees do not qualify for government Figure 2.19: Presence of any disability a. Overall b. By age group 6 40 5 30 4 Percent 20 Percent 3 10 2 1 0 Hosts Refugees Hosts Refugees Hosts Refugees 0 In camp Addis Ababa Total Hosts Refugees Hosts Refugees Hosts Refugees In camp Addis Ababa Total <=18 years >18 to 60 years >60 years Source: World Bank Staff based on SESRE 2023. 29 At least having difficulty with seeing, hearing, walking, remembering, selfcare or communicating. 30 Temporary shelters have walls mainly made of tent, plastic cover, and irons sheet. 31 Permanent shelters have walls mainly made of wood, mud, non-plastered blocks. 21 Sociodemographic Profile Figure 2.20: Dwelling type Figure 2.21: Housing quality 100 120 80 100 80 60 Percent Percent 60 40 40 20 20 0 Hosts Refugees Hosts Refugees Hosts Refugees 0 Hosts Refugees Hosts Refugees Hosts Refugees In camp Addis Ababa Total In camp Addis Ababa Total Owned Rented UN/NGO temporary UN/NGO permanent Other Overcrowded Improved wall Improved roof Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. housing schemes (such as Kebele housing). OCP percent. Housing conditions in terms of wall and roof refugees thus tend to rent in the private housing construction materials is worst for South Sudanese market, typically more extensive and better quality, refugees, none of which live in a house with an but which increases the cost of renting. High rents improved wall, with only 8 percent of households are a large challenge for OCP refugees, with rent having an improved roof (Annex D, Table D.4). expenditure taking the highest share of their total non-food expenditure (56 percent), compared to Refugees have better access to drinking water only 37 percent for host households in Addis Ababa. compared to hosts since the international community provides water, sanitation, and hygiene Housing quality varies across camps. Housing (WASH) services. The share of in-camp refugee quality is measured using three indicators: households with access to safe drinking water35 is overcrowding, quality of the wall, and roof higher than for host households. This is not surprising construction materials. Both in camps and in considering that the international aid sources Addis Ababa, refugee households live in more prioritize access to drinking water when setting up overcrowded32 conditions compared to hosts. camps. South Sudanese hosts have relatively lower Overcrowding is highest among Eritreans, with 66 access to safe drinking water. Despite good access percent of refugee households living in dwellings to drinking water in camps, the proportion of in- with more than three people per room. Most in-camp camp refugee households with improved bathing refugees and their respective host households live in facilities36 is low, especially among Somali refugees homes with low-quality walls,33 with only 2 percent and their hosts. In addition, few refugee and host of in-camp refugee households live in dwellings with households have a place or item designated for hand an improved wall. In Addis Ababa, the percentage washing in their dwellings. Availability of water or of refugee households living in houses with good detergent for hand washing is low, especially among quality walls is 82 percent, even higher than hosts at refugees (Annex D, Figure D.15). Regarding rented 58 percent. This is related to the fact that refugees homes, OCP refugee households live in houses with cannot access public housing schemes, such as better bathing facilities (82 percent) than hosts (66 Kebele housing, often of lower quality. Regarding the percent). Also, around 80 percent of refugee and quality of roofing, more than half of refugee and host host households in Addis Ababa have a place for households live in dwellings with improved roofs,34 hand washing, and more than half of refugees and except for South Sudanese refugees which have 8 hosts have water or soap. 32 Overcrowding occurs when if more than three people live per room (UN-Habitat). 33 Improved wall is made of stone & cement, blocks-plastered with cement or bricks. 34 Improved roof is made of corrugated iron sheet or concrete/cement. 35 Improved sources of drinking water are piped, bottled, sachet, or tanker water. 36 Improved bathing refers private or shared bathtub, shower, separate room for bathing. 22 Sociodemographic Profile Figure 2.22: Access to drinking water and hygiene Figure 2.23: Access to toilet facility and waste disposal 100 120 80 100 80 60 Percent Percent 60 40 40 20 20 0 0 Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees In camp Addis Ababa Total In camp Addis Ababa Total Improved source of drinking water Improved bathing facilities Improved toilet facility Improved waste disposal method Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. Refugees’ access to improved sanitation facilities Eritrean refugees, with 78 percent of Eritrean refugee is similar to hosts. Refugees’ access to improved households get lighting from solar energy. Almost all toilet facilities37 is the same as hosts, or even better South Sudanese refugee households have no access in some cases. Eritrean, Somali, and OCP refugees to electricity either from meter or solar sources. All and their hosts have similar toilet facilities. Even refugees in Addis Ababa use electricity for lighting, though the percentage of households with access similar to their hosts. to improved toilet facilities is lower among South Figure 2.24: Source of lighting Sudanese refugees than other refugees, it is higher 100 compared to their hosts. Moreover, refugees have 80 higher access to improved waste disposal methods38 60 Percent than hosts. 40 Both refugee and host households have low access 20 to electricity, except for hosts of Eritrean refugees. 0 Hosts Refugees Hosts Refugees Hosts Refugees Hosts around Eritrean refugees have better access to In camp Addis Ababa Total electricity (meter private or shared) for lighting (74 Electricity (meter) Electricity (meter, generator, solar) percent). The use of solar energy is common among Source: World Bank Staff based on SESRE 2023. 37 Improved toilet facility includes toilets flush to septic tank/pit latrine/piped sewer system, pit latrine with slab, or composting toilet. 38 Improved waste disposal refers waste not thrown to field or yard, into river and burnt. 23 3. Jobs and Livelihoods T his chapter presents findings on labor market outcomes and livelihood choices of refugees and hosts. It discusses how sociodemographic Over the last decade, negative repercussions from the concurrent overlapping crises have threatened this marginal progress. In 2020, the COVID pandemic characteristics such as age, gender, education level, and closed markets, albeit relatively briefly, and this location of residence relate with labor market outcomes. immediately decreased job opportunities. While most reentered the labor market, many changed Ethiopia has experienced steady economic growth their work situation, and some permanently exited for much of the last two decades, but even before the workforce or reduced working hours. the country’s concurrent crises, economic growth did not transform the labor market structure. At the same time, more and more people— Between 2004 and 2020, Ethiopia’s GDP annual rural women in particular—are unemployed or growth averaged 10 percent, helping to reduce the out of the labor market altogether. Nationally, poverty by about ten percentage points. But during unemployment doubled between 2013 and 2021, this period, the distribution of workers across and it tripled in rural areas. Women and youth both sectors and geographies shifted very little. According saw particularly sharp increases in unemployment. to newly released 2021 Labor Force Survey (LFS) There was also a decrease in the labor force data, about 80 percent of Ethiopians live in rural participation rate (LFPR), from 86 percent in 2013 areas, where roughly 75 percent work in agriculture. to 74 percent in 2021, following a long period of In urban areas, approximately 70 percent of people steady LFPR in the two previous LFS surveys. Like work in services. Nationally, self-employment for unemployment, LFPR was much more affected accounts for about half of jobs, with unpaid family in the rural labor market, and these trends are work second most common and wage work a distant particularly striking for women. third. Again, the urban context is different: wage employment is prevalent but often poorly paid. 24 Jobs and Livelihoods Within this challenging context, vulnerable Refugees endure trauma and loss of assets and populations—including refugees—face unique livelihoods resulting from their flight. Stabilizing barriers to accessing quality work. On average, their livelihoods, improving their economic rural women, urban youth, people with disabilities, opportunities, and placing them on a path of self- and rural-urban migrants are more likely to be reliance can help refugees overcome these conditions inactive and less likely to have improved their and avoid short-term survival strategies that have livelihoods over the last two decades. In this negative long-term consequences, such as putting context, it is unsurprising that refugees cite “lack of children to work, early marriage of children, or selling economic opportunity” as one of their most critical remaining assets (World Bank, 2017). Development challenges. In all domains, when refugees are asked approaches enabling and incentivizing refugees’ to list the top three challenges they face as refugees self-reliance can improve refugee outcomes and in Ethiopia, the most common responses are lack reduce the burden on host communities by reaping of work or business opportunities and high cost of economic benefits from refugees’ presence. living. To refugees, these are much more important than poor services, lack of community networks, or Refugees are forced to suddenly leave their countries insecurity and discrimination. This highlights the and settle in foreign lands without necessarily severity of the labor market challenges refugees face selecting their destination or having favorable in Ethiopia. employment prospects. Compared to economic migrants, refugees often arrive without connections Inclusion, rather than marginalization, can benefit to employers or time to invest in applicable human both refugees and host communities. Defining capital, language, or other skills (Brell et al., 2020). development approaches and better situating Nonetheless, many refugees have valuable skills them within the agenda of international protection and experience to contribute to local economies and national, regional, and local development (World Bank, 2023; Lebow, 2023). Strengthening plans can enable refugees and their hosts to refugees’ human capital during displacement— fulfill their potential. Development approaches that is, strengthening their skills, knowledge, and are most successful when they focus on building experience and the ability to apply them in the host self-reliance—including offering refugees secure country setting—is essential to enable refugees to terms of stay, mobility to access better economic realize their potential, become productive members opportunities, and access to the labor market— of society, and achieve self-reliance. and supporting refugees’ pursuit of economic opportunities while simultaneously supporting Upon arriving in the host country, the first few refugee-hosting communities (Betts et al., 2014; years have an outsized effect on economic Clements et al., 2016; Krause and Schmidt, 2020). opportunities and wages. Enabling refugees’ labor market participation from a very early stage is Figure 3.1: Top 3 difficulties with being a refugee critical to achieving positive long-term integration 0.8 as it limits long-term scarring effects, such as long- 0.7 0.6 term unemployment or inactivity (Fasani et al., 0.5 2022; Slotwinski et al., 2019). Refugee employment 0.4 after arrival depends on policies in the host country 0.3 0.2 concerning work permits and mobility (Fuller, 2015; 0.1 World Bank, 2017). On the other hand, arrival of 0 Addis Refugees Camp Refugees refugees may have a complex range of positive Lack of work/business opportunities Poor services or institutional support High cost of living Lack of freedom or mobility and negative effects on local labor markets in host Lack of community/family networks Insecurity or discrimination communities, including on sectoral employment, Source: World Bank Staff based on SESRE 2023. wages, and prices. Studies in Ethiopia show that 25 Jobs and Livelihoods refugees may decrease employment among hosts While the GoE has made commendable steps in rural areas (Ayenew, 2021), while other studies towards granting refugees the right to work, most find no effect on employment and increases in refugees do not have work permits for wage or self- consumption (von der Goltz, 2023) and product employment. Technically, refugees in Ethiopia have diversification and livestock sales (Walelign et al., a right to participate in the labor market. However, 2022) as refugees increase consumer demand for this has not been implemented in practice due to a agricultural products, with variations in effects lack of clarity on what “most favorable treatment across the different regions of Ethiopia. accorded to foreign nationals” means. Though progress is made by clarifying the legal framework Most refugees in Ethiopia do not move to locations and issuing work permits for different employment with better economic opportunities and live in pathways, such as for joint projects, wage- camps. About 88 percent of refugees in Ethiopia are employment, and self-employment (see Annex B for living in camps, where they cannot take advantage details on pathways of employment), few refugees of economic opportunities to be self-reliant and have obtained work permits or business licenses. thereby improve their economic opportunities to Instead, refugees typically work in the informal reduce their dependence on support from their sector in surrounding communities or the camps., hosts and the international community (World Bank, this may include selling aid rations on the local 2017). Denying refugees mobility to settle where they market, informal trade, and economic exchange, would like comes at a cost, as the choice of location or working for local NGOs and UN agencies within the host country affects refugees’ labor market (ReDSS, 2018). Extensive research has shown that outcomes. Placing refugees in areas of lower economic not being able to enter local labor markets legally opportunity while unable to relocate to better areas is detrimental to refugees’ ability to earn for their makes it hard for them to work (Azlor et al., 2020; Eckert families, and for them to find an occupational et al., 2020; Fasani et al., 2022). Therefore, development match that maximizes the benefits they contribute approaches promoting refugees’ mobility to where to Ethiopia (World Bank, 2023). economic opportunities are highest are most likely to contribute to local economies. Box 3.1: Eritrean refugee sample in the SESRE Because of conflict, the SESRE data collection could not include refugees in the Tigray region in Ethiopia, where most Eritrean refugees were hosted before the conflict. Since outbreak of the conflict in November 2020, many Eritrean refugees moved to Addis Ababa, and many fled from the Mai Ani and Adi Harush refugee camps in Tigray to the newly established refugee hosting site of Alemwach, Dabat in the Amhara region. Between February and July 2022, over 15,000 refugees relocated from the Tigray camps to Alemwach, and an additional 7,000 refugees were resettled in November 2022 following the cessation of hostilities (UNHCR, 2022). Before the conflict, 64 percent of all Eritrean refugees were hosted in camps in Tigray and 36 percent in camps in Afar (UNHCR, 2020b). Refugees going to either Tigray or Afar are distinct culturally and linguistically. Eritrean refugees in Afar are Muslim and speak Afar, as do the Ethiopian hosts in Afar. Many Eritrean refugees in Tigray—and thus the ones who moved to Amhara during the conflict—are Orthodox Christians and speak Tigrinya. The SESRE sample, therefore, includes in-camp Eritrean refugees in two regions: Afar (216 households) and Amhara (216 households). This means that Eritrean refugees in Amhara, representing half of the Eritrean refugee sample, were displaced from Tigray only a few months before the SESRE was implemented, thus have had less time to integrate into the surrounding community and labor market. Among Eritreans in Afar and Amhara, the share working is 43 percent and 9 percent, respectively. Among workers, Eritreans in Amhara are three times as likely to work for NGOs or RRS, and very few work outside the camp. 26 Jobs and Livelihoods Box 3.2: OCP Refugees under the Amnesty Program During the conflict in Ethiopia’s Tigray region, many Eritreans fled Tigray and moved to Addis Ababa without OCP documentation. Given the needs, Ethiopia’s Refugee and Returnee Service (RRS) implemented an “amnesty” program, providing OCP documents to all refugees who came to Addis Ababa after November 2020. Between November 2020 and 2022, approximately 43,000 Eritreans migrated to Addis Ababa. These represent 27 percent of Addis Ababa refugees in the SESRE sample. The Eritreans who migrated after November 2020 are slightly less educated and, given they have had less time to integrate into the Addis Ababa labor market, are less likely to be employed (10 percent relative to 19 percent for Eritreans who came before November 2020). These households are also more likely to rely on remittances as their primary source of income. Throughout this chapter, we will keep Eritreans who arrived to Addis Ababa before and after November 2020 combined for analysis. Ethiopia relies on a camp-based model, with 88 workforce (in the week before the survey), meaning percent of refugees hosted in camps. As outlined, they were employed or available to work and actively in-camp refugees generally do not have work permits searching (strict unemployment). This compares or business licenses and largely depend on work to 52 percent for hosts. This figure increases to 43 inside the camp or informal work outside the camp. percent for refugees and 57 percent for hosts if you On the other hand, the GoE introduced an out-of- include all available to work regardless of whether camp policy (OCP)39 in 2010 that provides refugees they are searching (relaxed unemployment). The the opportunities to live in Addis Ababa and different remaining 57 percent of refugees are inactive, and non-camp locations of their choice. Roughly 71,000 just over half are currently in school, leaving 23 Eritreans were under the OCP regime as of 2022, with percent of refugees neither working nor studying, over 90 percent living in Addis Ababa. In practice, compared to 20 percent of hosts. most of those approved for the OCP have family and friends in Ethiopia who support them with In-camp refugees have high unemployment rates remittances—fewer than 1,500 OCP work permits relative to hosts. Figure 3.2 shows that only 25 were issued by 2022. The permit allows refugees to percent of in-camp refugees performed paid work in freely move and establish residence in all areas of the week before the survey, compared to 48 percent the country except restricted areas. for hosts. At the household level, only 54 percent of refugee households have any workers, relative to OCP refugees are systematically different from 86 percent for hosts. The strict unemployment rate in-camp refugees, as seen in their livelihood is 21 percent for refugees and 7 percent for hosts, strategies. Therefore, the remainder of this chapter while the relaxed unemployment rate is significantly divides the analysis between in-camp refugees and higher at 43 percent for refugees and 15 percent OCP refugees. for hosts. Across camp domains, Eritreans have the highest rate of relaxed unemployment at 55 3.1 Labor market outcomes of in-camp percent, while it is 45 percent and 40 percent for refugees and their hosts Somalis and South Sudanese, respectively. South In-camp refugees have high inactivity rates (not Sudanese have the highest rates of inactive workers working or unemployed) and low labor force remaining in school, reflecting that they have a participation. Table 3.1 shows that only 31 percent of younger population and that more young adults stay all in-camp refugees aged 15-64 “participated” in the in school (mainly primary) after age 15.40 39 For a more detailed description of OCP see Box 2.2. 40 Refer to the Annex D Table D.5 for statistics broken down by survey domains. 27 Jobs and Livelihoods Figure 3.2: Work status Table 3.1: Labor force statistics Camp- Camp- Hosts Refugees Camp Refugees Labor force participation rate (strict) 52% 31% Unemployment rate (strict) 7% 21% Labor force participation rate (relaxed) 57% 43% Unemployment (relaxed) 15% 43% Camp Hosts Employment-to-population ratio 48% 25% Source: World Bank Staff based on SESRE 2023. Note: Labor force participation ratio is the share of working-age people 0 10 20 30 40 50 60 70 80 90 100 who are engaged in the labor market, either employed or unemployed. Percent Unemployment is the share of people participating in the labor force who are not employed. The “relaxed” definition of labor force participation includes anyone who is available to work. The “strict” Employed Unemployed, searching Unemployed, not searching definition of labor force participation includes only those who are Inactive not in school Inactive in school available to work and also actively searching for work. Employment-to- population ratio is the share of working-age people who are employed. Source: World Bank Staff based on SESRE 2023. Working in-camp refugees tend to work in lower- These patterns mask important variations across skill jobs than hosts, and many rely on employment refugee domains. While the share of in-camp with NGOs, international organizations, and RRS. working refugees does not vary by country of In-camp refugees are more likely than surrounding origin, Eritrean refugees are more likely to work in hosts to be self-employed (71 percent) or work in crafts and related trades (49 percent; see Annex D, private households (6 percent), and 15 percent rely Figure D.19). South Sudanese refugees are likelier on work with NGOs, international organizations, to be in elementary occupations (77 percent). As a or RRS.41 Refugees are less likely to be in high- result, Eritreans concentrate more in the industrial skill occupations, which include managerial sector and less in services. Somali refugees stand or professional jobs (based on ISCO-2008 out because they work more in services and sales classifications), and more likely to be in elementary (29 percent) and skilled agricultural (24 percent) occupations, crafts, and services. This largely reflects with higher livestock ownership relative to other the fact that in-camp refugees have lower education domains. Somalis are also more likely to work for than hosts. Notably, refugees are much less likely to private households, including household services, work in agriculture, even though many households construction, and agricultural work. previously relied on agriculture in their home country, reflecting refugees’ lack of access to land. Figure 3.3: Work type Figure 3.4: Occupation 100 100 90 90 80 80 70 Percent 70 60 60 50 Percent 50 40 30 40 20 30 10 20 0 Camp Hosts Camp Refugees 10 0 Elementary occupations Machine operators/assemblers Camp Hosts Camp Refugees Cra /related trade workers Skilled agricultural workers Self-employed Employee Public Service/sales workers Clerical support workers NGO/RRS Private household Unpaid Tech/associate professionals Managers/professionals Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. 41 In Ethiopia, in-camp refugees can work as incentive workers, with standardized pay scales according to their skills, in different organizations, including RRS, a government entity. Thus in-camp refugees who indicated that they work in the public sector were assumed to be incentive workers under RRS. 28 Jobs and Livelihoods Figure 3.5: Work status by gender Figure 3.6: In-camp refugee share employed by age 0.70 Camp Refugees Female 0.60 Camp Hosts Female 0.50 0.40 Camp Refugees Male 0.30 0.20 Camp Hosts Male 0.10 0 20 40 60 80 100 Percent 0.00 Employed Unemployed, searching Unemployed, not searching 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 Inactive not in school Inactive in school Female Male Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. Female refugees have high employment rates labor market outside of camps is a critical element relative to male refugees of all ages and make a of sustainability—both financially and socially—to critical contribution to refugee household incomes. reduce dependence on host government assistance Figure 3.5 shows that, on average, in-camp refugee (World Bank, 2023). Finally, Figure 3.9 shows that, women and men are equally likely to be employed while women are less likely than men to work outside (around 25 percent), while among hosts, men are the camps, they still do so at relatively high rates—35 twice as likely to be employed (62 percent compared percent of working women work outside of camps, to 37 percent for women). Like men, refugee compared to 47 percent of men. women are more likely than host counterparts to be self-employed and less likely to be in high-skill Not only are refugees less likely to work, but those occupations. who do also have lower earnings than hosts. Figure 3.11 shows that, on average, camp-based refugees’ Despite not having work permits, many in-camp hourly earnings are 57 percent lower than hosts’ refugees work outside the camps, which presents hourly earnings, and this gap is higher for women many more income-generating opportunities at 68 percent. For both men and women, the hourly than working in camps. On average across refugee earnings gap is largest in the South Sudanese camps, 40 percent of working refugees work outside domain (77 percent) and lowest in the Somali the camp. This rate is 42 and 44 percent in Somali domain (52 percent). Because refugees work fewer and South Sudanese camps, respectively. It is only hours on average, the average monthly earnings gap 10 percent for Eritrean refugees, but this low number is even larger at 62 percent. is a result of the Amhara camps, where refugees were more recently displaced due to the conflict in Tigray; Lower wages for refugees are not explained in Amhara, the rate is 5 percent relative to 34 percent by differences in education, demographics, in Afar. Out-of-camp work is primarily a mix of occupation, or sector. We can demonstrate this elementary occupations, skilled agricultural work— statistically by using a regression analysis to especially among Somalis, for whom it accounts for compare the earnings of refugees and hosts and how 51 percent of work outside the camp—and, to a lesser this wage gap changes after adjusting for the effects extent, services and sales. As the next section shows, of demographic and job characteristics. Column these workers earn much more than inside the 1, Annex D, Table D.6 shows that, after controlling camp, highlighting the greater income-generating for the domain, monthly earnings is 70 percent opportunities outside the camp. This demonstrates lower for refugees than for hosts. After controlling the importance of allowing refugees to work outside for age, gender, and education in Column 2, this of camps to support their self-reliance; access to the earnings gap remains at 64 percent. This means 29 Jobs and Livelihoods Figure 3.7: Female work type Figure 3.8: Female occupations Camp Refugees Female Camp Refugees Female Camp Hosts Female Camp Hosts Female 0 20 40 60 80 100 Percent Elementary occupations Machine operators/assemblers 0 10 20 30 40 50 60 70 80 90 100 Cra /related trade workers Skilled agricultural workers Percent Service/sales workers Clerical support workers Self-employed Employee Public NGO/RRS Private household Unpaid Tech/associate professionals Managers/professionals Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. that, after adjusting for any differences explained primary are more likely to work if they are older; by age, gender, and education, refugees still earn among hosts who completed primary education, significantly less than hosts. Column 3 indicates the share who are employed rises to 89 percent by that refugees face this same earnings gap even after age 45-54 compared to 69 percent for hosts without adjusting for differences explained by occupation primary. However, for refugees, employment does and industry. Only after restricting the analysis to not depend on education. Similarly, employment refugees who work only outside the camp does this increases dramatically with age for hosts in all wage gap fall to 41 percent, indicating that policies to education groups, but this is less the case for allow refugees to work outside the camp are crucial refugees. A regression analysis (Annex D, Table D.7) for improving refugees’ ability to generate income, confirms these patterns after controlling for country though they still face significant disadvantages in of origin, gender, and years spent in Ethiopia – hosts the labor market even after adjusting for their age, enjoy greater increases in earnings as they age or gender, and education. if they are better educated compared to in-camp refugees. This may partly explain why many refugee Compared to hosts, camp-based refugees’ households are unwilling to send their children to employment depends little on education and school, as highlighted in Chapter 2. increases less with age. Among hosts, those who completed secondary are much more likely to Education and experience are also more associated work across all ages, while people who completed with working in a high-skill occupation for hosts than for refugees, though there is a positive Figure 3.9: Refugee work location relationship between refugee education and working in a high-skill occupation. Regression results (see Annex D, Table D.7) also confirm that, Camp Refugees Female for hosts, older and more educated people are, in addition to having higher monthly earnings, more likely to work in a high-skill occupation (managerial Camp Refugees Male or professional occupations). For refugees, older and more educated people are no more likely to work 0 20 40 Percent 60 80 100 outside the camp. Yet, refugees with higher levels Inside the camp Outside the camp of education are more likely to be in a high-skill Source: World Bank Staff based on SESRE 2023. occupation when they are able to find work. 30 Jobs and Livelihoods Figure 3.10: Hours per week Figure 3.11: Hourly earnings 45 60 40 50 35 30 40 25 30 20 15 20 10 10 5 0 0 Camp Hosts Camp Refugees Camp Hosts Camp Refugees Camp Hosts Camp Refugees Camp Hosts Camp Refugees Male Male Female Female Male Male Female Female Source: World Bank Staff based on SESRE 2023. Note: Hourly earnings are past-month earnings in the main occupation in Birr divided by the typical hours worked in a month over the past year. Only refugees working outside of camps start to see demographic characteristics, there is a slight increase earnings improve with schooling. Working outside in the probability of working for each year that a the camp is associated with a 42 percent increase refugee is in Ethiopia, by around one percentage in earnings. Most importantly, the relationship point per year, but no change in the likelihood of between completing secondary and post-secondary being in a high-skill occupation, likelihood of working schooling and wage earnings becomes significant outside the camp, or in monthly earnings. only when the refugee sample is restricted to workers outside the camps (see Annex D, Table D.7). This Agriculture is an important source of livelihood indicates that refugees only benefit from education for host households, but refugee households have and are incentivized to invest in education if they can low agricultural holdings, reflecting their inability access work outside the camps. This highlights the to own land legally. Refugee households are less importance of refugees’ access to the labor market than half as likely as hosts to report an agricultural without restrictions, particularly outside the camps, holding with crops (19 percent versus 41 percent as a critical component to achieving positive long- of host households, Figure 3.14). Refugee livestock term effects. ownership is similarly low (22 percent of households own livestock versus 48 percent for host households) Refugee outcomes in the labor market do not but average livestock ownership is higher for Somali improve over time in Ethiopia. Annex D, Table D.7 households (41 percent) (Figure 3.15). also shows that, after adjusting for domain and Figure 3.12: Share employed by age – camp refugees Figure 3.13: Share employed by age – camp hosts 1.00 1.00 0.90 0.90 0.80 0.80 0.70 0.70 0.60 0.60 0.50 0.50 0.40 0.40 0.30 0.30 0.20 0.20 0.10 0.10 0.00 0.00 15-24 25-34 35-44 45-54 55-64 15-24 25-34 35-44 45-54 55-64 Less than Primary Primary Secondary Less than Primary Primary Secondary Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. 31 Jobs and Livelihoods Figure 3.14: Household owns crops Figure 3.15: Household owns livestock 0.6 0.6 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 0.1 0.1 0 0 Camp Hosts Camp Refugees Camp Hosts Camp Refugees Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. When refugees own livestock, the value and flock motorcycles, or Bajaj, a cause of a significant portion size of this livestock is low. Somali refugees mostly of the gap in total value of assets between refugees own sheep, goats, and donkeys, and compared and hosts. to Somali hosts, they have smaller flock sizes and lag them in terms of cattle ownership. As Figure With low employment rates, earnings, and value of 3.17 shows, the lower value of livestock in refugee household income-generating activities, in-camp households also reflects lower reported monetary refugee households rely heavily on aid. Expanding value of equivalent livestock (per Tropical Livestock refugee access to agricultural land, livestock, and Unit). For the most part, however, it reflects the legal work outside of camps are vital for refugees lower value of the type and number of livestock to maintain their livelihoods without depending refugees own. on donations. On average, 78 percent of in-camp refugee households report that NGOs or government In Eritrean and Somali camps, refugees and donations are their primary source of income hosts report a similar rate of non-farm business (Figure 3.20), increasing to 88 percent for South ownership; but the value of productive assets Sudanese households. On the other hand, their in refugee businesses is low, indicating they host counterparts rely most on employment, and are primarily small-scale and low-income. The 32 percent rely on agricultural income compared to exception is Somali refugees, partially driven by only 3 percent of refugee households. This contrasts ownership of animal-drawn carts. Across all domains, with refugees’ previous livelihoods in their country refugees have a lower value of productive assets of birth, where they relied on traditional income such as farming tools and construction equipment. sources, especially agriculture and remittances, They are also less likely to own commercial cars, along with a smaller amount of aid. Figure 3.16: Total value of livestock Figure 3.17: Value per tropical livestock unit 60,000 25,000 50,000 20,000 40,000 15,000 30,000 10,000 20,000 5,000 10,000 0 0 Camp Hosts Camp Refugees Camp Hosts Camp Refugees Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. 32 Jobs and Livelihoods Figure 3.18: Household has non-farm business Figure 3.19: Value of productive assets among households with non-farm business 0.25 25,000 0.2 20,000 0.15 15,000 0.1 10,000 0.05 5,000 0 0 Camp Hosts Camp Refugees Camp Hosts Camp Refugees Source: World Bank Staff based on SESRE 2023. Productive assets include the subset of assets with production value, such as farm tools and water pumps, sewing and building equipment, and commercial cars. Outliers are treated, and values are adjusted for inflation. Reliance on aid is even larger for female-headed outside the camp. The regression in Annex D, households, which are larger and have more Table D.8 shows that, after adjusting for household children. While female refugees have comparable demographic characteristics and education, employment rates to men, they are more likely households with at least one member working to work in elementary occupations, more likely outside the camp rely much less on donations as the to work inside the camp, and earn substantially primary source of income—specifically, 8 percentage lower hourly earnings. Female-headed refugee points less overall and 34 and 18 percentage points households are half as likely to rely on salary as less in Eritrean and Somali camps, respectively. their primary source of income (11 versus 22 percent There is little benefit to working outside the camp to for male-headed households) and instead rely reduce aid reliance in South Sudanese households. more on aid and donations (Figure 3.20b). Female- On the other hand, years spent in Ethiopia are only headed households are also larger, on average, and associated with a gradual decrease in dependence on have more children under age 15, highlighting the donations, with only a 1 percentage point reduction importance of creating livelihoods opportunities for for every year spent in the country. these households. The lack of employment outcomes contrasts with As with labor market outcomes, household reliance other refugee-hosting countries in East Africa, on donations improves little over time in Ethiopia highlighting the lack of labor market access for but improves once a household member works Ethiopian refugees. Evidence suggests that refugees Figure 3.20: Household primary income source a. Pre-post migration b. By gender of head 100 80 Camps-Female-Headed 60 Percent 40 Camps-Male-Headed 20 0 0 20 40 60 80 100 Camp Hosts Camp Refugees Current Camp Refugees COB Percent Salary (employment/casual labor) Crops/livestock Donations (NGO/gov) Salary (employment/casual labor) Crops/livestock Donations(NGO/gov) Remittances (local/international) Other (rental income, PSNP, pension) Remittances (local/international) Other (rental income, PSNP, pension) Source: World Bank Staff based on SESRE 2023. Note: “COB” refers to livelihood strategies in their country of birth. 33 Jobs and Livelihoods Box 3.3: Refugee Vocational Training and Cooperatives Refugees across all regions of Ethiopia have benefited from livelihood training interventions provided by RRS, domestic and international NGOs, and humanitarian organizations. For example, the Ikea Foundation, through UNHCR, invested around US$100 million in the Dollo Ado camps in Somalia between 2012 and 2019. Much of this funding supported economic development and livelihood opportunities for refugees and the host community, including creating livelihood cooperatives in agriculture, livestock value chain, energy, firewood, and microfinance (Betts et al., 2020). In the north of the country, UNHCR worked with various partners to provide Eritrean refugees with vocational skills training, tools, and start-up capital for crafts, such as leather products, weaving, and tailoring. UNHCR records indicate that more than 8,000 Eritrean refugees received such training, which could explain the high share of Eritreans working in crafts and related trades in the SESRE. As another example, the German Agency for International Cooperation (GIZ) has a multi-million-dollar program to improve quality and access to vocational training for refugees and Ethiopians across all refugee-hosting regions (Giordano et al., 2021). World Bank programs include the Economic Opportunities Program (EOP) and the Urban Safety Net and Jobs Project (UPSNJP), which provide economic opportunities for Ethiopians and refugees through various social protection and labor market interventions like public works employment and job search assistance. How effective have these livelihoods and vocational training programs for refugees been in Ethiopia? This question is difficult to answer due to lack of consolidated information on programs, the number of beneficiaries, and the economic outcomes of beneficiaries. Case-study evaluations indicate that vocational training programs have helped to increase incomes, diversify income sources, and, in some cases, promote local infrastructure development. However, these income increases are often modest, and trained individuals and cooperatives often fail to become self-sustainable in the long term and continue to rely on external inputs, especially in more remote camps with poor market linkages (Betts et al., 2020; Giordano et al., 2021; Holzaepfel, 2015). This is consistent with the results of the SESRE finding that, despite the scale of livelihood training investments across the country, most working-age refugees still do not work, and most refugee households rely on aid as a primary source of income. Given the protracted nature of refugee hosting in Ethiopia, it is essential to better understand how vocational training programs and cooperatives can become self-sustainable, especially in the more remote border regions with poor market linkages. arrive in Uganda with few assets, in a state of all of the remaining 81 percent of refugees in Addis high poverty, and with similarly low employment Ababa rely on remittances. As in the camp domains, rates. However, unlike Ethiopia, employment remittance reliance is even higher for female- rates for refugees in Uganda improve over time, headed households (84 instead of 72 percent for approximately doubling after five years or more male-headed households). The refugee households (World Bank, 2023b). Uganda is also notable for in Addis Ababa are also used to relying on providing work rights to refugees in practice (Ginn et remittances; 45 percent relied on remittances even al., 2022) (Box 3.3). before migrating to Ethiopia, and only 12 percent relied on crops or livestock. This contrasts with the 3.2 Labor market outcomes of OCP refugees Eritreans in camps, who previously relied primarily and their hosts on agricultural and labor income, and demonstrates Refugee households in Addis Ababa rely heavily the large differences between Eritrean households on remittances as their primary source of income. that could and could not acquire OCP status. This, This reflects the fact that the Eritrean OCP refugees coupled with the finding that OCP refugees have who move to Addis Ababa must provide proof of much higher levels of education compared to in- formal employment or guarantor to support them. camp Eritrean refugees (as highlighted in Chapter Figure 3.21a shows that only 19 percent of refugee 2), again indicates that OCP refugees were relatively households in Addis Ababa rely on employment well-off before displacement, and they still have income, compared to 79 percent of hosts. Almost family members or other support systems. 34 Jobs and Livelihoods Figure 3.21: Household primary income source a. Pre-post migration b. By gender of head Addis Refugees COB Addis Female-Headed Addis Refugees Current Addis Male-Headed Addis Hosts 0 20 40 60 80 100 0 20 40 60 80 100 Percent Percent Salary (employment/casual labor) Crops/livestock Donations(NGO/gov) Remittances (local/international) Other (rental income, PSNP, pension) Source: World Bank Staff based on SESRE 2023. Note: “COB” refers to livelihood strategies in their country of birth. In line with high reliance on remittances, labor Among the 17 percent of working-age OCP force participation and employment rates among refugees who are working, many have completed OCP refugees are low, with rates similar for male secondary schooling; yet, they face considerable and female refugees. Table 3.2 presents labor force occupational downgrading relative to hosts, and participation (LFP) data; among those aged 15-64, this is especially severe for female refugees. While LFP is 46 percent for refugees and 66 percent for refugees in Addis Ababa are less likely than hosts to hosts. This increases to 67 and 72 percent when have a post-secondary degree (4 percent versus 20 you use the relaxed definition of unemployment, percent for hosts), many of them have completed reflecting the large number of OCP refugees who secondary (25 percent versus 21 percent for hosts). say they are ready to work but are not actively Yet, Figure 3.24 shows that refugees are less likely searching. The unemployment rate is an astonishing than hosts to be in a high-skill occupation, which 63 percent for refugees relative to 12 percent for includes managers, professionals, and technical hosts, and it increases to 75 percent under the and associate professionals. This gap increases relaxed definition. Only 17 percent of OCP refugees further when restricted to workers with completed work, and 23 percent are inactive and not in school. secondary education. Among workers with Female refugees have similarly large rates of completed secondary, 30 percent of refugees and inactivity and unemployment, and the total share 50 percent of hosts are in high-skill occupations working is similar to male refugees. Female refugees (Figure 3.25). Among only women, these numbers also have high rates of self-employment relative to are 20 percent and 50 percent, respectively. Instead, female hosts. refugee men and women in Addis Ababa are over- represented in crafts and related trades (typically classified as medium-skill occupations). Table 3.2: Labor force statistics Addis Hosts Addis Hosts Labor force participation rate (strict) 66% 46% Unemployment rate (strict) 12% 63% Labor force participation rate (relaxed) 72% 67% Unemployment (relaxed) 19% 75% Employment-to-population ratio 58% 17% Source: World Bank Staff based on SESRE 2023. Note: Labor force participation ratio is the share of working-age people who are engaged in the labor market, either employed or unemployed. Unemployment is the share of people participating in the labor force who are not employed. The “relaxed” definition of labor force participation includes anyone who is available to work. The “strict” definition of labor force participation includes only those who are available to work and also actively searching for work. Employment-to-population ratio is the share of working-age people who are employed. 35 Jobs and Livelihoods Figure 3.22: Work status by gender Figure 3.23: Work type by gender 100 - Addis Refugees Female 80 Addis Hosts Female 60 Percent Addis Refugees Male 40 Addis Hosts Male 20 0 20 40 60 80 100 0 Percent Addis Hosts Addis Refugees Addis Hosts Addis Refugees Male Male Female Female Employed Unemployed, searching Unemployed, not searching Inactive not in school Inactive in school Self-employed Employee Public NGO/RRS Private household Unpaid Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. Occupational downgrading among the small concentration of refugees if they were in the same share of OCP refugees who work is not explained occupations as hosts with their same demographic by education, gender, and age. Another way to characteristics. If the actual share of refugees visualize the scale of occupational downgrading in an occupation is different from this predicted among refugees in Addis Ababa is to calculate how share, then it is due to other factors not related to they should be distributed across occupations if education, gender, and age. The results in Figure they were in the same occupations as hosts within 3.26 show that refugees are under-represented in their education, gender, and age group. For example, high-skill occupations and services and sales relative suppose that among male hosts under age 25 with to what we would expect based on their age, gender, primary education, 70 percent work in elementary and education, and substantially over-represented occupations and 30 percent in crafts and related in crafts and related trades. trades. Imagine taking all male refugees under age 25 with primary education and assigning 70 percent Among employee workers, OCP refugees also earn of them to elementary occupations and 30 percent less than hosts, though the wage gap is smaller to crafts and related trades – to represent their than for in-camp refugees. Annex D, Table D.9 shows occupational concentration if they worked the same that refugees in Addis earn 25 percent less than occupations as hosts – and repeating this for all their hosts. Even after adjusting for demographic demographic groups. The final share of refugees in characteristics, occupation, and sector of work, this each occupation now represents the occupational wage gap remains at around 19 percent. Figure 3.24: Occupation Figure 3.25: Occupation among completed secondary or more 100 100 90 90 80 80 70 70 60 60 Percent Percent 50 50 40 40 30 30 20 20 10 10 0 0 Addis Hosts Addis Refugees Addis Hosts Addis Refugees Addis Hosts Addis Refugees Addis Hosts Addis Refugees Male Male Female Female Male Male Female Female Elementary occupations Machine operators/assemblers Cra /related trade workers Elementary occupations Machine operators/assemblers Cra /related trade workers Skilled agricultural workers Service/sales workers Clerical support workers Skilled agricultural workers Service/sales workers Clerical support workers Tech/associate professionals Managers/professionals Tech/associate professionals Managers/professionals Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. 36 Jobs and Livelihoods Figure 3.26: Refugee occupation concentration 0.35 0.3 0.25 Share of Refugees 0.2 0.15 0.1 0.05 0 Elementary Occupations Cra /Related Trades Service/Sales Machine Operators Associate Prof. Clerical Support Professionals Observed Predicted according to observables Source: World Bank Staff based on SESRE 2023. These results indicate that the OCP model is Instead of keeping refugees in camps, it is vital to not ideal for refugees’ labor market inclusion. allow them to realize their potential and thus benefit Few refugees primarily enroll through an existing host communities as productive members of society. formal employer; thus, the OCP is mainly open to refugees with networks that can support them with 3.3 Refugee youth remittances, and this makes these households less Refugee inactivity and unemployment are high likely to work in Addis Ababa. Therefore, refugee among youth (aged 15-24) in Addis Ababa. Table 3.3 labor market outcomes in Addis Ababa would be shows that, in Addis Ababa, the youth participation very different if all refugees had the possibility rate is 60 percent for refugees, just higher than the to move in response to economic opportunities. 54 percent rate for hosts. However, this falls to 37 In fact, households without an existing support percent for refugees and 48 percent for hosts under system are precisely those who can benefit most the “strict” definition, reflecting the large number from access to labor markets and will contribute of refugee youth who are available to work but not the most economically. However, the occupational actively searching. Of the 60 percent of refugee downgrading and lower wages among Ethiopia’s youth who are available to work, 82 percent are OCP refugees who work show that challenges persist unemployed, compared to 21 percent for hosts. As a even after refugees are granted access to urban labor result, only 11 percent of refugee youth work relative markets. This has been well-documented in many to 43 percent of hosts. Also, while the inactivity rates other settings around the world (Lebow, 2023; World are broadly similar, refugee youth in Addis are half as Bank, 2023). likely to be in school (21 percent relative to 40 percent among hosts). The lower schooling among refugees Solutions that integrate refugees and host in Addis Ababa only emerges after age 18, indicating a communities throughout the displacement cycle lower probability of attending post-secondary when have proven most promising for achieving good they instead enter inactivity or unemployment. Youth development outcomes (World Bank, 2023). participation and employment rates are also lower This requires safeguarding refugees from harm for refugees than hosts in the camp domains, though while integrating them as workers, students, and this difference is starker in Addis Ababa. The share neighbors. Despite Ethiopia’s goodwill towards who are not in employment, education, or training refugees, and the global recognition that responses (NEET) in Addis Ababa is 19 percent for refugees and to forced displacement need humanitarian and 6 percent for hosts, and in the camps is 13 percent for development responses, too few efforts exist to refugees and 12 percent for hosts. better integrate refugees within host communities. 37 Jobs and Livelihoods Table 3.3: Youth labor force statistics (age 15-24) Camp Hosts Camp Refugees Addis Hosts Addis Refugees Participation (strict) 25% 12% 48% 37% Unemployment (strict) 15% 25% 12% 71% Participation (relaxed) 30% 22% 54% 60% Unemployment (relaxed) 30% 57% 21% 82% Employment-to-population ratio 21% 9% 43% 11% Source: World Bank Staff based on SESRE 2023. Note: Labor force participation ratio is the share of working-age people who are engaged in the labor market, either employed or unemployed. Unemployment is the share of people participating in the labor force who are not employed. The “relaxed” definition of labor force participation includes anyone who is available to work. The “strict” definition of labor force participation includes only those who are available to work and also actively searching for work. Employment-to-population ratio is the share of working-age people who are employed. Figure 3.27: Youth work status NEET (23 percent for girls relative to 15 percent for boys—Figure 3.28). In camps, refugee girls look more Camp Refugees like their host counterparts regarding schooling and labor force participation. However, among those in Camp Hosts the workforce, their relaxed unemployment rate is Addis Refugees much higher (55 percent for female refugees and Addis Hosts 36 percent for female hosts). Boys in camps have 0 20 40 60 80 100 higher schooling rates (75 percent relative to 65 Percent percent for hosts), reflecting their higher propensity Employed Unemployed, searching Unemployed, not searching to stay enrolled in primary or secondary schooling Inactive not in school Inactive in school after the typical completion age, especially in South Source: World Bank Staff based on SESRE 2023. Sudanese camps where 84 percent of boys aged 15- In Addis Ababa, refugee girls—like boys—are 24 are in school. The relaxed unemployment rate is much less likely than host counterparts to work high among boys in the workforce, as it is for girls or attend school, and they are more likely to be (60 percent for male refugees and 24 percent for unemployed or NEET. They are even less likely than male hosts). refugee boys to be in school and more likely to be Figure 3.28: Male youth work status Figure 3.29: Female youth work status Addis Refugees Male Addis Refugees Female Addis Hosts Male Addis Hosts Female Camp Refugees Male Camp Refugees Female Camp Hosts Male Camp Hosts Female 0 20 40 60 80 100 0 20 40 60 80 100 Percent Percent Employed Unemployed, searching Unemployed, not searching Employed Unemployed, searching Unemployed, not searching Inactive not in school Inactive in school Inactive not in school Inactive in school Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. 38 Jobs and Livelihoods 4. Refugees’ Aspirations T his chapter looks at how low sociodemographic and labor market outcomes shape how refugees perceive their future prospects and aspirations. pathways of refugee admission to third countries outside of UNHCR processes—was below 2 percent over the past twenty years (World Bank, 2023). According to government statistics, there has also While “resettlement” to a high-income country been a downward global trend in the number of is an attractive solution, the share of refugees resettlement opportunities, fluctuating from 99,000 resettled globally is marginal. Resettlement is in 2010 to just 34,000 in 2020, even as the number considered one of the three “durable solutions” of forcibly displaced persons increases globally. In for refugee protection under the 1951 Refugee Ethiopia, resettlement numbers are similarly low; in Convention, alongside naturalization and return. Yet, 2022, only 309 refugees departed for resettlement the share of refugees resettled globally—including (UNHCR, 2022). private sponsorship and other complementary Figure 4.1: Desired location in three years Figure 4.2: Expected location in three years 100 100 90 90 80 80 70 70 60 Percent 60 Percent 50 50 40 40 30 30 20 20 10 10 0 0 Eritrean Somali South Addis All Eritrean Somali South Addis All (camps) Sudanese Ababa Refugees (camps) Sudanese Ababa Refugees Ethiopian refugee camp Ethiopian city Country of birth Ethiopian refugee camp Ethiopian city Country of birth Other African country Western country Other African country Western country Source: World Bank Staff based on SESRE 2023. Note: Household respondents’ responses to the questions “Where do you hope to be living in 3 years?” and “Realistically, where do you think you will be living in 3 years?” 39 Refugees’ Aspirations Most refugees hope to go to a Western country points for refugees over age 45 (Annex D, Table D.10).42 in the next three years. When asked in the SESRE However, it does not depend on gender. Only in where they would like to live in three years, most Eritrean camps is intention to migrate abroad lower, refugees say they would like to live in a Western by 5 percentage points, among those with completed country. This rate is highest among OCP refugees (90 secondary education. Intention to migrate abroad percent) and Eritreans in camps (83 percent), lower also does not vary with time in Ethiopia. in Somali camps (66 percent), and lowest in South Sudanese camps (29 percent). More Somalis and Refugee aspirations and expectations for South Sudanese hope to stay in Ethiopian refugee resettlement may be important determinants camps than Eritreans. South Sudanese refugees of how much they will invest in their skills and stand out in that almost 20 percent hope to return socio-economic integration. Evidence from various to their country of birth in the next three years, while settings shows that when migrants expect to spend this rate is meager for other groups. more time in a country, they invest more in their skills and socio-economic and labor market integration Despite the low probability of being resettled, (Adda et al., 2022). For example, this could include refugees hold an unrealistically high belief that investing in language proficiency, starting a business, they will migrate to a Western country in the acquiring legal work documents, or studying for an next three years. To distinguish between desires occupational license. It could also mean building and expectations of reality, in addition to asking social networks in Ethiopia, which needs to improve households where they hope to live in three years, considering, for example, the tiny share of refugees the SESRE also asks where they “realistically” think who report having an Ethiopian friend (as will be they will be living in three years. One-third of all discussed in Chapter 7). refugees indicated they realistically believe they will be resettled to a Western country, which is in stark Better alignment of refugees’ expectations contrast to the low resettlement numbers worldwide for resettlement with reality could improve and in Ethiopia. As with aspirations, the share socio-economic outcomes. Based on evidence who expect to be in a Western country is highest of resettlement over the past ten years, better- among OCP refugees (57 percent), similarly high for aligning expectations with reality could be essential Eritreans in camps (53 percent), lower for Somalis to encourage refugees to make more significant (42 percent), and lowest for South Sudanese (18 investment in skills and socio-economic integration. percent). While these numbers may be higher than Humanitarian organizations have long understood true beliefs if they reflect a response bias, they are the importance of managing resettlement so high that they strongly indicate an over-optimism expectations (UNHCR, 2023b). Better understanding about relocation. Most of those who do not believe the reasons for unrealistic expectations in Ethiopia, they will be in a Western country believe they will and the role that policymakers and the international remain where they are. Many South Sudanese (13 community play in this, could support refugees’ percent) also think they will return to South Sudan. long-term trajectory. While the intention to migrate abroad is lower for “Locus of control” (LOC)—a feeling of personal older refugees, on average, it does not differ widely control over events in one’s life—is significantly depending on gender or education, and persists related to willingness to invest in one’s future across most subgroups, including youth. Intention and has been shown to improve refugees’ socio- to migrate abroad is lower by around 8 percentage economic integration. LOC is a psychological concept 42 This analysis is based on an individual-level question regarding intention to migrate abroad, which has rates similar to the question to the household head discussed above. 40 Refugees’ Aspirations indicating the degree to which people believe that lives and destinies, and this difference is statistically they, as opposed to external forces, have control over significant. This difference, however, is driven by the outcomes of events in their lives (Rotter, 1966). South Sudanese refugees. When comparing LOC by For example, if a harvest is good or bad, a farmer with country of origin, we find that there is only for South low LOC is likelier to attribute it to chance or external Sudanese there is a significant difference between forces than their skill. Higher LOC has been shown refugees and hosts on perception of control over to affect schooling decisions, occupational choice, their lives. LOC for South Sudanese refugees is 0.15 and savings (Cobb-Clark et al., 2016; Heckman et points (0.33 standard deviations) lower than that of al., 2006). In Ethiopia, higher LOC has been shown hosts, highlighting that they feel that they have less to predict farmer adoption of modern agricultural personal control over their lives and destinies than technologies (Taffesse and Tadesse, 2017). In refugee their hosts. populations, low LOC also correlate with depression, anxiety, and psychological distress (Schlechter et al., When considering different dimensions, we find 2023; Tsionis et al., 2022). Higher LOC has also been that refugees’ LOC is driven by a feeling of lower found to improve employment and socio-economic internal control over the future. LOC can be grouped integration among immigrants and refugees in into three categories: a sense of internal control, Germany (Hahn et al., 2019; Thum, 2014). the role of chance or fate, and the role of “powerful others” (Levenson, 1981). Lower LOC among refugees Compared to hosts, South Sudanese refugees is driven mainly by a feeling of lower internal control perceive less personal control over their lives over fate, as opposed to a greater sense of chance and destinies. Based on SESRE data, the index or the role of other individuals (though in South used to construct a measure of personal control Sudanese camps, the role of chance and powerful over one’s life is an unweighted average of 10 LOC- others is more important). LOC increases with higher related questions. The index (Likert scale) ranges levels of education for both hosts and refugees. Still, from 1—”little control over one’s life”—to 4—”more it has little relationship with age or gender, or how control over one’s life”. The index is 0.12 points (.25 much time refugees have spent in Ethiopia, nor their standard deviations) lower for refugees than hosts aspirations to go abroad. indicating they feel they have lower control over their Figure 4.3: Locus of control Figure 4.4: Locus of control by type of control 2.3 3 2.25 2.5 2.2 2 2.15 2.1 1.5 2.05 1 2 0.5 1.95 1.9 0 Eritrean Somali South Sudanese Addis All Internal Control Chance or Fate Powerful Others Hosts Refugees All Hosts All Refugees Source: World Bank staff based on SESRE 2023. Note: This index is the unweighted average of 10 questions about feelings of control over one’s fate. The index ranges from 0 to 4, where more positive indicates greater control. The internal control index uses four questions regarding personal control over destiny. The chance index uses five questions regarding the role of chance or determinism. The role of the powerful others index is 1 question on whether others determine fate. 41 Refugees’ Aspirations 5. Welfare and Equity T his chapter examines the state of welfare and poverty levels of refugees and host communities in Ethiopia. The emphasis on to specific potential policies to help refugees and their hosts. refugees and host communities acknowledges both 5.1 Welfare dimensions groups’ mutual—and sometimes interdependent— 5.1.1 Monetary poverty and inequality development needs. In-camp refugees have lower welfare outcomes than their hosts. In-camp refugees have significantly We assess multiple dimensions of welfare and higher monetary poverty based on strikingly low poverty of refugees and hosts in Ethiopia using average expenditures. A staggering 75 percent household-level consumption data. The data of refugees live below the international poverty presents a comprehensive set of social and line of US$2.15 in 2017 Purchasing Power Parity economic indicators to determine poverty incidence, (PPP) per day per capita. Though still high, host food security, and standard of living. In addition communities have a relatively lower poverty rate to refugees overall, we look at welfare differences across refugee groups—Eritreans, Somalis, and of 25 percent (Figure 5.1). Considering in-camp South Sudanese—and compare differences in refugees and their hosts only, we find higher contexts and situations. In more detail, we analyze poverty incidence; roughly 84 percent of in-camp (i) poverty incidence and inequality, (ii) expenditure refugees and 32 percent of hosts live in poverty. patterns, (iii) multidimensional poverty, and (iv) Although poverty incidence is higher for refugees, food security, perception of standard of living, and the high poverty rates among host communities shocks. The chapter also provides a poverty profile also imply that they live in severely resource- and determinants of the welfare of refugees and constrained conditions. This calls for development host community households and estimate of the approaches that invest in refugee-hosting areas cost of basic needs for refugees. Insights on poverty in a manner that benefits both refugees and drivers and living conditions contribute to deeper hosts alike (Annex D, Table D.11 presents detailed understanding of displacement dynamics and point poverty rates by refugee domains). 42 Refugees’ Aspirations Figure 5.1: Poverty incidence Figure 5.2: Income inequality, Gini index 45 40 84% Poverty headcount rate (%) 35 75% 30 25 20 32% 15 25% 18% 10 5 7% 0 In camp Addis Ababa Total In Camp Addis Ababa Total Hosts Refugees Hosts Refugees Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. Refugees in Addis Ababa are less poor than their in rented houses. The data show that Addis Ababa hosts, as well as in-camp refugees and their hosts. refugees pay higher rents (ETB 31,600 per year, per Poverty incidence in Addis Ababa for refugees living adult equivalent) than hosts (ETB 18,700 per year, per under the OCP is lower (7 percent) than their hosts adult equivalent). Moreover, rent expenditures make (18 percent).43 This difference is driven primarily up 56 percent of refugees’ non-food expenditure. by the high rent expenditures of refugees since they cannot benefit from public housing schemes, In-camp refugees in Ethiopia are much poorer increasing their overall consumption expenditure. As than their hosts, but because everyone suffers discussed in Chapter 2, about 97 percent of refugee from similarly low expenditures, inequality is and 39 percent of host households in Addis Ababa live also low for refugees. As measured by the Gini Box 5.1: Consumption aggregation and poverty measurement Most of the analysis presented in this chapter is based on detailed consumption data from the Socioeconomic Survey of Refugees in Ethiopia (SESRE) conducted between October 2022 and February 2023. All consumption of food and non-food items is included, regardless of whether these items are purchased on the market, come from own production, or received as gifts. For own-consumption and gifts, the quantities consumed are valued at prevailing prices in the enumeration area. Although consumption is expressed annually, the reference period used during data collection varies based on the nature of the items. For example, questions related to information on food and food-related items was asked by visiting households twice a week using the “last three days” and “last four days” as reference periods. For house rent, durable goods, clothing, health and education expenditures, and some other categories, the survey questions used the “last three months” and “last 12 months” as references. Imputed rent for owner-occupied houses is calculated by the Ethiopian Statistical Service (ESS) team and is included in the consumption expenditure data shared with the Bank team. Spatial and temporal price deflators adjust for price variations across time and space. First, nominal consumption is adjusted for price differences across survey domains using spatial deflators calculated using the Household Welfare Statistics (HoWStat 2021) survey data. Second, spatially-deflated consumption levels are expressed in December 2022 prices using the food and non-food Consumer Price Indexes produced and provided by the ESS. Finally, to adjust for variations in household size and composition, the spatially and temporally adjusted consumption expenditure is divided by household size. This is because the poverty rates presented in this chapter are calculated using the international poverty line of USD 2.15 per capita in 2017 PPP. The US$2.15 poverty line was converted to local currency in 2017 using the PPP conversion factor, and then the value was inflated to December 2022 prices using the national CPI. Given that international poverty estimates reported at the global level are based on consumption aggregates not spatially deflated, the poverty reports presented in this report are not strictly comparable to global poverty rates. 43 For details on the OCP policy, see Box 2.2. 43 Refugees’ Aspirations index, income inequality averages 28.7 for in-camp Stark differences in food and non-food expenditures refugees and 32.7 for their hosts (Figure 5.2). While exist between refugees and hosts, with in-camp welfare is generally unevenly distributed in Ethiopia, refugees receiving most of their food and non- inequality tends to be lower among refugees than food expenditures as transfers. Refugees rely on aid. hosts, except in Addis Ababa. Yet, when looking at While 83 percent of the in-camp refugees depend on the whole sample (in-camp and OCP), inequality is transfers and gifts to cover their food consumption very high among refugees (39.2), much higher than needs, more than two-thirds of the host community their hosts (33.2). This result is driven by the stark households depend on market purchases for their food welfare disparity between in-camp and OCP refugees consumption (Figure 5.5a). Similarly, most refugees (Figure 5.3). Moreover, differences in employment depend on transfers or gifts for non-food consumption, opportunities, and mobility create an uneven playing while their hosts depend on market purchases (Figure field for in-camp refugees and OCP (see Chapter 3). 5.5b). A large share of refugees in Addis Ababa also rely on transfers or gifts driven by remittances. 5.1.2 Expenditure patterns Average consumption expenditures for in-camp Expenditure patterns vary by refugee domain and refugee households is nearly half of hosts. Average in-camp and OCP refugees. Analysis of expenditure annual expenditure per capita is around 45,600 Birr patterns helps to understand differences in dietary for host households and 27,700 for refugees (Figure preferences that affect food poverty and well-being. 5.3). The average food and non-food expenditure With increasing income, more affluent households (such as utilities and supplies, clothing and footwear, are more likely to spend a greater share of their and rent) for refugees is more than half that of their budget on high-value food items such as animal- hosts, except for refugees in Addis Ababa, where the origin diets, processed food, and food away from average expenditure of OCP refugees is considerably home, as well as on non-food items. Except for Addis higher than that of hosts. The strikingly low Ababa, food consumption patterns, as indicated by average expenditure for refugees could be related expenditure shares, differ by food groups (Figure to measurement errors (See Box 5.2 for additional 5.6). Overall, refugees expenditures are higher information). Food expenditure shares are higher on cereals and less on animal-origin food items for refugees, indicating a high dependence on food associated with the types of food aid provided. This associated with higher poverty. Except for Addis could be because refugees receive assistance for Ababa, the share of expenditures on food is slightly cereals/grains, not animal-origin food items. Food higher for refugees than hosts; about 68 percent of all away from home is lower for refugees than hosts, expenditures of in-camp refugees are spent on food, except in Addis Ababa. while hosts spend 61 percent on food, consistent with lower poverty rates in host communities (Figure 5.4). Figure 5.3: Expenditure components (in birr) Figure 5.4: Shares of food expenditure 70,000 68% 60,000 65% 56% 50,000 40,000 30,000 20,000 61% 58% 59% 10,000 - Hosts Refugees Hosts Refugees Hosts Refugees In Camp Addis Ababa Total In Camp Addis Ababa Total Food Non-food non-durables Durables Rent Hosts Refugees Source: World Bank staff based on SESRE 2023. Note: The expenditures are in December 2022 values. 44 Refugees’ Aspirations Box 5.2: Disparity between refugee ration aid and reported consumption quantities Expenditures for in-camp refugees is almost half that of hosts, despite the sizeable food aid and cash transfers (in selected camps) the WFP and UNHCR provide. The significantly lower expenditures (food and non-food) among refugees compared to host populations led to higher poverty rates. The team cross-checked the food aid received by in-camp refugees based on administrative data from the UNHCR and WFP and food consumption data from SESRE. The information received from UNHCR on food aid provided to refugees in each camp includes quantities per food item per month and cash transfers per person per month for each camp and period. The food items include cereal, wheat, maize, rice, sorghum, CSB/famex (CSB+), pulse, biscuit, date biscuit, dates, oil, vegetable oil, salt, and cash (see Annex E for detailed information). Food aid information received from WFP includes five food items and their quantities distributed to refugees: cereal (mainly wheat but in some camps rice), pulses (mostly yellow split peas), CSB+, vegetable oil, and salt. We have computed the per person, per month in-kind aid quantities into annual values using the same prices as other food items based on SESRE data, mapping them to the closest food item in SESRE (this was not straightforward as the items are different). We further considered the changes in quantities of food rations that took place across survey months due to funding shortages, which can significantly affect the overall wellbeing of refugees in Ethiopia. Based on this information, we compare items refugees should have received with what refugees reported regarding food consumption. The results show that refugees reported quantities lower than UNHCR food aid admin data for every item except biscuits. Refugee households still report lower quantities, even when valuing the food ration quantities indicated as sold in markets. Possible explanations for lower food quantities are that food rations are only received once a month, which may not coincide with the interview date. Moreover, SESRE asks what food people consumed (not based on a pre-set list of food items), not food received as aid. Refugees may sell more than indicated. Valuing quantities of food aid with prices from SESRE suggests that if UNHCR food aid quantities were received/reported by refugees, refugees’ food expenditures would be much more comparable to those of hosts. Using WFP food aid information, we found a picture similar to UNHCR’s. Quantities consumed in SESRE are lower than food aid, as reported by WFP, except for CSB+ and salt (See Annex E for details of the disparity in food aid between the admin data disparity SESRE report). Figure 5.5: Food and non-food expenditures shares by sources a. Food expenditure b. Non-food expenditure 100 100 90 90 80 80 70 70 60 60 Percent Percent 50 50 40 40 30 30 20 20 10 10 0 0 Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees In Camp Addis Ababa Total In Camp Addis Ababa Total Own production Market purchase Transfers/gi s Own production Market purchase Transfers/gi s Source: World Bank Staff based on SESRE 2023. 45 Refugees’ Aspirations Figure 5.6: Food expenditure shares by food groups 100 90 80 70 60 Percent 50 40 30 20 10 0 Hosts Refugees Hosts Refugees Hosts Refugees In Camp Addis Ababa Total Cereals Other grains (pulses, oilseeds) Fruits, root crops, and vegetables Animal origin food Other food (spices, fats & oils, beverages) Food away from home Source: World Bank Staff based on SESRE 2023. 5.1.3 Multidimensional poverty for in-camp refugees. About 56 percent of in-camp Refugees are more vulnerable to multidimensional refugees and 24 percent of their hosts are both poverty than hosts. The multidimensional poverty monetarily and multidimensionally poor. The rate is relatively high among refugees, driven picture differs for OCP refugees, less than 2 percent primarily by low living standards and poor access are poor in both monetary and non-monetary to education. Trends in monetary poverty are dimensions. The percentage of households who are mirrored using the Multidimensional Poverty Index multidimensionally poor but not monetarily poor (MPI)—an index measuring deprivations across stands at 10 percent for refugees living in camps and three dimensions of well-being: education, health, 32 percent for the communities hosting them. and standard of living (see Box 5.3). MPI provides a general picture of the extent of deprivation (Alkire et Low living standards and low educational al., 2021). The results show that 50 percent of refugees attainment drive deprivation for all refugee groups. and 23 percent of hosts are multidimensionally poor Low living standards due to low-quality cooking fuel, (Figure 5.7). Looking at in-camp refugees and their poor housing, and low asset holdings contribute hosts, multidimensional poverty is 64 percent for more than 50 percent to non-monetary poverty, refugees and 50 percent for hosts. Unlike monetary followed by education. For in-camp refugees, poverty, multidimensional poverty as measured the contribution of education, health, and living here appears to be relatively lower for refugees. This standards to overall non-monetary poverty is 30, 18, reflects improvement and ease of providing public and 52 percent, respectively. Few years of schooling services in high-density areas with high-refugee and child malnutrition are the dimensions that concentrations. There is a considerable correlation contribute most to poverty (Figure 5.7). However, between monetary and multidimensional poverty child mortality and access to improved water Figure 5.7: Multidimensional poverty incidence, severity, and vulnerability a. Poverty incidence b. Contributions by dimensions 70 100 90 60 80 50 70 60 Percent 40 50 40 30 30 20 20 10 10 0 Hosts Refugees Hosts Refugees Hosts Refugees - Hosts Refugees Hosts Refugees Hosts Refugees In Camp Addis Ababa Total In Camp Addis Ababa Total Years of education School attendance Child mortality Nutrition Headcount Vulnerable to poverty Severe poverty Electricity Sanitation Water Housing Cooking fuel Assets Source: World Bank Staff based on SESRE 2023. 46 Refugees’ Aspirations Box 5.3: MPI methodology Refugee and host communities could differ in multiple dimensions over and above consumption. The Multidimensional Poverty Index (MPI) explores this multiple deprivation, capturing differences across three dimensions of well-being: health, education, and living standards (Alkire et al., 2021). MPI provides a general picture of the extent of deprivation. In this context, deprivation in education is assessed using school attendance for school-age children and years of schooling among adults. Health is proxied by the presence in the household of a stunted child or death of a child in the last 12 months before the survey. Living standards are assessed by access to electricity, improved water, sanitation, cooking fuel source, housing, and economic assets. The MPI ranges from 0 to 1, with 1 representing a high level of deprivation. It is the product of two partial indices: the headcount ratio (H) and the intensity of poverty (A) i.e. (MPI = H*A). The headcount ratio is the share of poor people in the population, while the intensity shows how much deprivation poor people experience on average. A cut-off point of 0.33 is used for the multidimensional poverty headcount ratio; that is, a household is multidimensionally poor if the MPI is greater than 0.33. The population vulnerable to poverty is defined as those who experience 20-32.9 percent intensity of deprivation, and the population in severe poverty are those with an intensity of 50 percent or higher (that is, if the MPI is 0.50 or higher). contribute less to multidimensional poverty across have deteriorated. While in-camp refugees are more all refugee groups. There is a similar pattern for the pessimistic about the changes in their households’ host community around the refugee camps. Low living standards, there is no significant difference education, together with limited access to electricity, in perceptions among OCP refugees in Addis Ababa housing, assets, sanitation facilities, and drinking and their hosts. The considerably high negative water, mean that low living standards contribute perception about changes in household living more to overall poverty among refugees. standards indicates that well-being has been worsening for everyone over the past few years, but 5.1.4 Food security even more so for refugees in camps. Refugees and hosts perceive that household living standards have deteriorated over time. To capture Refugees, on average, have poor food and nutrition subjective well-being, the survey asks if the living security outcomes compared to their hosts. The standard of the household or their community has extent of food insecurity measured by the food improved or worsened in the past five years (Figure insecurity scale is significantly higher for refugees 5.8a) and in the past 1 year (Figure 5.8b). Overall, than hosts, both for in-camp and for out-of-camp most households feel that their living standards refugees (Figure 5.9). While the food insecurity Figure 5.8: Perceived changes in household living standards a. Last 5 years b. Last 1 year 100 100 90 90 80 80 70 70 60 60 Percent Percent 50 50 40 40 30 30 20 20 10 10 0 0 Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees In Camp Addis Ababa Total In Camp Addis Ababa Total Worse Same Better Worse Same Better Source: World Bank Staff based on SESRE 2023. Note: The survey asks how the household living standard has changed compared to last year and the last five years. 47 Refugees’ Aspirations Figure 5.9: Food insecurity scale for refugees and hosts score—the number of food groups consumed out of twelve—is 7.5 for hosts and 6.5 for refugees (Figure 8.1 5.10a). Overall, the average dietary diversity score 7.0 is also lower for in-camp refugees than their hosts. The share of households with acceptable food 3.4 consumption status—food consumption score of 4.0 35 or above—is considerably lower among in-camp 2.9 2.1 refugees (49 percent) than their host (74 percent). The relatively lower dietary diversity could be due In Camp Addis Ababa Total Hosts Refugees to refugees having limited access to diverse food as they depend on aid. Most Addis Ababa refugees and Source: World Bank Staff based on SESRE 2023. their hosts have an acceptable food consumption scale44 gap between refugees and hosts is status (Figure 5.10b). higher for in-camp refugees, the gap is relatively narrower for Addis Ababa refugees. The average 5.1.5 Shocks and coping strategies food insecurity scale for in-camp refugees is “8” Market-related shocks are common, but refugees and for their hosts it is “4” out of 10; that is, in- are exposed to more diverse shocks than their hosts. camp refugee households experienced about eight While Ethiopian households face a plethora of risks food insecurity events while host households that affect their livelihoods—risks to assets, income, experienced about 4 in the past year. Consistent and food supply (Dercon et al 2005; Woldehanna et with other welfare indicators discussed, food al 2008)—market shocks related to rising food prices, insecurity tends to be more severe among in-camp food shortage, and health shocks appear to be most refugees than their hosts or OCP refugees. prevalent (Figure 5.11). High food prices drive the market shocks. Food shortage seems to represent In-camp refugees have less diverse diets and a crucial economic shock among refugees—roughly poor food consumption status compared to their 31 percent of in-camp refugees are affected by hosts. The average household dietary diversity food shortage—but not for their host communities. Figure 5.10: Dietary diversity and food consumption status a. Dietary diversity score (out of 12 groups) b. Food consumption status 100 8.4 7.8 90 7.0 8.2 80 70 60 6.5 6.0 50 Percent 40 30 20 10 0 Hosts Refugees Hosts Refugees Hosts Refugees In Camp Addis Ababa Total In Camp Addis Ababa Total Poor (0-21) Borderline (21-35) Acceptable ( > 35) Hosts Refugees Source: World Bank Staff based on SESRE 2023. Note: Dietary diversity score is calculated as the total number of food groups (out of 12) consumed by the household in the last seven days before the survey. The food groups are cereals, roots and tubers, vegetables, fruits, meat (including poultry and offal), eggs, fish and seafood, pulses and legumes and nuts, milk and milk products, oils and fats, sugar/honey, and others. Food consumption status is determined based on food consumption score. 44 Food insecurity experience is measured based on a scale that ranges between 0 and 10 and calculated by adding household’s experience related to the following events in the past year: (i) worried about having enough food, (ii) unable to eat healthy/nutrition food, (iii) only ate a few kinds of food, (iv) had to skip a meal, (v) adults ate less, (vi) ran out of food, (vii) adults were hungry but did not eat, (viii) went without eating for a whole day, (ix) restricted consumption so kids could eat, and (x) borrowed food or relied on friend/relative for help. 48 Refugees’ Aspirations Figure 5.11: Type of shocks experienced Figure 5.12: Shock coping strategies 50 60 45 40 50 35 30 40 Percent 25 Percent 20 30 15 10 20 5 0 10 Hosts Refugees Hosts Refugees Hosts Refugees In Camp Addis Ababa Total 0 Hosts Refugees Hosts Refugees Hosts Refugees Health shock (death/illness) Market shock (output/food/input prices) In Camp Addis Ababa Total Employment shock Production shock (drought, crop damage, livestock loss) Consumption smoothing Sale of household assets (incl. livestock) Political shock (insecurity, loss of land, displacement) Food shortage Borrow food or cash or purchase on credit Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. Moreover, insecurity and displacement-related 5.2 Determinants of welfare shocks are common for Eritrean refugees, with 14 The poverty profile in this section compares the percent having experienced a recent displacement characteristics of poor compared to non-poor event. This result is driven by refugees in the people. The previous section presents refugees’ and Alemwach refugee hosting site, all of whom moved host communities’ poverty and welfare patterns. to the refugee site within a few months before the This section substantiates the earlier discussions survey as a result in the conflict in Tigray, and would on poverty levels by describing the demographic, have reported a recent displacement event. geographic, and socioeconomic characteristics by expenditure quintiles for each refuge and host group Both refugees and host communities use separately, along with the poverty headcount rate “consumption-smoothing” to cope with the various across grouping variables (see Annex D, Table D.11). shocks they face. Households utilize a mix of coping The descriptive statistics are substantiated by results strategies to mitigate harm to their welfare that from a regression analysis examining correlates shocks cause. “Consumption smoothing”, among of poverty while holding other things constant. the major risk coping strategies, mainly involves The dependent variable is the natural logarithm relying less on preferred food and more on less of consumption per capita. That is, we compare expensive food (diet changes) and reducing the level of consumption to other variables to identify number of meals eaten daily (negative food intake). characteristics that correlate to a household being Borrowing food or cash from friends and relatives poor. Table D.12 and Table D.13 in Annex D show the and purchasing food on credit second represent the full results of the regressions on the determinants of second and third most common coping strategies. consumption per capita separately for in-camp and Refugees in camps and in Addis Ababa are more out-of-camp refugees and hosts. likely to rely on these coping strategies than their hosts (Figure 5.12). The results further show that Poor in-camp refugees have higher household both refugee and host households do not engage sizes, dependency ratios, and male heads. in adverse coping strategies, such as the sale The poorest refugee and host households have of (productive) assets that would make them significantly larger household sizes and dependency vulnerable to poverty. This could be because either ratios than the richest counterparts (Figure 5.13). they do not have enough assets to sell or because The average household size of the poorest in-camp the strategies they utilize are enough to cope with refugee and host households is more than double the effects of shocks. that of the richest households (Figure 5.13). Larger 49 Refugees’ Aspirations Figure 5.13: Household composition by quintiles household sizes than hosts, except in Addis Ababa. In light of the discussion above, the highest poverty 90 2.0 80 1.8 incidence among South Sudanese refugees could 70 1.6 be associated with their high dependency ratio and Dependency ratio 60 1.4 1.2 50 Percent 40 1.0 high number of female-headed households. In-camp 0.8 30 20 0.6 refugees working inside the camp tend to exhibit 0.4 10 0.2 0.0 lower poverty incidence (81 percent) than those 0 working outside the camp (88 percent). Poorest 2 3 4 Richest Poorest 2 3 4 Richest Poorest 2 3 4 Richest Poorest 2 3 4 Richest In Camp In Camp Addis Ababa Addis Ababa Refugees Hosts Refugees Hosts The poor tend to live in households headed by Household size Adult equivalent Age of head (year) individuals with limited education. This trend is Source: World Bank Staff based on SESRE 2023. Note: Primary axis labels represent household size/adult equivalent. evident among both refugees and hosts, where a lower level of educational attainment by household household sizes for the poor are mainly driven by a heads and members correlates with increased larger number of children (under age 15). The data poverty. While building human capital represents further show that the poorest refugees and hosts an essential pathway out of poverty, there appears are more likely to have married and older household to be low human capital among refugee and host heads compared with the richest counterparts households, as indicated by the household head (Figure 5.14). Richest in-camp and out of camp and members’ low education. The data reveals refugee households are more likely to have female- that poverty incidence is more prevalent among headed households compared to the poorest. There households with no or minimal education (Annex is no difference in the gender of the household head D, Table D.11). Conversely, poverty incidence tends among the poorest and richest host households to decline with increased education level of the (Annex D, Table D.11). household head and members. These findings underscore the critical role of education as a means Location is an essential determinant of monetary to alleviate poverty among refugees and host poverty. Monetary poverty is highest among South communities in Ethiopia. The regression results Sudanese refugees (89 percent) (Figure 5.15). There also indicate that increasing years of schooling of is a significant difference in poverty rates between the household head is associated with increased in-camp refugees and their hosts, the gap being household consumption (Annex D, Table D.12); the highest in the Eritrean domain. As discussed average household expenditures linearly increase, in Chapter 2, refugee households have larger and poverty headcount decreases, as the education Figure 5.14: Demographic characteristics by quintile Figure 5.15: Poverty headcount rate for in-camp refugees and their hosts, by domain 90 50 100 80 45 70 40 90 60 35 80 Age (years) 30 50 Percent 25 70 40 20 60 30 Percent 15 20 50 10 10 5 40 0 0 30 Poorest 2 3 4 Richest Poorest 2 3 4 Richest Poorest 2 3 4 Richest Poorest 2 3 4 Richest 20 10 In Camp In Camp Addis Ababa Addis Ababa 0 Refugees Hosts Refugees Hosts Eritrean Somali South Sudanese Female headed Married head Age of head (year) Hosts Refugees Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. Note: Primary axis labels represent gender and marital status. 50 Refugees’ Aspirations level of the household head increases. This is only host households (Annex D, Table D.12). Mobile phone the case for Addis Ababa refugees and hosts (Figure ownership and ownership of a nonfarm enterprise 5.16). However, for in-camp refugees, there appears positively correlate with household welfare for out- to be no response to expenditure on an additional of-camp refugees and hosts. Ownership of a non- level of education of the household head compared farm enterprise also appears to correlate positively to other refugees. In-camp host community with welfare for in-camp refugees. households are higher, on average, than for in-camp refugees, and returns to the education level of the The poorest refugees and hosts tend to have household head for these households appears to be worse labor market outcomes than the richest. slowly increasing. Figure 5.18 summarizes labor market outcomes by expenditure quintiles for in-camp and Addis Ababa Household welfare is linked to possession of certain refugees and their hosts. Labor force participation assets or access to services. We assessed a number and employment-to-population rates increase with of indicators, including whether the household welfare, and unemployment rates fall with increasing has access to electricity; possesses livestock; has welfare (Figure 5.18a). This underscores the critical a mobile phone; owns agricultural land; runs a role labor market participation or employment non-farm enterprise; or has bank accounts. These plays for poverty reduction among refugees and “wealth” indicators show stark differences between their hosts. Looking at the sectoral distribution of the poorest and richest in-camp refugees and their employment, the poorest refugees—in- and out-of- hosts. The poorest in-camp refugees and their hosts camp—tend to be employed in the service sector. tend to have limited access to electricity, mobile While employment in the industry sector is low for phones, bank accounts, and non-farm enterprises refugees, the poorest are less likely to be employed (Figure 5.17). Livestock and agricultural holding do in the industry sector than the richest. The poorest not show a clear pattern among the poorest and the hosts of in-camp refugees are more likely to be richest. For Addis Ababa refugees and hosts, there employed in agriculture, and the richest appear tends to be increased access to electricity, ownership to be employed in the industry or service sectors of mobile phones, bank accounts, and non-farm (Figure 5.18b). Not surprisingly, the poorest hosts enterprises across expenditure quintiles. Regression of in-camp refugees and the poorest refugees work results show that possessing a bank account, a in low (or medium)-skilled occupations, while the mobile phone, and access to electricity positively richest are employed in high-skill occupations correlate with consumption for in-camp refugees and (Figure 5.18c). Figure 5.16: Poverty incidence decreases with education of the household head 90,000 1.00 Predicted total expenditure 80,000 0.90 Predicted poverty rate 70,000 0.80 60,000 0.70 0.60 50,000 0.50 40,000 0.40 30,000 0.30 20,000 0.20 10,000 0.10 - - No Primary Primary Secondary Secondary Post- No Primary Primary Secondary Secondary Post- education incomplete complete incomplete complete secondary education incomplete complete incomplete complete secondary In Camp Refugees In Camp Hosts Addis Ababa Refugees In Camp Refugees In Camp Hosts Addis Ababa Refugees Addis Ababa Hosts All Refugees All Hosts Addis Ababa Hosts All Refugees All Hosts Source: World Bank Staff based on SESRE 2023. 51 Refugees’ Aspirations Figure 5.17: Household wealth indicators by expenditure quintiles 100 90 80 70 60 50 Percent 40 30 20 10 0 Poorest 2 3 4 Richest Poorest 2 3 4 Richest Poorest 2 3 4 Richest Poorest 2 3 4 Richest In Camp Refugees In Camp Hosts Addis Ababa Refugees Addis Ababa Hosts Electricity Mobile phone Any livestock Bank account Agricultural holding Nonfarm enterprise Source: World Bank Staff based on SESRE 2023. Figure 5.18: Labor market outcomes by expenditure quintiles a. Key labor market indicators b. Employment by sector 90 100 80 90 70 80 60 70 Percent 50 60 Percent 40 50 30 40 20 30 10 0 20 10 Poorest 2 3 4 Richest Poorest 2 3 4 Richest Poorest 2 3 4 Richest Poorest 2 3 4 Richest 0 Poorest 2 3 4 Richest Poorest 2 3 4 Richest Poorest 2 3 4 Richest Poorest 2 3 4 Richest In Camp In Camp Addis Ababa Addis Ababa Refugees Hosts Refugees Hosts In Camp In Camp Addis Ababa Addis Ababa Labor force participation Employment to working age population ratio Refugees Hosts Refugees Hosts Unemployment rate Agriculture Industry Services c. Employment by skill (occupation-based) d. Work location for in-camp refugees 80 70 70 60 60 50 Percent 50 40 Percent 40 30 20 30 10 20 0 10 Poorest 2 3 4 Richest Poorest 2 3 4 Richest Poorest 2 3 4 Richest Poorest 2 3 4 Richest 0 Poorest 2 3 4 Richest In Camp In Camp Addis Ababa Addis Ababa In Camp Refugees Refugees Hosts Refugees Hosts Low-skill Medium-skill High-skill Inside the Camp Outside the Camp Source: World Bank Staff based on SESRE 2023. A larger proportion of refugees work inside the who are poorer overall. Regression results show that camp across the distribution. Yet, better-off an increase in the share of employed household refugees are more likely to work inside the camp. members is associated with increased household Regarding location, although working outside the expenditure for in-camp refugees, their hosts, and camp is shown to have significant wage effects (see out-of-camp refugees (Table D.12 in Annex D). The Chapter 3), the data show that the poorest in-camp predicted poverty rate decreases with the share refugees are more likely to work outside the camp of employed household members, indicating that than the richest (Figure 5.18d), an effect apparently employment is essential to lowering poverty for in- driven by refugees from South Sudan and Somalia, camp refugees (Figure 5.19). 52 Refugees’ Aspirations Figure 5.19: Poverty rates and employment for refugees and hosts 1 .15 .8 .1 .6 .4 .05 .2 0 0 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 Share of employed HH members Share of employed HH members In Camp Refugees In Camp Hosts Addis Ababa Refugees Addis Ababa Hosts Source: World Bank Staff based on SESRE 2023. Note: Predicted marginal probabilities of poverty based on the share of employed members after controlling for other factors. Poverty relates to lack of access to markets for refugee households in Ethiopia are vulnerable in-camp refugee households. Families with better to various shocks, including market shocks that access to essential resources such as education, harm their well-being. Poor households are more healthcare, clean water, and stable employment likely to experience shocks; concurrently, they are are more likely to experience improved economic less equipped to devise coping strategies. Market stability and well-being. Access to these services shocks, often manifested through escalated provides a foundation for building a more secure food prices, seem to predominantly harm host financial future, enabling households to invest in communities, while refugees are less affected their growth and development. Consequently, (Annex D, Table D.12). This disparity may stem from communities with better access to resources refugees’ heavy dependence on food assistance tend to have lower poverty incidences, as these and remittances, coupled with the international critical assets empower individuals to break free community’s concentrated efforts on enhancing from the cycle of economic hardship. For analysis, refugee livelihoods. A significant observation is resource access is proxied by remoteness or that political shocks, closely linked to displacement proximity to resource hubs and market accessibility. and insecurity issues, consistently result in adverse Descriptive statistics show that poverty rates are welfare outcomes for out-of-camp refugee and host higher in medium market-accessibility areas and community households (Annex D, Table D.12). lowest in high-market accessibility areas (Annex D, Table D.11). Poverty incidence also tends to be 5.3 Cost of basic needs for refugees lower in areas closer to Woreda capitals. Results This section estimates the cost of basic needs for from regression analysis for in-camp refugees in-camp refugees in Ethiopia and analyzes the show that consumption expenditure per capita determinants of these costs. This section identifies negatively correlates with distance to a Woreda how much it costs to meet basic needs through capital; with a 1 percent increase in mean distance aid. It also shows how the need for assistance to the capital reducing consumption expenditure depends on the degree of economic inclusion of per capita by 0.09 percent, holding other factors refugees, following the work of Atamanov et al. constant (Annex D, Table D.13). Moreover, living (2023). Basic needs are defined using monetary in high market-accessibility areas is associated poverty lines. The approach captures the cost of a with a 0.24 percent increase in consumption minimum standard of living, grounded in a well- expenditure per capita. established methodology. (see Box 5.4 for in-depth explanation of the methodology used). On average, Market and political shocks harm the welfare of in-camp refugees receive about 56 percent of total refugees and host communities. As mentioned, consumption from aid or assistance (Figure 5.20). 53 Refugees’ Aspirations Figure 5.20: Share of consumption provided in-kind or for free by Figure 5.21: Poverty incidence at consumption and pre-assistance consumption per capita quintiles among in-camp refugees consumption levels 80 100 90 70 80 60 70 60 50 Percent 50 Percent 40 40 30 30 20 20 10 0 10 Hosts Refugees Hosts Refugees Hosts Refugees Addis Ababa All camp Total 0 Poorest 2 3 4 Richest Total consumption Pre-assistance income Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. Note: Quintiles are constructed for in-camp refugees only. Note: Poverty rates are calculated based on $2.15 in the 2017 PPP line using total consumption and pre-assistance income The share is substantially higher for refugees from refugee to cover, depending on economic inclusion the bottom quintile (67 percent) than those from the across the three scenarios. Under a “no economic top quintile (24 percent). opportunities” scenario—in which refugees do not work and must rely solely on aid or assistance, We estimate poverty levels for refugees and host the annual cost of basic needs per refugee is communities using the standard consumption approximately US$378. Under the “current” aggregate and the pre-assistance consumption scenario—where refugees can find opportunities aggregate. We present poverty rates separately to earn money or work—the amount of assistance for camp and out-of-camp refugees (Figure 5.21). needed to cover basic needs reduces annual costs Notably, poverty headcount for refugee camps is by 44 percent to US$210 per capita. The saving can markedly higher when considering pre-assistance be viewed as an economic-inclusion dividend made consumption; that is, consumption after deducting possible by Ethiopia’s prevailing refugee policies. aid or assistance received (96 percent) as opposed Under a hypothetical “full inclusion” scenario— to total consumption (84 percent). However, for where in-camp refugees have equal opportunities as refugees living outside of camps and for host hosts—the cost of basic needs decreases further to communities, the changes in poverty rates are not only US$78 per refugee, per year. large. A similar trend is observed with the poverty Figure 5.22: Costs of basic needs per refugee per year under gap. These findings underscore humanitarian different scenarios aid’s vital importance for refugee camps. The lack USD per capita per year, Dec 2022 prices of substantial change in poverty among out-of- camp refugees is due to their greater reliance on remittances rather than direct aid. Compared to the “no economic opportunities” scenario (see Box 5.4), Ethiopia’s “current” economic integration model reduces costs by 44 Scenario 1: Scenario 2: Scenario 3: percent to an annual cost of US$210 per capita. no economic opportunities current scenario full inclusion Figure 5.22 shows the yearly costs of basic needs per Source: World Bank Staff based on SESRE 2023. Note: The costs are in December 2022 prices. 54 Refugees’ Aspirations Box 5.4: Estimation of the cost of basic needs for refugees In addition to their own resources, refugees rely on humanitarian aid to cover their expenditures on food, sanitation, hygienic products, and essential non-food items. “Successful” integration and economic inclusion—that is, earning sufficient income to be no longer poor and to consume more than the (international) poverty line—of refugees bring higher self-reliance and less reliance on humanitarian assistance. This opens two tracks for investigation: (i) First, how much aid would be needed if the policy objective were to bring refugee consumption up to the poverty line. The answer to this question is found by identifying the poverty gap for refugees. (ii) Explore the factors that determine, or at least are associated with, the size of the poverty gap. In Ethiopia, the policy on living out-of-camp is somewhat unique in that refugees who live in camps are eligible for humanitarian assistance; all refugees receive the complete package. However, that package regularly changes when funding gaps arise. Refugees who live out-of-camp forego any assistance. Still, they can access education and health services. Regarding selection for the OCP, only those who are “better off”— that is, they can rely on remittances—qualify and are selected for OCP. This implies that the OCP refugees have a vastly different profile from those living in camps. The focus of this section is first to identify how much it would cost if basic needs were met through aid alone, and next, how the need for assistance depends on the degree of economic inclusion of refugees. For this purpose, basic needs are defined using monetary poverty lines (Atamanov et al., 2023). Monetary poverty lines are used because they both capture the cost of a minimum standard of living and follow a well-established methodology combining: (i) a food allowance for adequate nutrition/minimum caloric intake using a national basket of goods, and (ii) a non-food allowance that captures the cost of essential non-food items such as clothing, shelter, and private expenses on health and education (Ravallion, 1998). The preferred poverty line was $2.15 per capita per day in 2017, and PPP converted to Ethiopian Birr in December 2022. The first step in estimating the cost of basic needs for refugees is the calculation of pre-assistance income or refugees—a proxy for income earned by refugees. The pre-assistance income for in-camp and out-of-camp refugees are calculated separately. For in-camp refugees, this involves deducting assistance from total consumption, including humanitarian assistance and housing—a proxy for gifts received from (international) donors. For those out-of-camp, only humanitarian assistance is deducted. The information about food and non-food consumption provided in-kind or free to households provides a measure of the role of existing humanitarian assistance. We assume that aid organization and the government provide these food and non-food products and services. The expenditure sources mapped to humanitarian aid to calculate pre-assistance income are consumption use of donation items from government or NGOs, sale of donation items from government or NGOs, donations in cash from government or NGOs, and imputed value of owned or subsidized dwelling units for in-camp refugees. The cost of basic needs for refugees is assessed using three scenarios: (i) “No economic opportunities”—the costliest scenario that assumes that refugees do not earn any income and need aid to cover all their basic needs. In this baseline scenario, the poverty line’s full value is used as a proxy for costs. (ii) “Current”—based on the premise that, in practice, refugees find opportunities to earn money, even in the most restricted environments. By allowing refugees to work, the assistance needed to cover basic needs is lower. This could be a stringent assumption in light of Ethiopia’s refugee policy that does not facilitate swift access to work permits and refugee mobility within camps. The cost of basic needs under this current scenario is measured by removing humanitarian aid from total household consumption, then taking the difference between the poverty line and pre-assistance consumption. This difference indicates how much assistance is needed to bring the consumption of refugees to the poverty line. The value is lower than the costs under the “no economic opportunities” scenario, with the savings viewed as an economic inclusion dividend made possible by Ethiopia’s prevailing refugee policies. (iii) “Full inclusion”—uses the current poverty gap of Ethiopian hosts as a proxy for basic needs costs, where an “average” refugee resembles an “average” Ethiopian in terms of human capital, access to productive assets, and economic opportunities. The “no economic opportunities” and “full economic inclusion” scenarios are hypothetical and only serve as upper and lower bounds for aid necessary to cover the costs of basic needs. 55 Markets and Opportunities 6. Markets and Opportunities R efugees must be able to engage in local markets to find better livelihoods and sustainable economic opportunities. Local labor markets that varies widely by country income status. Roughly half of refugees hosted in low-income countries live in camps (UNHCR, 2022b), but this share is shape the employment trajectories of refugees. much higher in Ethiopia (88 percent). Long-term Restrictions on land access for refugees restrict their encampment policies leave refugees isolated with access to rural labor markets, primarily shaped by limited or no economic rights, a situation that wastes agricultural activities. Livelihood activities in cities their human capital and capacity for work (World or work similar to that found in urban areas are Bank, 2017; Ibáñez et al., 2022). Although it may most promising for refugees to utilize their labor and appear practical to keep refugees in camps from the skills. Yet, many refugee camps are in more agrarian perspective cost, the speed of setting-up, delivering locations, and local labor market characteristics and services, identifying individuals, and other reasons, connectivity drive refugees’ labor market outcomes refugees in camps (or specific hosting areas) live (Hedberg and Tammaru, 2013; Kalter and Kogan, unproductive, unfulfilled lives that do not contribute 2014; Kogan and Kalter, 2020; Schuettler and Caron, to the local economy (World Bank, 2017). Usually, 2020; Dorian and Burmann, 2023). the only option for economic participation these refugees have is to work or in the informal sector in The GoE vision to create sustainable livelihood surrounding host communities. opportunities and build refugees’ self-reliance and resilience has yet to be fully implemented; roughly In Ethiopia, refugees live in 24 camps located across 88 percent of refugees in Ethiopia remain in camps different regions45. Refugee camp locations are based on SESRE data. Globally, approximately one- diverse. Some camps are part of Woreda cities, some quarter of all refugees live in camps, a proportion are close to Zone capital cities, some are remote, 45 In Ethiopia, the refugee camps are located in Tigray, Afar, Amhara, Somali, Benishangul-Gumuz, and Gambella regions. Eritrean refugees who speak Tigrigna are basically located in Tigray region, though they moved to Amhara region following the North Ethiopia Conflict (IOM, 2023). Refugees who speak the Afar language from Eritrea are settle in Afar region. Refugees from Somalia are located in different parts of the Somali region. Sudanese refugees live in Gambella region, whereas the South Sudanese settled in Benishangul-Gumuz. 56 Markets and Opportunities some are near a border to their home country, some hubs, connectivity, and local markets, and highlight are in the lowlands and some in the highlands. They differences in refugee communities depending on are spatially dispersed, have different geographic, location. Second, we identify in-camp refugees’ social, and economic contexts, and are in different performance in the labor market and investigate if ecological Zones. For example, about 38 percent there is a spatial disparity in such outcomes among of refugees live in drought-prone lowland and refugees. We discuss refugees’ spatial disparity in pastoralist areas, whereas 60 percent live in humid labor market access and outcomes based on their reliable lowland areas (Annex D, Figure D.31). In proximity to Zone capital cities, Woreda cities, many refugee-hosting areas in Ethiopia, except and the nearest international border. In addition, for a few places such as Addis Ababa, refugees and we investigate their level of accessibility to the host communities share cross-border cultural and given market where they are located. Third, using economic connections; and common ties of kinship, an econometric model, we assess to what extent language, and ethnicity (Vemuru et al., 2020). refugees’ group differences in terms of labor market outcomes correlates with several variables: local Location greatly affects socio-economic outcomes, factors, proximity to resource hubs, and connectivity. economic activities, and livelihood opportunities, and poverty levels vary profoundly by location in 6.1 Spatial disparities in refugees labor Ethiopia. Livelihood activities vary throughout the market access and outcomes country and the refugee-hosting zones. Rural labor To better understand refugees’ spatial disparities, concentrates in the agricultural sector, with low we look at their remoteness—measured as non- and off-farm employment in rural areas and proximity to the nearest Zone capital cities, Woreda small towns (Pimhidzai et al., 2022), while work in cities, and the international border—as well as the service and manufacturing sectors concentrates their market accessibility. We selected the capital in urban centers46. Livestock production and sale city of each Zone as it is a resource and market hub represent the main livelihoods in lowland pastoral for surrounding Woredas and Kebeles (Box 6.1). areas. Poverty rates are higher among households Usually, these cities serve as a commerce center in the drought-prone lowlands, and the likelihood of for agricultural goods and manufacturing products escaping poverty is higher for households in lowland and provide better employment opportunities. pastoral areas than those in moisture-reliable Moreover, Zone Capital and Woreda cities offer highlands (World Bank, 2020). Refugees in camps better education and health services and improved can neither choose nor participate in the local transportation and communication infrastructure. In agricultural economy, so disparities in livelihood addition to cities and towns, people often use border opportunities depending on location matter for areas to trade and purchase goods at better prices. refugees. For example, employment rates differ depending on the hosting zones, helping to explain In Ethiopia, refugees’ locations differ vastly in the different labor market outcome of refugees’ terms of proximity to resource and economic experience across the country. hubs. About 37 percent of refugees live within 10 kilometers of the nearest Woreda city and another 43 This chapter aims to better understand how camp percent live within 10 to 20 kilometers. Zone capital locations determine labor market outcomes and cities are farther away, but almost half (45 percent) highlights the importance of refugees’ location of in-camp refugees live within 20 kilometers of the as part of the development strategy for refugees nearest Zone capital city (Figure 6.1a). Borders seem in Ethiopia. First, we define refugees, resource farther, with 18 percent of refugees living within 30 46 Labor Force and Migration Survey 2021. 57 Markets and Opportunities Box 6.1: Measurement of proximity and market access index in Ethiopia The analysis measures the nearest Zone and Woreda capital cities and the closest international border from refugee camps using straight-line distance in a projected coordinate system (Euclidean distance). The study indexed refugees’ proximity to resource hubs by classifying their presence to a combination of distance to cities and borders. Level one is the presence of refugees within a radius of 20km from a Zone capital city. Level two is 10km away from a Woreda city but not within a radius of 20km from the Zone capital city. Level three is for refugees located 30km from the nearest international border but not within a radius of 20km from the Zone capital city and 10km from Woreda city. We classify level four as “remote”; that is, not located 30km from the nearest international border, not within a radius of 20km from the Zone capital city, and not 10km from Woreda city. The market access indicator in Ethiopia is measured at the Woreda level. Accessibility for a Woreda is estimated as the sum of the travel time of the weighted population to the destination Woredas. With Woreda- to-Woreda origin-destination matrices, we calculate market accessibility by the following equation (Donaldson and Hornbeck, 2016; World Bank, 2019b): where is market access at Woreda “o”, is the trade cost between two Woredas “o” and “d”, is the population of Woreda “d”, and is the trade elasticity. Trade costs between two Woredas, is defined by =exp ( ) with = 0.02 and the optimal travel time between Woredas using the transport network of 2020. The trade elasticity, has a value of 8.28 (Eaton and Kortum, 2002). kilometers of the nearest border. When defining Woreda city, and 17 percent of in-camp refugees live mutually exclusive location categories to measure in remote areas far from a Zone capital city, Woreda proximity to resource hubs, we see that 44 percent city, or a border. When looking at accessibility, as of refugees live closest to the nearest Zone capital defined by a market accessibility index, more than city. Another 28 percent live closest to a Woreda City, one-third of the refugees are located in areas with which is not a Zone capital city. About 11 percent low market accessibility (Figure 6.1b). live close to a border but not the Zone capital or Figure 6.1: Refugee incidence a. Against distance to cities and borders b. By market accessibility, proximity to resource hub To Woreda City Proximity to economic hub 37 43 19 44 28 11 17 1-10km 10-20km >20km Nearest to Zone Nearest to Woreda but not Zone To Zone capital Nearest to border but not Zone & Woreda Remote 45 26 29 1-20km 20-100km >100km Market accessibility 24 40 36 To nearest border 18 59 23 High accessibility Medium accessibility Low accessibility 1-30km 30-50km >50km Source: World Bank Staff based on SESRE 2023. Note: High, medium, and low accessibility refers to the level of market access, with >0, [-0.5, 0], and <-0.5 standard deviations from the average, respectively. 58 Markets and Opportunities Labor market outcomes47 differ by proximity to rate increases by 14 percentage points for refugees resource hubs and connectivity. The labor force closer to Zone capitals and with higher market participation rate for refugees is highest in remote accessibility (Figure 6.3b). locations and areas with poor connectivity (Figure 6.2). Yet, most refugees in the labor force in remote The likelihood of refugees working in the and low-connected locations are unemployed, agriculture sector increases with remoteness. The highlighting challenges for those who want to work share of employed refugees in agriculture rises from to find employment opportunities (Figure 6.3a and 7 percent in camps nearest Zone capital cities to 40 Figure 6.3b). In contrast, refugees near borders and percent in remote camps (Figure 6.4a). Similarly, Zone capitals have the highest employment rates. refugees in areas with low market accessibility have (Figure 6.3a and Annex D, Figure D.32). Refugees a higher share of employment in the agriculture benefit from being close to Zone capital cities sector (Figure 6.4b). The labor market in remote and as the cities are resource hubs, creating many less connected areas is predominantly agrarian,48 positive economic and social spillover effects on providing worse employment opportunities for surrounding areas. For example, the employment refugees other than the agriculture sector. Yet, they Figure 6.2: Labor force participation rate by proximity to resource do not have easy access to land to work in agriculture. hub, market accessibility As a result, a higher share of the economically active 55 working-age refugee population in these areas 44 49 remains unemployed (54 percent). 42 37 40 40 Proximity to resource hubs and better accessibility increase the likelihood for refugees to work in non-agriculture sectors. For example, in the camps LFPR LFPR nearest to Zone capital cities, 27 percent of the Nearest to Zone High accessibility Nearest to Woreda but not Zone Medium accessibility employed workers engage in trade activity and Nearest to border but not Zone & Woreda Low accessibility Remote 62 percent work in other service sectors (Annex D, Source: World Bank Staff based on SESRE 2023. Figure D.33). Since refugees are not better positioned Note: High, medium, and low accessibility refers to the level of market access, with >0, [-0.5, 0], and <-0.5 standard deviations from the average, to work in the formal private or public sector, their respectively. participation in trade and service relates to economic Figure 6.3: Refugees’ labor market outcomes a. By proximity to resource hub b. By market accessibility 68 64 60 60 56 54 50 50 46 44 40 40 36 32 Employment rate Unemployment rate Employment rate Unemployment rate Nearest to Zone Nearest to Woreda but not Zone Nearest to border but not Zone & Woreda Remote High accessibility Medium accessibility Low accessibility Source: World Bank Staff based on SESRE 2023. Note: High, medium, and low accessibility refers to the level of market access, with >0, [-0.5, 0], and <-0.5 standard deviations from the average, respectively. 47 This analysis uses the relaxed definition to measure the current employment status of the host community and refugees. 48 As the previous chapter highlighted, most of refugees engaged in agriculture activity are livestock holders (see Chapter 3). 59 Markets and Opportunities Figure 6.4: Sectoral employment a. By proximity to resource hub b. By market accessibility Nearest to Zone High accessibility Nearest to Woreda Medium accessibility Nearest to border Remote Low accessibility 0 20 40 60 80 100 0 20 40 60 80 100 Share of individuals (%) Share of individuals (%) Trade Industry Other service Agriculture Trade Industry Other service Agriculture Source: World Bank Staff based on SESRE 2023. Note: High, medium, and low accessibility refers to the level of market Note: We classified the service sector as trade and another service, aiming to access, with >0, [-0.5, 0], and <-0.5 standard deviations from the average, shed light on refugees’ engagement in trade activity. respectively. activities inside camps or in the informal sectors lifetime earnings (Zanfrini and Giuliani, 2023). About in surrounding areas, such as construction, small 38 percent of the working-age refugee population shops, and street trades. The likelihood of engaging is between ages 15 and 24, and one-fifth of these in the industry sector is higher (30 percent) for youth is NEET. Spatial inequalities in NEET are refugees in highly-accessible areas, highlighting that significant, with more girls being NEET in any area. refugees can participate in different employment About 58 percent of young women and 34 percent sectors as long as their location is well-connected to of young men in remote camps are NEET, but markets. only 26 percent of young women and 11 percent of young men in camps nearest to Zone capital The prevalence of youth refugees not in cities are NEET (Figure 6.5a). Similarly, a higher employment, education, or training (NEET) share of young women (47 percent) and young increases with remoteness and poor connectivity. men (27 percent) are NEET in low-connected areas Youth without employment, education, or training compared to others (Figure 6.5b). decreases their future labor market outcomes and Figure 6.5: The share of refugee youth who are NEET a. By proximity b. By market accessibility 58 47 40 32 34 34 26 27 26 17 13 15 11 9 Nearest to Zone Nearest to Woreda Nearest to border Remote High accessibility Medium accessibility Low accessibility but not Zone but not Zone & Woreda Boys Girls Boys Girls Source: World Bank Staff based on SESRE 2023. Note: High, medium, and low accessibility refers to the level of market access, with >0, [-0.5, 0], and <-0.5 standard deviations from the average, respectively 60 Markets and Opportunities 6.2 Effects of local factors on refugees’ labor refugees have a higher prospect of employment. market outcomes Across all model specifications, men are more likely This section estimates the effect of local factors to be employed than women. For example, in a local on refugees’ employment outcomes. It shows how market where only 50 percent of the working-age local factors matter for employment opportunities population is active, employment prospects are 12 by looking at refugees aged 18 to 64 not currently percent for female refugees but 22 percent for male studying. More specifically, it sheds light on the refugees (Figure 6.6). However, this difference in importance of accessible locations and proximity the probability of employment between male and to economic and resource hubs for refugees to female refugees disappears in local labor markets perform better in local labor markets and to access where more of the working-age population is active. sustainable economic opportunities. The analysis The finding implies that refugees will perform better uses household and individual information from in a labor market with better employment prospects. SESRE data and geospatial information. The Moreover, local unemployment levels affect the odds estimation applies logistic regressions to predict the of being employed for refugees, regardless of the effects of the various indicators on the probability gender of the refugee. The higher the unemployment of being employed and working in different sectors rates in local area, the lower the refugees’ chance of of employment (see Annex D, Table D.14). Annex D, obtaining jobs (Figure 6.7), consistent with existing Table D.15 shows the average marginal effects of the evidence (Azlor et al., 2020). explanatory variables. We discuss the results using predicted marginal probabilities of being employed Proximity to resource hubs increases refugees’ based on various local factors. chances or working, regardless of gender. In all of our proximity measurements, refugees nearest The local labor market structure affects49 the to the Zone capital cities have a higher chance of possibility of refugees finding jobs. Consistent with being employed (Annex D, Table D.16). Only about 34 existing evidence (Andersen et al., 2023), the study percent of male and 23 percent of female refugees reveals that in a local economy where most of the living 100 kilometers from a Zone capital city are working-age host population is in the labor market, employed. In contrast, the chance of obtaining Figure 6.6: Local labor supply effect of refugee’s odds of employment Figure 6.7: Local unemployment level matters to obtain jobs 1 .8 Probability of being employed Probability of being employed .8 .6 .6 .4 .4 .2 .2 50 60 70 80 90 100 2 4 6 8 10 12 14 16 18 20 Labor force participation rate Unemployment rate (5%) Female Male Female Male Source: World Bank Staff based on SESRE 2023. Note: Predicted marginal probabilities of being employed based on the Note: Predicted marginal probabilities of being employed based on the unemployment rate of the local market, tabulated by gender. labor force participation rate of the local market, tabulated by employment experience. 49 The analysis proxies the local labor market by the aggregate market of urban areas of each Zone where refugee camps are located. Seven Zones host refugees; this study calls them hosting Zones. North Gondar Zone hosts Eritreans in Alemwach camp. Awsi (Zone 1) hosts Eritreans in Asayita camp. Liben Zone hosts Somali refugees in Bokolmanyo, Buramino, Hilaweyn, Kobe, and Melkadida camp. Fafan Zone hosts Somali refugees in Aw-barre, Kebribeyah, and Sheder camp. Agnuak Zone hosts refugees from South Sudan in Pinyudo 1 and 2, Jewi and Okugo camp. Itang Special Zone hosts refugees from South Sudan in Tierkidi, Kule, and Nguenyyiel. Assesa hosts refugees from South Sudan in Bambasi, Sherkole, and Tsore. 61 Markets and Opportunities Figure 6.8: Distance to the nearest city and the chance of Figure 6.9: Employment and proximity to resource hubs obtaining a job for refugees .6 0.7 Probability of being employed 0.6 Probability of being employed .5 0.5 .4 0.4 .3 0.3 0.2 .2 0.1 .1 0.0 2 12 22 32 42 52 62 72 82 92 102 112 122 132 142 Nearest to Nearest to Nearest to Remote Distance to nearest capital city (Km) zone Woreda border Female Male Male Female Source: World Bank Staff based on SESRE 2023. Note: Predicted marginal probabilities of being employed based on proximity Note: Predicted marginal probabilities of being employed based on the to resource, tabulated by gender. distance to the nearest Zone capital city, tabulated by gender. a job increases to 59 percent for male refugees less likely to be employed then male refugees in and 47 percent for female refugees living within remote areas, but 13 percentage points in locations 10 kilometers of a Zone capital city (Figure 6.8). near Zone capital cities (Figure 6.9). Overall, proximity to resource hubs leads to better employment outcomes for refugees. The chance Refugees in Woredas well-connected to markets of being employed is higher for male refugees have better prospects of being employed. proximate to resource hubs by 41 percentage Compared to other locations, Woredas with above- points compared to those living in remote locations average market accessibility increase employment (Figure 6.9). probability for refugees (Annex D, Table D.15). Regardless of education level and gender, the chance Irrespective of distance to economic hubs, the of being employed is below 25 percent for refugees gender employment gap persists, with female living in a Woreda with a level of market access >1.5 refugees having lower chances of being employed. standard deviations below the average. Moreover, However, the employment gap between male and the gender gap in employment persists at any level female refugees narrows as the location gets more of market access but becomes more pronounced remote. Female refugees are 8 percentage points with decreased market accessibility. 62 Markets and Opportunities 7. Social Cohesion S ocial cohesion is vital for refugees’ ability to integrate and contributes to social development. While social cohesion is often defined Despite challenges, forced displacement does not always lead to poor social cohesion between refugees and hosts. Social cohesion can actually differently in different contexts, we define it here as improve due to the benefits refugees bring to “a sense of shared purpose, trust, and willingness host communities, and with positive interactions to cooperate” (Barron et al., 2023). To be socially between refugees and hosts. In remote areas, sustainable, communities must work together to refugees often increase local economic development overcome challenges, provide public goods, and by increasing the availability of labor and demand allocate resources fairly, and social cohesion has for products and services. Aid inflow accompanying long been seen as critical for solid institutions refugees can also promote economic development and economic growth (Easterly et al., 2006). This in the host community. Across the world, studies is often challenging in a refugee context, where show that refugees are more likely to have positive, refugees not only experienced traumatic shocks to rather than adverse, economic effects on the host their social, economic, and emotional wellbeing, community. Economic studies of refugee camps in but they also face host communities’ concerns East Africa tend to find benefits for local economic regarding how refugees affect the local labor development (Verme et al., 2021; Alix-Garcia et al., market, the availability of goods and services, 2018; Maystadt et al., 2014). In Ethiopia, Walelign and the environment (World Bank, 2023a). These et al. (2022) find that refugees increase income challenges are even more significant when refugee diversification and livestock product sales for hosts camps are in underdeveloped and underserved and increase local market activity. Similarly, in regions of the country, where there is greater Uganda, Zhou et al. (2022) found that increased competition over scarce resources, livelihood refugee inflows improved local access to health, opportunities, and services. education, and roads, and had no detectable effect 63 Markets and Opportunities on hosts’ attitudes toward refugees. Other evidence in the Somali domain and worse in the South from Uganda finds that interactions between hosts Sudanese domain, but socio-political tensions and refugees may help improve hosts’ attitudes over ethnic composition in the Gambella region (Betts et al., 2023). In some contexts, refugee inflows is also a factor (see Box 7.1). Evidence from many have been found to harden in-group identification settings show that host attitudes and propensity for and increase support for ideological extremes. This positive host-migrant contact increase with cultural was the case with refugee inflows in Denmark, for proximity between migrants and hosts (World Bank, instance, but only in rural areas (Dustmann et al., 2023b; Hainmueller et al., 2014; Betts et al., 2023). As 2019). On the other hand, refugees hosted in Austrian an historical example, political backlash during the municipalities for extended periods, as opposed to U.S. age of mass migration (between roughly 1850 to those who passed through, were found to reduce 1910) was more significant against immigrant groups support for anti-immigrant parties, pointing to the that were more culturally distant (Tabellini, 2020). benefits of refugee-host interactions (Steinmayr, In East Africa, the relationship between refugee- 2021). All-in-all, there is little evidence that refugee host interactions in Uganda and positive attitudes hosting tends to worsen attitudes toward refugees in was higher when there was greater cultural overlap the Global South (World Bank, 2023b). (Betts et al., 2023). In Ethiopia, Somali refugees and hosts benefit from speaking a common language 7.1 Attitudes between refugees and hosts and having a common religion, which is not always SESRE data show that, while some hosts have the case in the South Sudanese domain. However, negative attitudes towards refugees, most most South Sudanese refugees still think they are attitudes are generally positive. Sixty-five percent culturally similar to hosts (Figure 7.16). The worse of hosts agree that refugees are friendly and good attitudes towards refugees in the South Sudanese people, and only 20 percent are uncomfortable domain are also related to socio-political tensions with having a refugee neighbor. This is an important over ethnic composition described in Box 7.1. finding, highlighting the potential for integration policies. Host attitudes are generally most favorable Attitudes among male and female hosts are similar, in the Somali region and most negative around South but female hosts have slightly more positive Sudanese camps; the share not comfortable with attitudes, especially in the Somali domain. Female having a refugee neighbor increases to 37 percent in hosts are more likely to agree that refugees are good the South Sudanese domain. people (66 compared to 63 percent for males), and this gap is largest in the Somali domain (where it The greater degree of cultural and linguistic overlap increases to 89 compared 76 percent for males). between Somali refugees and their Ethiopian hosts A similar pattern is observed regarding being may explain why attitudes are generally better comfortable with having a refugee neighbor. Figure 7.1: Host response to “Refugees are good people” Figure 7.2: Host response to “Would you feel comfortable having a refugee as a neighbor?” 100 100 90 90 80 80 70 70 Percent 60 60 Percent 50 50 40 40 30 30 20 20 10 10 0 0 Eritrean Somali South Sudanese Addis Ababa All Hosts Eritrean Somali South Sudanese Addis Ababa All Hosts Strongly agree Agree Disagree Strongly disagree Comfortable Neutral Not comfortable Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. 64 Markets and Opportunities Box 7.1: Socio-political tensions in the Gambella Region The South Sudanese population is ethnically and culturally diverse, with more than sixty cultural and linguistic groups. The South Sudanese refugees in Ethiopia mainly speak five languages—Nuer, Juba-Arabic, Dinka, Murle, and Luo—and the majority are ethnic Nuer, who in South Sudan are pastoralists (UNHCR, 2023c; Peters and Golden, 2019). Over 90 percent of South Sudanese refugees in Ethiopia are in the Gambella region, a multi-ethnic region dominated by two ethnic groups: the agro-pastoralist Anywaa (Anyuak) and pastoralist Nuer (Hagos 2021). While these groups have a long history of peaceful coexistence, they also have a history of conflicts over land and water resources and political representation (Vemuru et al., 2020; Hagos, 2021). The Anywaa were the majority of the population until the mid-1980s, when an influx of South Sudanese refugees shifted the demographic composition towards Nuer (Feyissa, 2015). This trend continued with the influx of more South Sudanese refugees in 2013, creating a sense of marginalization among the Anywaa in terms of changes in demographic composition, widening educational disparities, and increasing insecurity (Vemuru et al., 2020). The struggle between these two ethnic groups in the Gambella region has created socio-political tensions and influenced South Sudanese refugees’ social integration (ReDSS, 2018). Figure 7.3: Host response to “Refugees are good people” by gender Figure 7.4: Host response to “Would you feel comfortable having a refugee as a neighbor?” by gender 100 100 90 90 80 80 70 70 Percent 60 60 50 Percent 50 40 40 30 30 20 20 10 10 0 0 Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Eritrean Somali South Sudanese Addis Ababa Eritrean Somali South Sudanese Addis Ababa Strongly agree Agree Disagree Strongly disagree Comfortable Neutral Not comfortable Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. Most Ethiopian hosts want refugees to have access have add to economic opportunities in Ethiopia. to free primary education and healthcare and Walelign et al. (2022) similarly find that refugees’ the right to work, and to live where they choose. positively contribute to hosts’ income diversification. Eighty-seven percent of hosts believe that refugees Fewer think that refugees increase insecurity or should have the right to free primary education are taking their land, yet these concerns exist for and healthcare, increasing to 95 percent in Somali many. Among hosts in the Somali, Eritrean, and areas, and falling to around 80 percent in the South South Sudanese domains, 86, 57, and 61 percent, Sudanese domain. Rates are similarly high regarding respectively, think that refugees have increased the right to work and to internal mobility. While overall economic opportunities, though this rate there is more skepticism in the South Sudanese is lower (30 percent) in Addis Ababa. Fewer than domain, still 63 percent of Ethiopian hosts agree half of hosts in all domains believe refugees are that refugees should have the right to work and 51 increasing insecurity or taking land, though security percent agree that refugees should have the right to concerns are moderately higher (49 percent) in the move and settle freely in Ethiopia. South Sudanese domain. These questions refer to hosts’ perspectives on the effects of refugees in Social acceptability for integration of refugees in Ethiopia generally, not specifically towards them high; almost half of all hosts agree that refugees and their communities. 65 Markets and Opportunities Figure 7.5: Share of hosts who agree refugees Figure 7.6: Host beliefs about refugee impact in Ethiopia should have access to... 120 90 80 100 70 80 60 60 50 40 40 30 20 20 0 10 Eritrean Somali South Sudanese Addis Ababa All Hosts 0 Right to work Right to internal mobility Eritrean Somali South Sudanese Addis Ababa All Hosts Right to free primary education Right to free healthcare Increased economic opportunity Increased insecurity Are taking our land Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. Most hosts do not think they have experienced Somali and South Sudan domains (47 and 35 percent, adverse effects from refugees. However, a sizeable respectively), where refugees and hosts rely on minority are concerned about the effects on firewood for cooking fuel. Tesfaye (2021) also shows employment, inflation, security, and deforestation that hosts perceive negative environmental impact in their communities, with significant differences in terms of deforestation and loss of wildlife of South across domains. The most consistent perceived Sudanese refugees in Bambasi Woreda. Security effects are economic competition and price concerns are highest at 34 percent in the South increases. Across domains, 29 to 39 percent of Sudanese and Addis Ababa domains. Fewer than 15 hosts think they have experienced either wage or percent of hosts in each domain think refugees are employment competition due to refugees.50 Beliefs deteriorating infrastructure (not presented). that refugees have increased prices are especially prevalent in the Eritrea and Addis Ababa domains Some hosts think refugees have improved local (70 and 81 percent, respectively), possibly reflecting infrastructure and access to health and education concerns over housing costs in Addis Ababa. Other services. Thirty-eight percent of Somali hosts studies have shed light on this phenomenon in believe refugees have improved local infrastructure, more detail. In rural areas, Ayenew (2021) finds and 36 percent think they have improved local that hosting refugees increased prices of food and services. In the South Sudan domain, these rates are agricultural inputs. Deforestation is a concern in the 16 and 18 percent. Figure 7.7: Negative experiences due to refugees Figure 7.8: Positive experience due to refugees 90 40 80 35 70 30 60 25 50 20 40 15 30 20 10 10 5 0 0 Eritrean Somali South Sudanese Addis Ababa All Hosts Eritrean Somali South Sudanese Addis Ababa All Hosts Price Increases Economic Competition Insecurity Increases Deforestation Improved Infrastructure Improved Services Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. 50 The numbers are combined here, but mainly measures employment competition since very few are concerned about wage competition. 66 Markets and Opportunities Figure 7.9: Are most Ethiopians/refugees in Ethiopia trustworthy? 100 90 80 70 60 Percent 50 40 30 20 10 0 Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees All Hosts All Refugees Eritrean Somali South Sudanese Addis Ababa Neither are trustworthy Only my group Only the other group Both are trustworthy Source: World Bank Staff based on SESRE 2023. Note: Combines two questions regarding trust in Ethiopians and refugees, with identical wording to both Ethiopians and refugees. Overall, hosts and refugees show similar trust rates women. The above questions on attitudes and in each other; still, refugees are generally more trust towards refugees, the rights refugees should trusting. Questions about the trustworthiness of have, and the effects refugees have had on Ethiopia hosts and refugees reveal that most people either can be combined into an index. To construct the trust both (hosts and refugees) or neither group. index, we average the response to the ten questions Only 17 percent of hosts trust other Ethiopians but examined above, rescaled to range from 1-4. The not refugees, and only 10 percent of refugees trust index is highest in the Somali domain (.46 standard other refugees but not Ethiopians. In comparison, 39 deviations above the mean) and lowest in the South percent of hosts and 55 percent of refugees trust both Sudan domain (.48 standard deviations below the groups. Once again, hosts’ trust towards refugees is mean). It is higher for female than male hosts in both highest in the Somali domain (67 percent) and lowest domains, by .14 standard deviations from average in in the South Sudanese domain (29 percent). On the the Somali domain and .19 standard deviations in other hand, refugee trust towards hosts is lowest in the South Sudan domain. the Eritrean camps (38 percent). The difference in attitudes between male and Results of combining the survey answers into an female hosts is not statistically significant, and index echo our prior findings; attitudes toward inter-group attitudes vary little by age or education. refugees are better in Somali areas, worse in We can study the characteristics associated with South Sudan areas, and slightly better among attitudes by putting this index into a regression Figure 7.10: Host attitudes Index Figure 7.11: Host attitudes index by gender 3.1 3.1 3.0 3 2.9 2.9 2.8 2.8 2.7 2.7 2.6 2.6 2.5 2.5 2.4 2.4 2.3 Male Female Male Female Male Female Male Female Male Female 2.3 Eritrean Somali South Sudanese Addis Ababa All Hosts Eritrean Somali South Sudanese Addis Ababa All Hosts Source: World Bank Staff based on SESRE 2023. Note: This index is an average of ten questions regarding beliefs about refugees’ character, the rights they should receive, and their impact on the host community. The scale ranges from 1-4, where more positive indicates better attitudes. 67 Markets and Opportunities framework. This is presented in Annex D, Table D.18. the lack of integration into Ethiopian society and While attitudes are less positive for men, as we have social interaction with Ethiopians (discussed later seen, this difference is not statistically significant in this section). after controlling for age and education. There is no clear pattern in attitudes by age and education. However, refugees are more trusting if they believe they are more culturally similar to their hosts. Based on regression analysis, the most significant Refugees responding positively to the question, predictor of positive host attitudes and trust is “Do you agree that you are culturally similar to the whether they think the presence of refugees has host community?” are 20 percent more likely to improved local infrastructure or services. This believe Ethiopians are trustworthy. This is true after hints at the importance of local service delivery controlling for gender, age, education, and time in in driving host attitudes. On average, controlling Ethiopia. Importantly, it also controls for domains, for other characteristics and regions, hosts who so this implies that variations in cultural proximity think refugees have improved local infrastructure within the domain drive this result. This indicates and service delivery score .41 standard deviations that cultural similarity plays a role in facilitating higher on the Attitudes Index and are 12 percentage social cohesion. On the other hand, we see that points more likely to feel hosts are trustworthy. This cultural similarity explains some but not all of the is consistent with extensive evidence that service high trust in Somali camps relative to other domains. delivery and aid inflows are crucial in improving social cohesion between hosts and refugees (World 7.2 Social interactions Bank, 2023a). From a policy perspective, this points Social and community integration is fundamental to the benefits of ensuring that aid and programs to refugees’ ability to improve their livelihoods, to support refugees also benefit hosts (Baseler et support systems, and economic integration. Having al., 2021). friends in the host country is a valuable resource for refugees; Ethiopian friends can provide valuable Host trust for refugees does not significantly information, employer connections, assistance with depend on gender, age, or education, and it does language, and countless other types of social and not increase with time in Ethiopia. While trust of economic support. This can also promote more refugees is higher for women and less-educated positive attitudes between groups. Social integration refugees, these differences are not statistically has improved well-being, health, and educational significant. There is little relationship between achievement for refugee adolescents across various trust and time in Ethiopia. This may result from settings (Boda et al., 2023). Figure 7.12: Share with family or friends in Ethiopia Figure 7.13: Share with friends in Ethiopia by demographic group 35 40 30 35 25 30 20 25 15 20 15 10 10 5 5 0 Eritrean Somali South Sudanese Addis Ababa All Refugees 0 Family Friends Age Under 30 Age 30-44 Age 45-64 Age Over 64 Female Male Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. 68 Markets and Opportunities Despite the generally positive attitudes described, does not appear to improve ease the creation social integration—measured by the friends and of social interactions or sharing resources with family refugees have in Ethiopia—is low. Only 7 hosts. The results in Column 2 of Annex D, Table percent of refugees report having family in Ethiopia, D.19 show—controlling for other characteristics, and 25 percent report having an Ethiopian friend including region and year of arrival—that refugee outside the refugee camp. This rate is slightly higher men are 6.7 percentage points more likely to have among OCP refugees in Addis Ababa but still relatively an Ethiopian friend than refugee women, and low at 11 for having family and 31 percent for having those who completed secondary education are 22 a friend. The share with Ethiopian friends is higher percentage points more likely to have an Ethiopian for men, but similar across age groups (though lower friend. However, no significant difference exists in for refugees over age 64). This masks some variation the ease of which these groups find it to have social across domains; refugees under age 30 are likely to interactions with hosts. Men are notably worse in have friends than older refugees in the Eritrean and terms of their reported ease of sharing resources. South Sudan domains. In contrast, refugees under age 30 are less likely to have friends in Addis Ababa. As time passes, Ethiopian refugees become more likely to have Ethiopian family and friends and find Many refugees report that social interactions market interactions more accessible; but this and sharing resources with hosts is “not easy,” occurs slowly and only translates into greater especially in the South Sudanese domain. Overall, ease of socializing or sharing resources. A refugee 30 percent of refugees say it is not easy to have social who has spent an additional ten years in Ethiopia interactions with hosts, and 34 percent report that it is only six percentage points more likely to have is challenging to share resources such as water and an Ethiopian friend and 8 percentage points more food. These rates fall to 21 for social interaction and likely to find market interactions easier. Still, 15 percent for challenge in sharing resources in the there is no effect on ease of social interactions or South Sudan domain. On the other hand, refuges sharing resources. do not report that it is difficult to conduct market interactions. Because of low social integration, refugees rely little on the local population in times of need. Country-wide, the relationship between integration When refugees are asked, “Who is most reliable outcomes and demographic characteristics in a time of need?” almost none respond with the is complex. More-educated male refugees are local population. Instead, refugees heavily rely on more likely to have Ethiopian friends, but this donations and their family and friends in Ethiopia, Figure 7.14: Share of refugees who think interactions with hosts Figure 7.15: Who do refugees rely on in times of need are “easy to do” 100 100 90 90 80 80 70 70 60 Percent 60 50 50 40 40 30 30 20 20 10 10 0 0 Eritrean Somali South Sudanese Addis Ababa All Refugees Eritrean Somali South Sudanese Addis Ababa All Refugees Friend/family in Ethiopia Friend/family abroad Local population, Ethiopians Selling/buying items Social interactions Sharing resources NGOs/donations Myself Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. 69 Markets and Opportunities while 9 percent of refugees respond that they would Figure 7.16: Share of refugees who agree they are “culturally depend on themselves. Only 0.2 percent of refugees similar to hosts” in camps and 2.3 percent in Addis Ababa rely most on 100 90 the local population in times of need. 80 70 Refugees with Ethiopian friends are more 60 likely to be employed and to work in high-skill 50 40 occupations. Annex D, Table D.20 uses another 30 regression framework to study the relationship 20 between labor market outcomes and various social 10 0 integration outcomes, controlling for domain, Eritrean Somali South Sudanese Addis Ababa All Refugees gender, age, education, and years in Ethiopia. For Source: World Bank Staff based on SESRE 2023. the regression, the social integration outcomes are: “having an Ethiopian friend and family”, whether they merely show that having an Ethiopian friend they find social interactions with hosts “easy to and sharing cultural similarity with hosts are the do,” and their perceived cultural similarity to hosts. social integration measures most closely related to The results show that having an Ethiopian friend is improved labor market outcomes. However, they associated with a 6 to 7 percentage point increase in do not necessarily explain who can, or does, work the probability of refugee employment and, among outside the camp. those who are employed, a 4 to 5 percentage point increase in the likelihood of being in a high-skill Low levels of refugee social integration is not due occupation (manager, professional, or associate to lack of cultural similarity with Ethiopian hosts. professional). The variables for having an Ethiopian While this is undoubtedly a challenge in some cases, family and finding social interactions easy are notably among South Sudanese, refugees generally positive (not statistically significant). In contrast, believe they are culturally similar to their hosts. the variable for cultural similarity is both large, This rate averages 78 percent for all refugees. It is at around 8 percentage points, and statistically lowest in South Sudanese camps at 68 percent and significant, highlighting the continued importance highest in Somali camps at 87 percent. Many of of cultural similarity for employment outcomes. these refugees who say they are culturally similar Among refugees in camps, working outside of the to their hosts respond that they have no Ethiopian camp appears to be relatively unrelated to all these friends and that social interactions with Ethiopians characteristics. It is important to remember that are complex. none of these relationships are necessarily causal; Figure 7.17: Share or refugees involved in a community Figure 7.18: Share or refugees engaged in a community representative body representative body by demographic group 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 Eritrean Somali South Sudanese Addis Ababa All Refugees Age Under 30 Age 30-44 Age 45-64 Age Over 64 Female Male Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. 70 Markets and Opportunities Low refugee social integration is not because social integration barriers refugees in Ethiopia they have a low willingness to engage with their face. In principle, the positive attitudes among many communities. Refugees have an extremely high rate hosts towards refugees, the high degree of cultural of involvement in refugee community representative similarity between groups, and the willingness of bodies. This includes participation in refugee central refugees to be engaged in their community are all committees, refugee outreach volunteers, refugee promising signs for social integration. Yet, even in community leaders, and leaders in women and youth the Somali domain, where cultural similarity and associations. Sixty percent of refugees participate host attitudes are greatest, more than two-thirds of in an organization like this, the lowest in Eritrean refugees do not have an Ethiopian friend, and more camps at 32 percent. On average, these rates are than half do not find social interactions with hosts similar across age and gender groups. easy. Better employment outcomes for refugees with Ethiopian friends indicate the benefits of facilitating Main immediate refugee integration challenges social integration for refugee livelihoods and are: (i) to expand involvement outside of refugee economic integration. communities, and (ii) to better understand the Figure 7.19: Discrimination and harassment Figure 7.20: Discrimination and harassment by demographic group 35 30 30 25 25 20 20 15 15 10 10 5 5 0 0 Eritrean Somali South Sudanese Addis Ababa All Refugees Age Under 30 Age 30 -44 Age 45 -64 Age Over 64 Female Male Discrimination/harassment in past year Victim of crime in past 2 weeks Discrimination/harassment in past year Victim of crime in past 2 weeks Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. 71 8. Policy Recommendations A ddressing the challenges refugees face in Ethiopia requires a concerted effort to promote their self-reliance, economic integration, and access unable to move to locations with better economic opportunities and require a work permit (which is difficult to get for work outside of refugee camps) to to education. By leveraging data from initiatives like legally access the labor market. As a result, refugees SESRE and adopting a comprehensive approach in Ethiopia remain poor and depend heavily on that considers the needs of both refugees and host humanitarian assistance. communities, Ethiopia can maximize the benefits of hosting refugees while minimizing associated costs. Concerted effort and policy interventions are necessary to better integrate refugees and The GoE has proven its strong commitment to improve the well-being of both refugees and host protecting refugees, but the progressive policy communities. The existing cultural and ethnic- framework has not yet translated into tangible based affiliation between refugees and their hosts socioeconomic outcomes for refugees. The is critical in facilitating and enhancing socio- encampment model previously followed in Ethiopia economic integration. As highlighted in Chapter 7, neglected how refugees affect socio-economic and the context for an integrated solution is favorable: environmental conditions of hosting communities, social cohesion is high, creating a supportive context including the untapped potential for refugees to for policy rollout. Sixty-five percent of hosts agree contribute to the local economy. Despite strong that refugees are friendly and good people, and improvements in Ethiopia’s underlying legal only 20 percent are uncomfortable with having a framework to benefit refugee inclusion, and a refugee neighbor. Moreover, the social acceptability strong international aid response, refugees still face for integrating refugees is high; almost half of all various challenges accessing services and improving hosts agree that refugees have increased economic their socioeconomic outcomes. Refugees are opportunities in Ethiopia. 72 Policy Recommendations Improvements can be achieved by focusing policy Promote improved refugee access to labor markets attention on three areas: to provide sustainable economic opportunities, (i) Providing refugees with a path to self-reliance. improved labor outcomes, and better prospects for long-term self-reliance (Muna, 2019). As highlighted Implementing place-based interventions to (ii) in Chapter 3, not all refugees have favorable alleviate the pressures for refugees and hosts. labor market outcomes and benefit from national (iii) Continuing to implement the progressive policy economic opportunities. In-camp refugees mainly framework for refugees. rely on assistance, have low employment rates, Path to self-reliance and few opportunities to generate income. Some Pursue development approaches that enable and refugees work outside camps but without work incentivize refugees’ self-reliance in Ethiopia to authorization. This limits wages and job security. improve refugees’ outcomes and reduce burdens Easing restrictions on access to the labor market on host communities. Host communities can outside of camps and accelerating and automating reap economic benefits from refugees’ presence. issuance of work authorizations will have lasting In Ethiopia, the path of self-reliance includes, at a effects in improving refugees’ livelihoods in minimum: (i) encouraging mobility to access areas camps. Given the importance of labor market with better economic opportunities, (ii) facilitating participation for self-reliance, efforts to strengthen labor market access for refugees by easing restrictions the human capital of refugees during displacement and providing work permits, (iii) integrating refugee can have large payoffs. Strengthening their skills, children into national education system to improve knowledge, and experience could enable them their long-term prospects, and (iv) strengthening to realize their potential and become productive inclusive healthcare systems to address the health members of society. needs of refugees. Build inclusive education systems. Integrating Encourage refugees to move where economic refugees into functioning national education opportunities are highest, which can also benefit systems can improve future outcomes for refugee local economies. As outlined in Chapter 6, denying children and their hosts (UNHCR, 2020; Piper et al., refugees mobility to settle where they would like 2020; Abu-Ghaida and Silva, 2020; Crawford et al., comes at a cost, as the choice of location within 2015; Bilgili et al., 2019). As Chapter 2 highlighted, the host country matters for refugees’ labor market more than half of all refugees are children under outcomes. Placing refugees in areas with lower the age of 15. Although over 70 percent of primary economic opportunities without the ability to move school age children attend primary education, makes it difficult for them to work (Azlor, Damm, and they do not make it past primary education. Schultz-Nielsen 2020; Eckert, Hejlesen, and Walsh Integrating refugee children into educational 2020; Fasani, Frattini, and Minale 2022). Therefore, programs soon after their arrival in Ethiopia development approaches that allow refugees to avoids the loss of valuable years of education move to areas with high economic potential can and human capital accumulation, hindering provide refugees with more job opportunities prospects. It is critical to address the obstacles that and boost demand in local economies. Increased hinder children from transitioning to secondary economic demand can pull (host) people out of schooling, such as challenges in accessing school agriculture and contribute to rural transformation, a records, language barriers, or distance to schools. prerequisite for achieving structural transformation Supporting Regional Education Bureaus could in Ethiopia. increase the accessibility of secondary schools 73 Policy Recommendations to camp refugees.51 To improve the educational community. Place-based interventions focus on the attainment of refugee children, the focus should unique characteristics, needs, and resources of the be on increasing the number of qualified teachers particular area; they leverage local assets to address in primary education, increasing the currently low local challenges with the active participation of compensation to incentive teachers with similar community members. In Ethiopia, investments in qualifications as nationals, improving primary-to- refugee hosting locations should benefit refugees secondary transition rates, and reducing classroom and hosts. Development partners and the GoE overcrowding. should align their development plans to expand opportunities for refugees and host communities Build an inclusive health system. Good health is an sustainably. Leveraging development resources essential requirement to rebuild refugees’ lives after to increase investment in refugee areas can displacement, but as highlighted in Chapter 2, refugee support social cohesion by demonstrating to children are particularly prone to stunting and other host communities that the presence of refugees nutritional challenges. Refugees have, as any other can create new livelihood opportunities for all population, varied healthcare needs, including local people. non-communicable diseases, infectious diseases, trauma from injuries, and violence. Research shows Direct more educational resources to refugee- that conflict has extensive psychological impacts hosting school woredas. Despite the positive on refugees, particularly youth and children, externalities of integrating refugees into the public- which are often not addressed (Simpson 2018; school systems, a large influx of refugee children Bosqui and Marshoud 2018; Dong 2018). Aligned can exacerbate challenges in local schools where with the GCR, refugees should be able to access refugees are hosted. Where there are large inflows, healthcare and essential health services through the national education system might require the national health systems of the destination additional resources to integrate newly-arrived countries at affordable costs and sufficient children. Increasing the supply and improving the quality. Services should consider the challenges quality of schools in affected areas, supported refugees face, such as lack of familiarity with by external assistance and financing, can avoid administrative procedures, uncertainty about tension that may arise over competition for access the future, and psychological distress. This to education services. Better coordination between requires strengthening and expanding service humanitarian and development actors can support delivery in the national health sector. This could, efforts to expand and strengthen national education for example, be achieved by increasing the systems to benefit all students. Support is particularly enrollment of refugees in the Community-Based required in remote areas—where refugees are often Health Insurance (CBHI) scheme. hosted—where educational service is strained, even for local children (Abu-Ghaida and Silva, 2020). Place-based intervention Pursue place-based development approaches Expand refugee access to social safety nets. The complementing regional development policies most vulnerable refugees and hosts may not be to benefit both refugees and hosts. Place-based able to reap benefits from better development interventions are strategies or programs that approaches. Social protection (SP) systems can address issues in a specific geographic location or alleviate pressures and safeguard against risks 51 It may be noted that, through the General Education Quality Improvement Program for Equity (GEQIP-E) mainly funded by the World Bank, the progressive transfer of refugee camps’ secondary schools from DICAC to Regional Education Bureaus has started. For instance, in Gambella region, secondary schools in Jewi and Pinyudo I refugee camps have been taken over by Gambella REB in September 2023 although Gambella REB still requires support to cover all existing needs in concerned schools (i.e., Gambella REB covers education and administrative personnel’s salaries but cannot afford additional construction/maintenance of these schools’ facilities, teachers’ transportation and accommodation, teaching and learning materials, etc.). 74 Policy Recommendations for vulnerable populations. Social protection Global Compact of Refugees and the CRRF. This encompasses a wide-ranging set of policies and includes establishing a system to track progress programs to protect people against poverty and risks regularly in implementing the government to their livelihoods and well-being. Implementing pledges. To realize Government commitments, place-based SP approaches that allowing the most close coordination among many stakeholders is vulnerable hosts and refugees to participate in vital, including between RRS, line ministries, and national programs—such as done under the Urban humanitarian and development actors at all levels Safety Net and Jobs Project—can support social (federal, regional, Woreda, and Kebele). It may be cohesion and integration. necessary to leverage development resources to accomplish this. Continue implementing the progressive policy framework regarding refugees Redesigning the OCP to encourage mobility to Implement concrete actions to realize pledges realize greater socioeconomic opportunities and proclamations. Action is needed to continue for refugees while accelerating and automating implementing Ethiopia’s progressive framework the issuance of work authorizations can enable in refugee inclusion, service integration, and right sustainable improvements in refugees’ lives. to work. These relate to transforming camps into The current system of work authorizations is not human settlements, which facilitate socio-economic implemented effectively. Though definitions for work opportunities for refugees to absorb the refugee permits have been improved, few work permits are camp into the local population, encouraging mobility issued, and restrictions to both wage-employment to achieve self-reliance, accelerating and automating and self-employment around the areas of work work authorization by virtue of status to engage exist. Accelerating and automating the issuance refugees in three avenues of job opportunities (joint of work authorizations can achieve sustainable projects, wage-employment, and self-employment), improvements in refugees’ lives. expanding possibilities for access to land, and improving secondary legislation. Reduce challenges refugees face in accessing business licenses for self-employment. Regarding Harmonize national and sub-national laws and self-employment, automate the existing procedural policies to support the full implementation of requirements that restrict refugees more than the the Refugee Proclamation. Although the Refugee most favorably treated foreign nationals, including Proclamation has provisions to protect refugees, a requirement for an investment permit subject to some enabling regulations and directives to capital requirements. Moreover, the lack of access facilitate full implementation of the GoE pledges to finance and lack of credit—from financial service are still lacking. The absence of these regulations is providers, including microfinance institutions, delaying the implementation of most of the rights which are not yet able to give credit to refugees— set out in the Refugee Proclamation. Secondary are among the key challenges refugees who want to legislation is still required to provide additional open businesses face. guidance on the meaning and scope of the rights granted, to harmonize relevant national and sub- Improved cooperation and coordination national laws and policies, and to clarify the roles In order to make commitments for sustained and responsibilities of government agencies in their support to refugees, the GoE needs to have implementation. predictable support and streams of resources. The GCR and the CRRF represent a significant step Better identify and document best practices and towards improving the current system by providing lessons to better coordinate and implement the a renewed architecture for collective action. The 75 Policy Recommendations GCR is underpinned by the principles of greater connections between RRS and line ministries could international solidarity and responsibility-sharing. support initiatives seeking to mainstream refugees Yet, current development approaches to support into existing governance structures. Encouraging refugees and their hosts in Ethiopia are still limited these collaborative efforts of departments and and lack specific mechanisms for sharing the agencies of the GoE can achieve a successful responsibility of hosting refugees more equitably. implementation of development solutions. Humanitarian and development actors need Efforts to improve the coverage, accuracy, to swiftly invest in and accelerate inclusive reliability, quality, and comparability of data can approaches. Mobility to accelerate economic provide the analytical underpinning for policy opportunities and self-reliance can support the decisions. Better data enables better planning GoE in implementing the Refugee Proclamation of (and decisions). Integrating refugees as part of the 2019, while encampment undermines achieving national household survey system could provide the goals set out in the Proclamation. The GoE high-quality data on a regular basis. This would pledged to transform camps into settlements and include the ability to disaggregate data to a subset facilitate mobility for refugees to take advantage of the population surveyed and compare refugees of opportunities in the labor market and increase with other population groups. This requires work authorization to allow for the formalization of strengthening data collection and dissemination working conditions. Humanitarian and development mechanisms at all levels. The GoE pledged to partners should strongly support this pledge by include refugee data in national statistics. This swiftly investing in and accelerating inclusive would ensure that systems are systematically approaches through the engagement of line built to serve all people in a particular “place” ministries. Yet, large gaps in financing remain to fill regardless of status. This includes the need for a the needs of refugees and host communities. full population count (including refugees) across Ethiopia’s territory to inform decisions such as the Better coordination and engaging line ministries size of schools to ensure progress toward inclusive can achieve better outcomes for refugees systems that support refugees and their hosts can be and their hosts. Implementing an overarching made. Strengthening the use of statistics includes coordination mechanism across line ministries to facilitating access to data and disseminating results. track investments and progress on refugee inclusion The SESRE is an excellent start to this initiative. Yet, could leverage the existing humanitarian resources the need to systematically integrate refugees in to deliver the first mile investment into inclusive every round of the national household surveys and development approaches, led by development other data collection activities is key to allowing for actors. Improved communication, collaboration, and evidence-based policy making. 76 REFERENCES Abdelhady D., and Al Ariss, A. (2023). How Capital Shapes Refugees’ Access to the Labour Market: The Case of Syrians in Sweden. The International Journal of Human Resource Management, 34(16), 3144-3168, DOI: 10.1080/09585192.2022.2110845 Abu-Ghaida, D., and Silva, K. (2020). Forced Displacement and Educational Outcomes: Evidence, Innovations, and Policy Indications. Second issue. Quarterly Digest on Forced Displacement. Washington, D.C.: World Bank Group, UNHCR and JDC. Adda, J., Dustmann, C., and Görlach, J. S. (2022). The Dynamics of Return Migration, Human Capital Accumulation, and Wage Assimilation. The Review of Economic Studies, 89(6), 2841-2871. Alix-Garcia, J., Walker, S., Bartlett, A., Onder, H., and Sanghi, A. (2018). Do Refugee Camps Help or Hurt Hosts? The Case of Kakuma, Kenya. Journal of Development Economics, 130, 66-83. Alkire, S., Kanagaratman, U., and Suppa, N. (2021). The Global Multidimensional Poverty Index (MPI) 2021, OPHI MPI Methodological Note 51. Oxford Poverty and Human Development Initiative, University of Oxford. Altindag, O., Bakls, O., and Rozo, M. S. (2020). Blessing or Burden? Impacts of Refugees on Businesses and the Informal Economy. Journal of Development Economics, 146. https://doi.org/10.1016/j.jdeveco.2020.102490 Andersen, H.L., Osland, L. and Zhang, M.L. (2023). Labour Market Integration of Refugees and The Importance of the Neighborhood: Norwegian Quasi-experimental Evidence. Journal of Labour Market Research 57, 16. https://doi.org/10.1186/ s12651-023-00341-y Aracl, D., Demirci, M., and Klrdar, M. (2022). Development Level of Hosting Areas and the Impact of Refugees on Natives’ Labor Market Outcomes in Turkey. European Economic Review, 145. https://doi.org/10.1016/j.euroecorev.2022.104132 Atamanov, A., Hoogeveen, J., and Reese, B. (2023). The Costs Come Before the Benefits Follow. Why Should Donors Invest More in Refugee Autonomy in Uganda? 1–11. Azlor, L., Damm, A. P., and Schultz-Nielsen, M. L. (2020). Local Labor Demand and Immigrant Employment. Labour Economics, 63, 101808. https://doi.org/10.1016/j.labeco.2020.101808. Bahar, D., and Dooley, M. (2019). No Refugees and Migrants Left Behind. In H. Kharas, J. McArthur and Ohno, I. (Eds.), Leave No One Behind: Time for Specifics on the Sustainable Development Goals, 79-104. Washington, DC: Brookings Institution Press. Balcilar M., and Nugent J. B. (2019). The Migration of Fear: An Analysis of Migration Choices of Syrian Refugees. The Quarterly Review of Economics and Finance 73, 95- 110. https://doi.org/10.1016/j.qref.2018.09.007 Barron, P., Cord, L., Cuesta, J., Espinoza, S., and Woolcock, M. (2023). Social Sustainability in Development: Meeting the Challenges of the 21st Century. World Bank Publications. Baseler, T., Ginn, T., Hakiza, R., Ogude, H., and Woldemikae, O. (2021). Can Aid Change Attitudes Toward Refugees? Experimental Evidence from Microentrepreneurs in Urban Uganda. Betts, A., Bloom, L., Kaplan, J. D., and Omata, N. (2014). Refugee Economies: Rethinking Popular Assumptions. Oxford: University of Oxford, Refugee Studies Centre. Betts, A., Fryszer, L., Omata, N., and Sterck, O. (2019). Refugee Economies in Addis Ababa: Towards Sustainable Opportunities for Urban Communities? Oxford: University of Oxford, Refugee Studies Centre. Betts, A., Marden, A., Bradenbrink, R., and Kaufmann, J. (2020). Building Refugee Economies: An Evaluation of the IKEA Foundation’s Programmes in Dollo Ado. Oxford: University of Oxford, Refugee Studies Centre. Betts, A., Stierna, M. F., Omata, N., and Sterck, O. (2023). Refugees Welcome? Inter-Group Interaction and Host Community Attitude Formation. World Development, 161, 106088. Bevelander, P., and Lundh, C. (2007). Employment Integration of Refugees: The Influence of Local Factors on Refugee Job Opportunities in Sweden. IZA Discussion Paper No. 2551. Bevelander, P. (2020). Integrating Refugees into Labor Markets. IZA World of Labor, 269v2. https://doi.org/10.15185/ izawol.269.v2 Bilgili, Ö., Loschmann,C., Fransen, S., and Siegel, M. (2019). Is the Education of Local Children Influenced by Living near a Refugee Camp? Evidence from Host Communities in Rwanda.” International Migration, 57 (4), 291 - 309. https://doi. org/10.1111/imig.12541. Blumenberg, E., and Manville, M. (2004). Beyond the Spatial Mismatch: Welfare Recipients and Transportation Policy. Journal of Planning Literature, 19(2), 182–205. https://doi.org/10.1177/0885412204269103 Boda, Z., Lorenz, G., Jansen, M., Stanat, P., and Edele, A. (2023). Ethnic Diversity Foster the Social Integration of Refugee Students. Nature Human Behaviour, 1-11. Bosqui, T. J., and Marshoud, B. (2018). Mechanisms of Change for Interventions Aimed at Improving the Wellbeing, Mental Health and Resilience of Children and Adolescents Affected by War and Armed Conflict: A Systematic Review of Reviews. Conflict and Health 12 (1), 15. https://doi.org/10.1186/s13031-018-0153-1. Brell, C., Dustmann, C., and Preston, I. (2020). The Labor Market Integration of Refugee Migrants in High-Income Countries. Journal of Economic Perspectives, 34(1), 94-121. Clements, K., Shoffner, T., and Zamore, L. (2016). Uganda's Approach to Refugee Self-reliance. Forced Migration Review, 52. 77 References Cobb-Clark, D. A., Kassenboehmer, S. C., and Sinning, M. G. (2016). Locus of Control and Savings. Journal of Banking and Finance, 73, 113-130. Crawford, N., Cosgrave, J., Haysom, S., and Walicki, N. (2015). Protracted Displacement: Uncertain Paths to Self-Reliance in Exile. Humanitarian Policy Group Commission Report, Overseas Development Institute. Dercon, S., Hoddinott, J., and Woldehanna, T. (2005). Shocks and Consumption in 15 Ethiopian Villages, 1999-2004. Journal of African Economies, 14(4), 559–585. https://doi.org/10.1093/jae/eji022 Donaldson, D., and Hornbeck, R. (2016). Railroads and American Economic Growth: A “Market Access” Approach. The Quarterly Journal of Economics, 131 (2),2, 799 - 858. https://doi.org/10.1093/qje/qjw002 Dong, .D (2018). Mental Health and Psychosocial Support in Fragile, Conflict, and Violence Situations: Five Key Questions to Be Answered. FCV Health Knowledge Notes. Washington, D.C.: World Bank. https://openknowledge.worldbank.org/ handle/10986/30387. Dorian, T., and Burmann, M. (2023). Making the Match. The Importance of Local Labor Markets for the Employment Prospects of Refugees. Social Sciences, 12 (6), 339. Dustmann, C., Vasiljeva, K., and Piil Damm, A. (2019). Refugee Migration and Electoral Outcomes. The Review of Economic Studies, 86(5), 2035-2091. Easterly, W., Ritzen, J., and Woolcock, M. (2006). Social cohesion, institutions, and growth. Economics and Politics, 18(2), 103-120. Eaton, J., and Kortum, S. (2002). Technology, Geography, and Trade. Econometrica, 70(5), 1741–1779. http://www.jstor.org/ stable/3082019 Eckert, F., Hejlesen, M., and Walsh, C. (2022). The Return to Big-city Experience: Evidence from Refugees in Denmark. Journal of Urban Economics, 130, 103454. European Union Agency for Asylum (EUAA) (2023). Asylum Report 2023: Annual Report on the Situation of Asylum in the European Union. Luxembourg: Publications Office of the European Union. Fasani, F., Frattini, T., and Minale, L. (2022). (The Struggle for) Refugee integration into the Labour Market: European Evidence. Journal of Economic Geography, 22(2), 351-393. Feyissa, D. (2015). Power and Its Discontents: Anywaa’s Reactions to The Expansion of The Ethiopia Statae, 1950 – 1991. International Journal of African Historical Studies, 48(1), 31 - 49 FDRE. (2004). Refugee Proclamation No. 409/2004. FDRE. (2019). Refugee Proclamation No. 1100/2019. FDRE. (2020). Investment Proclamation No. 1180/2020. Fuller, S. (2015). Do Pathways Matter? Linking Early Immigrant Employment Sequences and Later Economic Outcomes: Evidence from Canada. International Migration Review, 49(2), 355-405. Gebreegziabher, T., and Regassa, N. (2019). Ethiopia’s High Childhood Undernutrition Explained: Analysis of the prevalence and Key Correlates based on Recent Nationally Representative Data. Public Health Nutrition, 22(11), 2099–2109. .org/10.1017/ S1368980019000569 Gebru, F. K., Haileselassie, M. W., Temesgen, H. A., Seid, O. A., and Mulugeta, A. B. (2019). Determinants of Stunting among under-five Children in Ethiopia: A Multilevel Mixed-effects Analysis of 2016 Ethiopian Demographic and Health Survey Data. In BMC Pediatrics 19 (1). BioMed Central Ltd. .org/10.1186/s12887-019-1545-0 Ginn, T., Resstack, R., Dempster, H., Arnold-Femandez, E., Miller, S., Guerrero, M., and Kanyamanza, B. (2022). 2022 Global Refugee Work Rights Report. Center for Global Development, Refugees International, and Asylum Access. Ginn, T. (2023). Labor Market Access and Outcomes for Refugees. World Bank – UNHCR Joint Data Center. Quarterly Digest on Forced Displacement, Seventh Issue. Washington, D.C.: World Bank Group. Giordano, N., Ercolano, F., and Makhoul, M. (2021). Skills and Labour Market Transitions for Refugees and Host Communities: Case Studies and Country Practices on Including Refugees in Technical and Vocational Education and Training (TVET) and Employment. UNHCR Gobillon, L. and Selod, H. (2014). Spatial Mismatch, Poverty, and Vulnerable Populations. In: Fischer, M.M., Nijkamp, P. (eds) Handbook of Regional Science. Springer, pp.93-107. https://doi.org/10.1007/978-3-642-23430-9 Hagos, S. Z. (2021). Refugees and Local Power Dynamics: The Case of The Gambella Region of Ethiopia. Hahn, E., Richter, D., Schupp, J., and Back, M. D. (2019). Predictors of Refugee Adjustment: The Importance of Cognitive Skills and Personality. Collabra: Psychology, 5(1), 23. Hainmueller, J., and Hopkins, D. J. (2014). Public Attitudes Towards Immigration. Annual Review of Political Science, 17, 225-249. Hainmueller, J., Hangartner, D., and Lawrence, D. (2016). When Lives Are Put on Hold: Lengthy Asylum Processes Decrease Employment among Refugees. Science Advances, 2. https://doi.org/10.1126/sciadv.1600432. Hanafi, L., Johns, M., Lacroix, R., Khadija Shaikh, Casaly, P., Aeschlimann, A., and Willy, A. (2021). Refugee Policy Review Framework. Washington, D.C.: World Bank. Heckman, J. J., Stixrud, J., and Urzua, S. (2006). Cognitive and Noncognitive Abilities Affect Labor Market Outcomes and Social Behavior. Journal of Labor Economics, 24(3), 411-482. Hedberg, C. and Tammaru, T. (2012). ‘Neighbourhood Effects’ and ‘City Effects’: The Entry of Newly Arrived Immigrants into the Labour Market. Urban Studies, 50(6). https://doi.org/10.1177/0042098012461674 78 References Holzaepfel, E. A. (2015). Evaluating the Effectiveness of Livelihoods Programs for Refugees in Ethiopia. Washington, D.C: United States Department of State. Hvidtfeldt, C., Schultz-Nielsen, M. L., Tekin, E., and Fosgerau, M. (2018). An Estimate of the Effect of Waiting Time in the Danish Asylum System on Post-Resettlement Employment among Refugees: Separating the Pure Delay Effect from the Effects of the Conditions under Which Refugees Are Waiting. Edited by Bernardo, L. Q. PLOS ONE 13 (11), e0206737. https:// doi.org/10.1371/journal.pone.0206737. Ibáñez, A.M., Moya, A., Adelaida Ortega, M., Rozo, S.V, and Urbina, M.J. (2022). Life Out of the Shadows: Impacts of Amnesties in the Lives of Migrants. Working Paper 15049. IZA Discussion Papers. https://www.econstor.eu/handle/10419/250710. Immergluck, D. (1998). Job Proximity and the Urban Employment Problem: Do Suitable Nearby Jobs Improve Neighborhood Employment Rates? A Reply. Urban Studies, 35(12), 2359-2368. https://doi.org/10.1080/0042098983926 IOM. (2023). A Region on the Move 2022: East and Horn of Africa. IOM, Nairobi. IPCC. (2019). Special Report on the Ocean and Cryosphere in a Changing Climate. Retrieved from https://www.ipcc.ch/srocc/ Kalter, F., and Kogan, I. (2014). Migrant Network and Labor Market Integration of Immigrants from the Former Soviet Union in Germany. Social Forces, 92(4), 1435 – 1456. https://www.jstor.org/stable/43287575 Kalter, F., and Kogan, I. (2020). An Empirical-Analytical Approach to the Study of Recent Refugee Migrants in Germany. Social Forces, 71 (1-2), 3-23. https://doi.org/10.5771/0038-6073-2020-1-2-3 Kassa, T., Mulatu, F., & Iddossa, J. (2019). Ethiopia’s Refugee Policy Overhaul: Implications on the Out-of-camp Policy Regime and Rights to Residence, Movement and Engagement in Gainful Employment. Ethiopian Journal of Human Rights, 4(1), 124–186. Kneebone, E., and Holmes, N. (2015). The Growing Distance Between People and Jobs in Metropolitan America. Metropolitan Policy Program. Brookings. Krause, U., and Schmidt, H. (2020). Refugees as Actors? Critical Reflections on Global Refugee Policies on Self-Reliance and Resilience. Journal of Refugee Studies, 33(1), 22-41. Kvittingen A., Valenta M., Tabbara H., Baslan D., and Berg B. (2018). The Conditions and Migratory Aspirations of Syrian and Iraqi Refugees in Jordan. Journal of Refugee Studies 32(1), 106-124. https://doi.org/10.1093/jrs/fey015. Lebow, J. (2023). Immigration and Occupational Downgrading in Colombia. Journal of Development Economics, Forthcoming Levenson, H. (1981). Differentiating among Internality, Powerful Others, and Chance. Research with the Locus of Control Construct, 1, 15-63. Liu, D. and Kwan, M. (2020). Measuring Spatial Mismatch and Job Access Inequity Based on Transit-Based Job Accessibility for Poor Job Seekers. Travel Behaviour and Society, 19, 184-193. https://doi.org/10.1016/j.tbs.2020.01.005 ESS and World Bank. (2023). Ethiopia Socioeconomic Panel Survey Report – Wave 2, 2021/22. Marbach, M., Hainmueller, J., Hangartner, D. (2018). The Long-Term Impact of Employment Bans on the Economic Integration of Refugees. Science Advances, 4. https://doi.org/10.1126/sciadv.aap9519. Martin, S. F. (2010). The Causes and Consequences of Forced Migration. Paper presented at the Annual Meeting of the Population Association of America, Dallas, TX. Maystadt, J. F., and Verwimp, P. (2014). Winners and Losers among a Refugee-Hosting Population. Economic Development and Cultural Change, 62(4), 769-809. MoE. (2022). Education Statistics Annual Abstract (ESAA). MoE MoLSA. (2019). Revised Directives on the Issuance of Work Permits for Expatriates in Ethiopia. MoTRI, MoLS , MoR and RRS. (2023). Memorandum of Understanding (MoU) to allow recognized asylum seekers and refugees to engaged in formal business. MoLS and RRS. (2023). MoU on technical and vocational training, work permit, job creation and livelihood improvement, and related issues for refugees. MoLS and RRS. (2023). Technical and vocational training for refugees, job creation and livelihood improvement, work permit and related issues service delivery manual (Standard Operating Procedure). Muna, S. (2019). In Pursuit of Self-reliance: Perspectives of Refugees in Jordan. Archnet-IJAR, 13(3), 612-626. https://doi. org/10.1108/ARCH-04-2019-0085 Nguyen, N. T. V., Savadogo, A., and Tanaka, T. (2021). Refugees in Chad: The Road Forward. The World Bank. Norman, T., Borjesson, M., and Anderstig, C. (2017). Labor Market Accessibility and Unemployment. Journal of Transport Economics and Policy. Pape, U. J., Petrini, B., and Iqbal, S. A. (2018). Informing Durable Solutions by Micro-Data: A Skills Survey for Refugees in Ethiopia. Peters, S. and Golden, S. (2019). Assessing Mental Health in Gambella, Ethiopia: A Representative Survey of South Sudanese Refugees in Nguenyyiel Camp. St. Paul, MN: The Center for Victims of Trauma. Pimhidzai, O., Chigumira, E., Tesfaye,W., and Yonis, M. (2022). Ethiopia - Rural Income Diagnostics Study: Leveraging the Transformation in the Agri-Food System and Global Trade to Expand Rural Incomes. Washington, D.C.: World Bank Group. https://documentsinternal.worldbank.org/search/33891468 Piper, B., Dryden-Peterson, S., Chopra, V., Reddick, C., and Oyanga, A. (2020). Are Refugee Children Learning? Early Grade Literacy in a Refugee Camp in Kenya. The Journal on Education in Emergencies, Inter-agency Network for Education in 79 References Emergencies, 5(2). https://doi.org/10.33682/f1wr-yk6y. Ravallion, M. (1998). Poverty Lines in Theory and Practice: Living Standards Measurement Study. LSMS Working Paper, Issue 133. https://documents1.worldbank.org/curated/en/916871468766156239/pdf ReDSS. (2018). Local Integration Focus: Refugees in Ethiopia - Gaps and Opportunities for Refugees Who have Lived in Ethiopia for 20 Years or More. ReDSS. Rotter, J. B. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological monographs: General and applied, 80(1), 1. RRS. (2017). Road Map for Implementing the Federal Democratic Republic of Ethiopia Government Pledges and the Practical Application of the CRRF in Ethiopia. RRS RRS. (2019). Directive to Determine Conditions for Movement and Residence of Refugees Outside Camps, Directive No. 01/2019. RRS RRS. (2019a). Directive to Determine the Procedure for Refugees Right to Work. Directive No. 02/2019. RRS. (2019b). Refugees and Returnees Grievances and Appeals Handling Directive. Directive No. 03/2019. RRS, and UNHCR. (2021). Ethiopia GRF Pledge Progress Report. RRS and UNHCR Shamsuddin, M., Acosta, P.A., Schwengber, R.D., Fix, J., and Pirani, N. (2021). Integration of Venezuelan Refugees and Migrants in Brazil. Policy Research Working Papers 9605. http://hdl.handle.net/10986/35358. Schuettler, K., and Caron, L. (2020). Jobs Interventions for Refugees and Internally Displaced Persons. Jobs Working Paper No. 47. Washington, D.C: World Bank Group. Schlechter, P., Hellmann, J. H., and Morina, N. (2023). Self-Efficacy and Locus of Control as Transdiagnostic Factors in Middle Eastern Refugees. European Journal of Psychotraumatology, 14(1), 2180707. Simposon, R. (2018). Peace Education and Psychosocial Support for Social Cohesion. Forced Migration Review, 3. Slotwinski, M., Stutzer, A., and Uhlig, R. (2019). Are Asylum Seekers More Likely to Work with More Inclusive Labor Market Access Regulations? Swiss Journal of Economics and Statistics, 155, 1-15. Steinmayr, A. (2021). Contact Versus Exposure: Refugee Presence and Voting for the far Right. Review of Economics and Statistics, 103(2), 310-327. Tabellini, M. (2020). Gifts of the Immigrants, Woes of the Natives: Lessons from the Age of Mass Migration. The Review of Economic Studies, 87(1), 454-486. Taffesse, A. S., and Tadesse, F. (2017). Pathways Less Explore: Locus of Control and Technology Adoption. Journal of African Economies, 26(suppl_1), i36-i72. Tasic, H., Akseer, N., Gebreyesus, S. H., Ataullahjan, A., Brar, S., Confreda, E., Conway, K., Endris, B. S., Islam, M., Keats, E., Mohammedsanni, A., Wigle, J., and Bhutta, Z. A. (2020). Drivers of Stunting Reduction in Ethiopia: A Country Case Study. American Journal of Clinical Nutrition, 112, 875S-893S. https://doi.org/10.1093/ajcn/nqaa163 Thum, A. E. (2014). Labor Market Integration of German Immigrants and their Children: Does Personality Matter? Thompson, A. M. (1997). The Impact of Spatial Mismatch on Female Labor Force Participation. Economic Development Quarterly, 11(2), 138–145. https://doi.org/10.1177/089124249701100203 Tsionis, A., Pantoglou, D., and Kasvikis, Y. (2022). Mental Health Locus of Control in Refugees With Clinically Established Psychopathology. Psychiatrike= Psychiatriki. Tsolak, D., and Bürmann, M. (2023). Making the Match: The Importance of Local Labor Markets for the Employment Prospects of Refugees. Social Sciences, 12,339. https://doi.org/10.3390/socsci12060339 Tumen, S. (2019). Refugees and ‘Native Flight’ from Public to Private Schools. Economics Letters, 154 - 159. https://doi. org/10.1016/j.econlet.2019.05.030. Tumen, S. (2021). The Effect of Refugees on Native Adolescents’ Test Scores: Quasi-Experimental Evidence from PISA. Journal of Development Economics, 150, 102633. https://doi.org/10.1016/j.jdeveco.2021.102633. UNHCR. (2020). Coming Together for Refugee Education. Education Report 2020. UNHCR. (2020a). Ethiopia Refugee Education Strategy: Towards inclusion. UNHCR. (2020b). Ethiopia Fact Sheet: September 2020. UNHCR. (2020c). Ethiopia 2019 Summary Pledge Progress Report. UNHCR. (2020d). Global Trends Forced Displacement in 2019. Retrieved from https://www.unhcr.org/5ee200e37.pdf UNHCR. (2021). Global Compact on Refugees Indicator Report 2021. UNHCR. (2022). Refugee Settlement Profile: Alemwach. UNHCR. (2022a). Ethiopia Annual Results Report 2022. UNHCR. (2022b). Global Trends: Forced Displacement in 2021. https://www.unhcr.org/publications/brochures/62a9d1494/ global-trends-report-2021.html. UNHCR. (2022c). Ethiopia Country Refugee Response Plan (ECRRP) Jan Dec 2022. https://data.unhcr.org/en/documents/ details/94099 UNHCR. (2022d). Serdo Refugee Camp Profile April 2022. https://data.unhcr.org/fr/documents/details/92436 UNHCR. (2022e). “Global Report 2021.” http://reporting.unhcr.org/globalreport2021/pdf. 80 References UNHCR. (2023). Ethiopia Global Refugee Forum Pledge Progress Report. UNHCR. (2023a). Addis Ababa Quarterly Urban Factsheet, April 2023. UNHCR. (2023b). UNHCR Resettlement Handbook. UNHCR. (2023c). Culture, Context and Mental Health and Psychological Well-being of Refugees and Internally Displaced Persons from South Sudan. UNHCR. (2023d). Mid-year Trends. Geneva. https://www.unhcr.org/mid-year-trends-report-2023 UNHCR. (2023e). Ethiopia Country Refugee Response Plan 2023. https://reporting.unhcr.org/ethiopia-country-refugee- response-plan-summary UNHCR. (2024). Global Refugee Forum 2023 Pledges. https://globalcompactrefugees.org/pledges-contributions UNHCR. (2024a). Press Release (Ethiopia Launches Inclusive ID System for Refugees, Boosts Access to National Services). https://www.unhcr.org/africa/news/press-releases/ethiopia-launches-inclusive-id-system-refugees-boosts-access- national-services UNHCR, and World Bank Group. (2020). Understanding the Socioeconomic Conditions of Refugees in Kenya Volume A: Kalobeyei Settlement. Results from the 2018 Kalobeyei Socioeconomic Survey. Washington, D.C. UNICEF. (2018). Situation and Access to Services of Persons with Disabilities in Addis Ababa. Briefing Note. UNICEF. (2019). Birth Registration for Every Child by 2030: Are WE on Track? New York. UNICEF. (2021). Education in South Sudan. Briefing Note. Vemuru, V., Sarkar, A., and Woodhouse, A.F. (2020). Impact of Refugees on Hosting Communities in Ethiopia: A Social Analysis. World Bank Group. Verme, P., and Schuettler, K. (2021). The Impact of Forced Displacement on Host Communities: A Review of the Empirical Literature in Economics. Journal of Development Economics, 150, 102606. Walelign, S. Z., Wang Sonne, S. E., and Seshan, G. (2022). Livelihood Impacts of Refugees on Host Communities. Wieser, C., Dampha,N.K., Ambel, A.A., Tsegay, A.H., Mugera, H.K., Tanner, J. (2020). Monitoring COVID-19 Impact on Refugees in Ethiopia : Results from a High-Frequency Phone Survey of Refugees (English). Monitoring COVID-19 Impact on Refugees in Ethiopia Washington, D.C.: World Bank Group. Whitaker B.E. (2023). Border Proximity and Attitudes Toward Free Movement in Africa. The Afro Barometer Working Papers No. 200. Woldehanna, T., Hoddinott, J., and Dercon, S. (2008). Poverty and Inequality in Ethiopia:1995/96 – 2004/05. May. World Bank (2017). Forcibly Displaced: Toward a Development Approach Supporting Refugees, the Internally Displaced, and Their Hosts. Washington, D.C: World Bank. https://doi.org/10.1596/978-1-4648-0938-5. World Bank. (2019). Informing the Refugee Policy Response in Uganda. Results from the Uganda Refugee and Host Communities 2018 Household Survey. World Bank. (2019b). Better Opportunities for All: Vietnam Poverty and Shared Prosperity Update Report. Washington DC: World Bank Group. World Bank (2020). Ethiopia Regional Poverty Report: Promoting Equitable Growth for All Regions. Washington, D.C.: World Bank Group. World Bank (2020b). Ethiopia Poverty Assessment: Harnessing Continued Growth for Accelerated Poverty Reduction. Washington, D.C.: World Bank Group. World Bank (2023). World Development Report 2023: Migrants, Refugees, and Societies. Washington, D.C: World Bank Group. https://doi:10.1596/978-1-4648-1941-4 World Bank. (2023a). Social Cohesion and Forced Displacement: A Synthesis of New Research. World Bank. (2023b). Welfare in Forcibly Displaced Populations: From Measuring Outcomes to Building Capabilities. Leveraging Harmonized Data to Improve Welfare among Forcibly Displaced Populations and their Hosts: A Technical Brief Series. World Bank. (2023c). Do Legal Restrictions Affect Refugees’ Labor Market and Education Outcomes? Evidence from Harmonized Data World Bank. (2023d). Ethiopia Economic Opportunities Program: Aide Memoire. World Bank Group. 2021. “IDA19 Mid-Term Refugee Policy Review.” Washington, D.C.: World Bank Group. Zanfrini L., and Giuliani, C. (2023). Look at Me, but Better: The Experience of Young NEET Migrant Women between Vulnerability and Stifled Ambitions. Social Sciences 12, 110. https://doi.org/10.3390/socsci12020110. Zetter R., and Ruaudel, H. (2016). Refugees’ Right to Work and Access to Labor Markets – An Assessment, Part 1. World Bank Global Program on Forced Displacement (GPFD) and the Global Knowledge Partnership on Migration and Development (KNOMAD) Thematic Working Group on Forced Migration. http://bit.ly/KNOMAD-Zetter-Ruaudel-2016-1 Zhou, Y. Y., Grossman, G., and Ge, S. (2022). Inclusive Refugee Hosting in Uganda Improves Local Development and Prevents Public Backlash. 81 ANNEXES Annexes Annex A: Description of Refugees by Country of Origin E thiopia hosts refugees from some 24 countries. By far the largest groups are refugees from South Sudan, Somalia, and Eritrea. Each of these groups population growth. In the case of the Dollo Ado area (Liben Zone), some new arrivals are anticipated due to the security situations and the prevalence is described below, including a short description of of climate-change-induced drought in Somalia. host communities around the camps where refugees Some of those residing in the Jigjiga area have been are hosted. assisted in Ethiopia for over three decades, while the majority of individuals in Dollo Ado have been in the South Sudanese Refugees region for eight years. Following the outbreak of hostilities in parts of South Sudan in December 2013, a massive influx of refugees In the Ethiopian Somali Region, the armed group in Ethiopia led to the establishment of new refugee of Al-Shabab based in Somalia perpetrated several camps. The South Sudanese are the largest refugee attacks in Afder, Liben, and Shabelle zones in July population in Ethiopia. Currently, South Sudanese 2022. The attacks prompted all aid partners to refugees are sheltered in nine camps located in suspend movements and operations along the the Gambella (seven camps) and Benishangul- affected areas temporarily affecting the drought Gumuz (two camps) regions of the country. Two response in the region. refugee camps in the Benishangul-Gumuz region, namely Tongo and Gure-Shombola, were impacted Even with sluggish implementation, the by the clashes in the region, and the refugees were Intergovernmental Authority on Development relocated to another camp called, Tsore, situated in (IGAD) Special Summit on Durable Solutions for the same region. Somali Refugees and Reintegration of Returnees in Somalia, the related Nairobi Declaration, and the In 2023, Ethiopia hosted 416,660 registered South accompanying Plan of Action are still expected to Sudanese refugees, which is as high as the local provide impetus for delivering durable solutions. The population in the Gambella region and the largest Nairobi Declaration is a declaration by the Heads of refugee population in the country (UNHCR, 2023d). State and Government of the IGAD Region on durable Despite ongoing international and regional peace solutions for Somali refugees and reintegration of efforts, including by the South Sudanese factions, returnees in Somalia adopted in Nairobi, Kenya on Ethiopia continues to receive new arrivals from the 25 March 2017 at the Special Summit on protection country, mainly in dire need of food assistance, and durable solutions for Somali refugees and indicating limited opportunities for voluntary return reintegration of returnees in Somalia. and reintegration. Eritrean Refugees Somali Refugees Since 2000, Ethiopia has received and hosted The Somali refugee inflow to Ethiopia started in early thousands of Eritrean refugees fleeing persecution. 1990s, with people seeking safety and protection. As of 2023, Ethiopia hosted 165,793 registered As of 2023, Ethiopia hosted 283,111 refugees from Eritrean refugees (UNHCR, 2023d) in six camps, Somalia who were forced to flee their homes as a and under the out-of-camp policy in Addis Ababa. result of insecurity, political instability, conflict, The five refugee camps are located in Tigray (two), and famine (UNHCR, 2023d). The Somali refugee Afar (three), and Amhara (one) regions of the population is currently supported in two Zones in country. Unlike other refugee groups, many Eritrean the Somali Region: Fafan Zone (three camps) and refugees leave their camps due to various pull and Liben Zone (five camps). The population of Somali push factors to pursue onward movement to urban refugees in the Jigjiga area (Fafan Zone) is expected centers within Ethiopia, including Addis Ababa, and to increase modestly mainly because of natural other countries, primarily Europe. 83 Annexes Fighting initially broke out in the Tigray region of settlements. Most refugee-hosting areas are found Ethiopia in November 2020 between Tigrayan forces in remote locations bordering major refugee- and the Federal Government. Two refugee camps in producing countries such as South Sudan, Somalia, Tigray (Hitsats and Shimelba) were destroyed due Eritrea, and Sudan. The great majority of the refugee to the conflict in November 2020. The refugees who hosting communities not only share common socio- previously resided in these camps were relocated economic practices but also have similar cultures and to Mai-Ayni and Adi-Harush refugee camps in the ethnicities with refugees from neighboring countries. region as well as to Addis Ababa. A new refugee Despite cultural and ethnic-based commonalities, site (Alemwach) was established in June 2021 in the overall level of socioeconomic integration the Northern Gondar Zone, Dabat Woreda of the between refugees and host communities varies Amhara region to shelter Eritrean refugees relocated across the refugee-hosting areas in the country due from Mai-Ayni and Adi-Harush refugee camps. The to several factors, including historical and resource- spreading of the conflict in Northern Ethiopia into related tensions and perceptions towards refugees. the Afar region also caused the destruction of the Berhale refugee camp, previously hosting 20,639 In most of the refugee hosting regions, there Eritrean refugees in the Afar region (UNHCR, 2022d). is a huge competition over the meager natural The refugees who fled Berhale and surrounding areas resources between refugees and host communities, were relocated to a new refugee site called Serdo, 40 which not only depletes the resource base but kilometers from the regional capital Semera. sometimes results in local-level conflicts. For instance, Gambella's social and political context is Sudanese Refugees exceptionally complicated due to a long history of The arrival of Sudanese refugees in Ethiopia started conflict among groups over land and political power. in 1997, and their number has significantly increased The presence of refugees is a significant component in 2011. Since 2011, the conflict in the Blue Nile State of these dynamics. Host communities access some has forced many Sudanese to flee to Ethiopia. As services provided within the refugee camps. In some of 2023, Ethiopia hosts 48,709 registered refugees operational areas, refugees have better access to from Sudan (UNHCR, 2023d), who are assisted in basic and social services than host communities. four camps in the Benishangul-Gumuz Region. While Relationships between refugees and hosts are some refugees have resided in Sherkole camp for generally largely amicable, except in Gambella. more than two decades, the majority of the Sudanese In Benishangul-Gumuz and in the Somali Region, refugee population has recently arrived and sheltered incidences of community-wide violent conflict in the other three camps. In addition to refugees with between refugees and hosts are rare. A long history Sudanese nationality, refugees from the African Great of displacement, shared ethnic identity, and shared Lakes region are also sheltered in Sherkole refugee cultural ties, along with other structural factors have camp in the Benishangul-Gumuz region. fostered some solidarity between the groups. Refugees in the Benishangul-Gumuz region have Among the major refugee-hosting regions, four also experienced the effects of internal conflict. Two regions—Afar, Benishangul-Gumuz, Gambella, and refugee camps (Tongo and Gure-Shobola) became Somali—are designated as “emerging regions,” and inaccessible to humanitarian actors as a result of Tigray is considered post-conflict. These regions attacks by armed groups, forcing refugees to self- are the least developed regions in the country, relocate and to be relocated to Tsore refugee camp, characterized by harsh weather, poor infrastructure, also in Benishangul-Gumuz. low administrative capacity, high poverty, and poor development outcomes. The arid environment in the Host Communities Afar and Somali regions and the small and scattered Except Harari and Sidama regions, the remaining nomadic populations make it more challenging to regional states of Ethiopia host hundreds of provide services. Many parts of the four regions are thousands of refugees in camps and camp-like inaccessible with poor or no roads. 84 Annexes Annex B: Refugee Policies in Ethiopia E thiopia ratified the first national Refugee Proclamation in 2004 (FDRE, 2004) to effectively implement international and regional conventions nine thematic areas: out-of-camp, education, work and livelihoods, documentation, social and basic services, and local integration (Table B.1) (RRS, 2017). of the 1951 UN Convention relating to the Status of Moreover, in 2019, at the first Global Refugee Forum Refugees and its 1967 protocol and the 1967 OAU (GRF), Ethiopia made four additional pledges on Jobs (Organization of African Unity) Convention Governing and Livelihoods, Education, Protection, and Energy the Specific Aspects of the Refugee Problem in (Table B.2) (RRS & UNHCR, 2021). As a result, a new Africa. According to Article 21 of the proclamation, refugee proclamation was adopted in 2019, Refugee refugees have the right to stay in Ethiopia and are Proclamation No. 1110/2019 (FDRE, 2019), replacing provided with identity cards and travel documents. the 2004 proclamation, which was not exhaustive and Moreover, Article 21 (3) states refugees are entitled to up to date with developments and progress made in the same rights and duties as foreigners concerning refugee protection. The 2019 Refugee Proclamation the right to education and work in wage-earning made major improvements concerning the rights employment. Even though the proclamation states and obligations of asylum-seekers, and recognized refugees are entitled to other rights and duties refugees under part four of the proclamation. These contained in the Refugee Convention and the include the right to access basic services (education, OAU Refugee Convention, it does not exhaustively health, banking, telecommunication, vital event address refugees’ rights to access basic services, registration), right to work, right to acquire and right to work, mobility to access better economic transfer property, special protection to vulnerable opportunities, and grounds for local integration. individuals (women, children, refugees with special needs), local integration and naturalization, and Four factors mainly drove the need for a new refugee right to association, freedom of movement, and policy. First, the nature of the refugees’ situation access to justice. Subsequently, three directives in Ethiopia made it difficult to provide sustainable were implemented to implement refugees’ right to solutions to refugees only through humanitarian movement and residence outside of camp, right to assistance and a camp-based approach. Second, work, and grievances and appeals handling. the increase in refugees entering the country due to unresolved crises and emerging conflicts from The first directive, Directive to Determine the neighboring countries was not accompanied by Conditions for Movement and Residence of Refugees financial support from the international community. Outside of Camps, No.01/2019 (RRS, 2019), was issued Third, the predominantly urban background of to enable refugees to establish residence outside of refugees made their accommodation in remote camp to broaden employment opportunities and camps challenging, leading to illegal migration to achieve self-reliance. The directive sets conditions cities. Finally, Ethiopia’s participated in the 2016 for refugees to be eligible for out-of-camp regular New York UN Summit on Addressing Large Scale residency and provides guidelines to obtain, renew, Movement of Refugees and Migrants, followed by and terminate a residence permit. Regular out-of- an agreement to implement nine pledges made at camp residency permits allow refugees to freely the Summit as part of the practical application of move and establish residence in all areas of the the Comprehensive Refugee Response Framework country except Refugees and Returnees Service (RRS, (CRRF) (Kassa et al., 2019). formerly Agency for Refugees and Returnees Affairs) restricted areas, for the interest of refugees’ safety In 2016, the GoE made pledges to improve the rights and to access basic protection and services. The and well-being of refugees following the adoption directive also includes provisions that allow refugees of the New York Declaration and initiation of the to benefit from the urban assistance program. CRRF at the UN Summit. The GoE made pledges on Temporary movement outside of refugee camps is 85 Annexes also granted to refugees through the issuance of pass foreign nationals upon obtaining the appropriate permits at refugee camps, the Zonal Coordination license according to national laws. The requirements Office, and the RRS Head Office. to engage in joint projects, wage employment, and self-employment are illustrated in Box B.1. The second Directive to Determine the Procedure for Refugees' Right to Work, No. 02/2019 (RRS, 2019a) Issuance of work permits or business licenses to provides detailed working procedures to implement refugees is put into practice with a limited scope due Article 26 of the 2019 Refugee Proclamation, to a lack of clarity related to what “most favorable the right to work for refugees to improve living treatment accorded to foreign nationals” refers (World condition and ensure economic benefits. Article 26 Bank, 2023d). For instance, the requirement for a of the proclamation states recognized refugees and work permit does not clearly state whether refugees asylum-seekers have the right to engage in wage- should be exempted from certain requirements, earning employment, agriculture, industry, small such as the minimum business investment amount and micro-enterprises, handicrafts and commerce, of US$150,000 and the need to present an employer and professional work (liberal professions) in the letter that justifies that their skills cannot be found same circumstance as the most favorable treatment among Ethiopian nationals. According to current accorded to foreign nationals under relevant laws. practice, the Ethiopian Diaspora, Djiboutians, and Refugees married to Ethiopian nationals, or who Rastafarians are exempted from these requirements. have children in possession of Ethiopian nationality, Thus, this created an implementation lag in issuing are exempted from restrictions imposed on the work permits or business licenses by the Ministry of employment of foreign nationals. Labor and Skills (MoLS, formerly Ministry of Labor and Social Affairs) or the Ministry of Trade and Regional In line with the proclamation, the directive covers Integration (MoTRI) (World Bank, 2023d). Moreover, detailed guidelines regarding refugees’ participation the 2020 Investment Proclamation of Ethiopia in joint projects, wage-earning employment, and provided a long list of areas52 exclusively reserved self-employment. The joint project approach for Ethiopian nationals, including small businesses works through projects funded by the international that would have been of interest to refugees (FDRE, community, either through the government or NGOs, 2020). Hence, it allows many business activities to go based on the agreement made with the government. beyond the reach of refugees. The projects, however, need to create economic opportunities for refugees and host communities. A draft MoU signed between RRS and MoLS/MoTRI/ Refugees get equal treatment as Ethiopian nationals MoR proposed exceptional treatment of refugees concerning participation in joint projects—rural and engaged in business due to their vulnerability. urban projects designed by the government and According to the MoU, refugees can own a business international community to benefit refugees and in agriculture, manufacturing, services, small and Ethiopian nationals. The second approach—wage- medium enterprises, handicrafts, and trade sectors earning employment—is defined in the directive through establishing private limited companies or as “the performance of professional or manual cooperative societies. Hence, refugees can obtain a work by a refugee who is employed permanently or business license as a private business or association temporarily in consideration for a wage.” Refugees by providing proof of refugee ID, support letter are allowed to be employed in areas that Ethiopian from RRS on the type of business, source of capital, nationals cannot cover. Under the third approach, utilization of profits, qualification certification self-employment, refugees are allowed to work, (depending on the sector), and tax identification individually or in a group, in areas such as agriculture, number. A business license renewal requires a valid industry, medium and small enterprise (MSME), refugee ID, tax clearance, and audit report for a handicraft, and commerce, and in sectors open for limited liability company. 52 Restaurants, tearooms, coffee shops, bars, nightclubs, catering services, producing bakery products and pastries, barbershop and beauty salon services, smithery, tailoring, sawmilling, timber manufacturing, brick and block manufacturing, quarrying, laundry services, translation secretarial services, security services, brokerage services, and attorney and legal consultancy. 86 Annexes Box B.1: Employment pathways of refugees • Refugees can work without a work permit just by obtaining a residence permit, and have obligations to use the residence permit only for the joint project and refrain from illegal activities. • For a residence permit, a refugee shall live for three years in Ethiopia a er granting refugee status, must fulfill conditions to be employed in a joint project and be free of crime charges. The residence permit is valid for five years. Joint projects • A refugee needs to have a renewed refugee identification paper, current residence permit, and employment contract or valid business license or evidence of membership in a cooperative union to renew a residence permit. • Refugees should obtain a work permit except refugees who obtained resident permit to participate in a joint project and who are legally married to Ethiopian nationals or have one or more child in possession of Ethiopian nationality. • An employer seeking to employ a refugee must obtain a work permit for a refugee up on providing documents including a support letter from RRS, an application form, a resume, authenticated educational and work experience qualifications, and a refugee ID. Wage employment • A refugee is not required to provide visa or residence permit to request a work permit. • A work permit is valid for three years. • Refugees are required to obtain residence permit, tax identification number, and one of the three license and certifications (cooperative membership, MSME registration, and business licenses and registration) to engage in self-employment. • Refugees with residence permits are allowed to be self-employed in joint projects without a work permit. • Refugees married to Ethiopian nationals or have one or more child in possession of Ethiopian nationality are Self-employment allowed to engage in business limited to Ethiopian nationals without residence permit upon obtaining the required license. • Self-employment in agriculture and irrigation requires agreement on lease arrangements between RRS and regional governments. Following the signing of the MoU between MoLS forced to compete with other foreigners for similar and RRS, a procedure on technical and vocational job opportunities. Thus, the law seems to have only training, work permits, and job creation and eased the challenges of a few qualified refugees livelihood improvement for refugees was ratified without addressing the needs of the broader and in September 2023. This authorizes RRS to issue largely unskilled refugee community. work permits for refugees. The procedure tasks the RRS and MoLS to ensure that refugees get proper The third directive, Refugees and Returnees information and services to have a work permit and Grievances and Appeals Handling Directive, to create employment opportunities for refugees. No. 03/2019 (RRS, 2019b), provides procedures To get a work permit, refugees must present a work for refugees to submit grievances and appeals permit application form, renewed refugee ID, and concerning any matters related to RRS services. proof of employment from employers. A work permit RRS is responsible for organizing a grievance and is issued for three years for a specific field of work. appeal handling unit at all levels of operation with The renewal of a work permit requires a renewal the mandate to receive and address refugees’ application request, prior work permit, proof of complaints. employment or business license, and tax payment verification. The procedure also includes provisions In the Global Refugee Forum 2023, Ethiopia made six for replacing a lost work permit and canceling a pledges on climate action, human settlement, the work permit. inclusion of refugees into existing national systems (GBV prevention, National ID, secondary school, Regarding wage employment, a directive by MoLS and TVET), private sector engagement, access states that foreigners could only be employed in to irrigable land, and access to documentation Ethiopia where there are no qualified nationals (UNHCR, 2024). A part of the National ID program, available for the job position in question, jobs in a Non- the government started a pilot to issue a digital Governmental Organization (NGO), an organization ID in March 2024 with a plan to reach 77,000 headquartered abroad, and employment because refugees in Addis Ababa. The digital ID is expected of a bilateral or multilateral agreement concluded to give refugees better access to services such as by the government (MoLSA, 2019). Hence, the law healthcare, school enrollment, financial services, does not grant refugees the right to engage in gainful and business registration (UNHCR, 2024a). employment on par with nationals, and refugees are 87 Annexes Table B.1: Pledges made at 2016 UN Leaders’ Summit and progress Pledges Progress as of 2019 Out-of-Camp Policy (OCP) The GoE introduced OCP in 2010 for Eritrean refugees giving opportunities • 5 % of the total refugee population are registered to to live in Addis Ababa and other non-camp location of their choice. benefit from OCP, with regional high regional variation, 83% and 28% in Addis Ababa and Afar, respectively. Expand OCP to all nationalities hosted by Ethiopia, which will benefit • Enacted directives on out-of-camp residence and rights 10% of the refugee population. Refugees who live for more than one to work. month in a camp can apply for a regular out-of-camp residency permit. To be eligible for out-of-camp residency, a refugee should be able to prove that he/she can cover the cost of living or provide a sponsor and receive a work permit. Out-of-camp residency permits can be issued with exceptions for refugees with special conditions (orphaned children, with medical issues, single mothers, elderly, and with urgent overseas travel). Refugees who are no longer beneficiaries of the urban assistance program can also get the permit if they meet the requirements of an out- of-camp residency permit. Education Increase school enrollment: Gross Enrollment Ratio (GER) improved at all school levels, pre-primary (51%), primary (67%) and secondary (13%, 19% • Pre-primary, from 44% to 60% for grades 9 to 10 and 6% for grades 11 to 12). • Primary school, from 54% to 75% • Male GER is higher than female GER, gap widens for • Secondary school, from 9% to 25% primary and secondary levels • Tertiary education, 1600 to 2500 students • GER shows variation across regions and camp sites • Primary and secondary school GER is higher for hosts compared to refugees in most of refugee-hosting regions. Work and Livelihood Provide work permits for refugees with permanent residence ID and Revision of the Refugee Proclamation (No. 1110/2019) refugee graduates based on the relevant domestic laws and in areas incorporating improvements on the right to work (Article permitted to foreign workers, both for in-camp and out-of-camp 26), and endorsement of implementation directive refugees. (Directive to Determine the Procedure for Refugees Right to Work, 02/2019). Avail 10,000 hectares of irrigable land for crop production to refugees 1,103 hectares of land have been made available, and 1,765 and local communities with a plan to benefit 20,000 households or 100, refugees and 1,463 hosts benefited, with the majority being 000 individuals. from the Somali region. Create job opportunities through the development of the infrastructure 4,412 refugees benefited from other livelihood for industrialization. opportunities (income-generating activities startup support, technical and vocational skills, livestock support). Social and basic services Strengthen, expand, and enhance refugees’ access to basic social • Refugees are included in national service provision services such as health, nutrition, immunization, reproductive health, programs related to TB, RH, HIV, mass immunization, HIV, and other medical services. and responses to disease outbreaks. • Health services provided in camps and health facilities through collaboration between RRS, Regional Health Bureaus, and NGOs. • Refugees have access to health posts and health centers within camps and hosting Woredas, and referral hospitals. 88 Annexes Pledges Progress as of 2019 Local integration Permit local integration of refugees who have lived for prolonged • The 2019 refugee proclamation gave RRS the mandate period (over 20 years) in Ethiopia. to facilitate the local integration of refugees who lived in Ethiopia for a prolonged period upon their request. • Positive developments were observed regarding socio- economic integration (skills and entrepreneurial training, access to farming land and peaceful coexistence). Documentation Provide services such as issuance of birth certificates to refugees’ Vital event registration service has been made available to children born in Ethiopia, opening bank accounts and obtaining a all refugees. driving license. • 8,080 events registered in 2019, with the majority being births (7,150), with variation across regions Opening a bank account is allowed by the 2019 Refugee Proclamation. • 13,960 bank accounts opened by refugees, where the majority (67%) in Tigray. Source: Ethiopia Pledge Progress Report 2019 (UNHCR, 2020) 89 Annexes Table B.2: GRF pledges and implementation progress Pledges Progress as of 2023 Jobs and Livelihoods “Create up to 90,000 economic opportunities through agriculture and • A total of 129,449 individuals (38,621 refugees and livestock value chains that benefit refugee and host communities in an 90,828 hosts) directly and indirectly benefited, equitable manner.” respectively, from agriculture, livestock, market system development, and financial inclusion-related services and training. • Major activities include providing irrigable land in Dollo Ado and implementing projects involving agricultural and livestock activities in Gambella, Assosa, and Jijiga. • Inclusion of refugees in urban social safety net program. Education “Expand government TVET system and facilities to provide quality • The Qualification and Employment Perspectives and accredited skills training that is linked to labor market demand to (QEP) initiative was implemented in the Addis Ababa, 20,000 hosts and refugees by 2024.” Tigray, Benishangul-Gumuz, and Gambella regions to integrate refugees in national TVET systems and strengthen the self-reliance of refugees and hosts. • Accredited skills training linked to the labor market is provided for 5,253 refugees and 6,696 hosts. Moreover, 13 TVET colleges are supported. • The government, UNHCR, and GIZ are working together to develop a national roadmap for including refugees in the national TVET system. Protection/Capacity “Strengthening Asylum System and Social Protection: (i) Refugee • The 2019 Refugee proclamation improved provisions Status Determination (RSD), refugee registration, civil documentation, related to registration, documentation, and protection and permits; (ii) National social protection system in refuge hosting of refugees and asylum seekers as well as refugee status areas-particularly for vulnerable individuals.” determination, and three implementation directives adopted. • RSD procedures are simplified for asylum seekers from Syria and Sudan. • Refugees are included in a Civil Registration and Vital Statistics Systems (CRVS) and the National Social and Behavior Strategy (awareness raising about the need for vital events registration). • A backlog of birth registration of 120,000 refugee children is cleared, and 72,286 vital events (62,816 birth, 8,177 marriage, and 757 divorce ) have been registered since 2017. • 890,825 refugees enrolled in the Level 3 Registration and Biometric Identity Management System (BIMS). Refugee ID cards and proof of registration are issued to 55 and 98 percent of refugees, respectively. Out-of-camp Permit is issued to 48,346 refugees. • One-stop shops have been established in 13 refugee camps (14 are under construction) to provide one-center registration, documentation, and protection services. 90 Annexes Pledges Progress as of 2023 • The National Strategy on Violence Against Women and Children (2021 -2026) recognizes refugee women and children. Also, refugees are included in the national Gender Based Violence (GBV). • Digital Request and Compliant System (DRCS) is established and implemented as part of the digitization of refugee protection services. • Refugees are getting mobile courts and free legal aid services. Energy/Environment “Provide market-based sustainable, reliable, affordable, culturally • A National Cooking Fuel Strategy is developed by EEWG acceptable, environmentally friendly clean/renewable energy to guide and define camp-specific cooking energy solutions for 3 million people. options. • In Afar, Gambella, and Melkadida, more than 382,000 refugees and 85,000 hosts have access to alternative cooking fuels and market-based clean electricity from solar-mini grids. • More than 1,739,726 seedlings were planted, 160m3 check dams were built, and 8 km of soil stone band were built to rehabilitate degraded land. Additional GoE GRF Pledges 2023 Climate Action: Protect and restore the environment, manage natural resources, afforestation of degraded lands, and expand renewable energy solutions for the benefit of both refugees and host communities. Recognize vulnerability of women to climate change, and prevent violence against women in all environmental policies and programs, and empower women to have agency and influence in environmental stewardship and adaptation to climate. Human Settlement: Transform selected refugee camps into sustainable urban settlements by enhancing the quality and availability of shelter, infrastructure, and public services, such as roads, electricity, water, sanitation, health, and education by aligning them with adjacent towns’ masterplan, by 2027. Inclusion of refugees into existing national systems: Enhance the capacity of GoE to include 1,000,000 refugees into the national Central Statistics Service (CSS), the national Gender-Based Violence (GBV) prevention and response programs, 814,000 refugees into the national ID program, refugee secondary schools into the national system and 30,000 refugees and host communities in the TVET systems with 70% job opportunities by 2027. Private Sector Engagement: Improve the enabling environment for private sector engagement and investment to foster socio- economic development and to boost productivity of refugee and hosting communities. Access to Land: Provide access to 10,000 hectares of irrigable land through lease arrangements and promote climate-smart agriculture and livestock value chain contributing to improved food security and socio-economic empowerment of refugees and host communities of which at least 50% being women and 30% refugees. Access to documentation: Enhance digital infrastructure in refugee hosting areas to facilitate refugee inclusion to the digital economy including digitally enabled livelihood opportunities and financial inclusion as well as to foster their access to socio- economic e-services, including standardized travel documents. Source: Ethiopia GRF Pledge Progress Report (RRS & UNHCR, 2021 and 2023) 91 Annexes Annex C: Survey Design and Methodology S ESRE is a separate but integrated survey alongside the Ethiopian Household Welfare Statistics Survey (HoWStat),54 the national household survey The SESRE covers all current major refugee camps: Eritreans, South Sudanese, and Somalis, as well as the out-of-camp refugees in Addis Ababa. In addition, to measure poverty and other socio-economic the survey covers the respective host communities outcomes. Like most national poverty surveys, around the camps, including the host communities of HoWStat excludes displaced populations—Internally Addis Ababa. Due to the conflict in the Tigray region Displaced People (IDPs) or refugees—including in of Ethiopia between 2020 and 2022, Eritrean refugees Ethiopia. To have up-to-date information on the living in camps in Tigray could not be included in socio-economic outcomes and poverty levels of this survey. To avoid exclusion of Eritrean refugees refugees and to allow comparison to Ethiopian in Ethiopia, we included Eritrean refugees living in host communities, the SESRE applied the same camps in the Afar region and the newly established questionnaire and data collection methods as refugee hosting zone Alemwach. Eritrean refugees the HoWStat, with some modifications. Training who were in the Tigray region prior to the conflict are of the enumerator team and implementation included in this survey in two ways: we sampled (i) arrangements of the survey followed the same refugees from Alemwach, where most of the refugees standards and procedures as the HoWStat. SESRE previously located in Tigray moved after conflict data was not collected alongside HoWStat due to broke out and (ii) from Addis Ababa, namely those security concerns at the time of data collection for refugees who arrived in Addis Ababa after November HoWStat, especially in the refugee areas. 2020. Data collection took place between November 2022 and January 2023. The SESRE aimed to solve two problems: (i) gaps in data on the socioeconomic dimensions of refugees, Sample population and (ii) gaps in analytical studies presenting the socioeconomic outcomes of refugees and hosts. The SESRE covers three types of groups, all of which Lack of up-to-date evidence is a significant obstacle require a distinct sampling procedure:55 (i) refugees to designing effective policies and support for in camps; (ii) refugees out-of-camps; and (iii) host refugees and host communities. To this end, the communities. This section discusses the sampling availability of the data helps to analyze refugee frames of each group. hosting areas’ social dynamics and longer-term socioeconomic viability by focusing on the: (i) (a) Refugees in Camps social impact of refugees on host communities, (ii) The sampling frame for refugee camps is based on socioeconomic interaction, (iii) social inclusion, and UNHCR’s proGRES database. The refugee camps (iv) social relations among refugees and between were grouped into three domains based on the refugees and host communities. The data provides concentration of refugees from the three major valuable information to development partners origin countries: South Sudan, Somalia, and and governments to inform policies to facilitate Eritrea.56 The first sampling stage divided each camp refugees’ integration and improve their lives, along into enumeration areas (EAs). Based on the proGRES with refugee hosting communities. database, we created pseudo EAs by taking 150-200 54 Formerly the Household Consumption and Expenditure Survey and Welfare Monitoring Survey. 55 Prior to the sampling process, the survey team conducted a pre-sampling assessment by visiting the camps to verify on-the-ground conditions. 56 Refugees from Sudan were not included in SESRE as, at the time of sampling, there were less than 50,000 Sudanese refugees in Ethiopia and inclusion was not deemed cost effective. 92 Annexes households in a row from the list; that is, 50-200 HHs frame57 meets the definition of host community In grouped as EA1 and the next 150-200 households the SESRE, host community members are defined as grouped as EA2, and so on. EAs and households from those who live adjacent to a refugee camp but within each sampled EAs were selected. a radius of 5km. We use the updated Ethiopian Statistics Service 2018 cartographic database of (b) Refugees in Addis Ababa enumeration areas (EAs) to define them. An EA is a We used a slightly different approach for refugees in defined area where 100-150 households live in rural Addis Ababa because of the difficulty of obtaining a areas, while in urban areas, it is an area where 150- reliable, complete list of locations for refugees living 200 households live. The first stage of sampling for there. The refugee sampling frame in Addis was the host community involved using simple random based on UNHCR’s proGRES registration data, sorted sampling to select EAs—the primary sampling unit— by location. The UNHCR list has information about from the list of EAs that are adjacent but within a how many refugee households live in each Woreda radius of 5km. in Addis Ababa, their contact details, location, and other information. We developed pseudo-EAs from Following EA selection, a fresh list of households was the list by location (sub-city and Woreda); some EAs prepared at the beginning of this survey, which was covered more than one Woreda, and multiple EAs used as a frame to choose sampled households were in a single Woreda. We selected a sample of from each sample EA. In Addis Ababa, a separate EAs and households from each EAs in collaboration host domain was developed as refugees spatially with UNHCR. Finding refugees in Addis Ababa concentrate in a few sub-cities and Woredas. We was challenging, as they change their location applied the ESS EA maps around the area where frequently. To minimize the burden of searching for refugees in Addis Ababa are located. We selected selected refugees, representatives of the selected EAs in the first stage and then conducted a households were contacted before the survey to complete listing. ask them to come to a UNHCR center to collect preliminary information, including their current Sampling design residential address. Since many Eritrean refugees in The sample for this survey was 3,456 households Addis Ababa had fled from the conflict in Tigray, out- from eight domains, with data was collected from of-camp refugees in Addis Ababa were stratified into 3,452 households (Table C.1).58 There are three two domains: refugees who arrived before the start domains for the three largest in-camp refugee of the conflict in November 2020, and those who groups—Eritreans, Somalis, and South Sudanese— arrived after November 2020. three for host communities of these major refugee (c) Host Communities groups, and one for refugees and one for host communities in Addis Ababa. In all categories, Populations around the refugee camps under each a stratified, two-stage cluster sample design domain were meticulously identified in consultation technique was used to select EAs and 12 households with UNHCR and RRS. We used the ESS EA maps to per EA, whereby the EAs were considered a Primary assess the settlement of communities around camps, Sampling Unit and the households as the Secondary ensuring a precise fit to the definition of a “host Sampling Unit. The SESRE is designed to estimate community”. The assessment highlighted that using demographic, socioeconomic, welfare, and refugee- the list of EAs obtained from the new cartographic specific indicators of the eight domains. 57 The cartographic map (frame) was prepared in 2018 for the upcoming Population and Housing Census. 58 See Annex C for sample size estimation. 93 Annexes Table C.1: The distribution of sampled and surveyed households by domains EA HH Sampled Covered Sampled Covered Eritrean refugee domain 36 36 432 432 Somalian refugee domain 36 36 432 432 South Sudanese refugee domain 36 36 432 432 Eritrean host domain 36 36 432 430 Somali host domain 36 36 432 431 South Sudanese host domain 36 36 432 432 Addis Ababa refugee domain 36 36 432 431 Addis Ababa host domain 36 36 432 432 Total 288 288 3,456 3,452 Note: We have three segmented EAs in Somali host domain. Sample size estimation (a) First Stage Sampling where the deft is the design factor defined as the In the first stage sampling, each domain is ratio between the square root of standard error considered an explicit sampling domain. We used using the given sample design and the standard the list of all EAs as a sampling frame and their error resulting from a simple random sample used. estimated population as a Measure of Size (MoS). Based on the above scenario, total sample size A sample is selected with Probability Proportional =3,456 Households, and EAs = 288 . to Size (PPS). The sample size is evaluated regarding the expected precision of the key An equal allocation method was used to ensure that indicator for the SESRE, the national household the survey precision was comparable across domains, survey to measure poverty as the percentage where 36 EAs were selected from each domain. Based is 0.235 (2016 Household Consumption and on a fixed sample take of 12 households per cluster, Expenditure). In the calculation, values for the measuring poverty rate (P) and design factor (deft) Equal Allocation formula 1.5, the expected Relative Standard Error (RSE) of 4.63%, and finally, an adjusted Response Rate of 99% at a 95% Confidence level used to represent the expected precision is acceptable at the domain level. To select a representative sample Where: from this population, first, the initial sample size was determined by using the following scientific = total number of sample households and = formula: Number of sample households allocated to stratum 94 Annexes (b) Second Stage Sampling (d) Implementation Plan In the second stage sampling, we selected 12 The survey implementation plan involved households per selected EA. The probability of collaboration with the Ethiopian Statistical Service of selecting a household in segment of EA of (ESS), World Bank, UNHCR, and RRS. The ESS was domain is given by responsible for administering the pre- and post- fieldwork implementation and management, including fieldworkers’ recruitment, training, field , where tests, data collection, data quality assurance, and data management The UNHCR supported in ◆ is the number of EAs in the domain’s sample, engaging refugee communities and leaders. The RRS ◆ is the estimated population of the EA, facilitated access to all camps for the survey teams; this is the first time that the RRS facilitated access ◆ is the estimated population of the domain, to all camps for such an extensive survey. Notably, ◆ is the number of segments listed in the the UNHCR and RRS facilitated the collaboration of EA (normatively always 2, or 1 if the EA is not the field workers with refugee leaders in each camp segmented), and Addis Ababa to support the teams in identifying ◆ is the number of households visited in the EA sampled households and maintaining the safety (normatively always 12), and of the field workers, and the sampled households ◆ is the total number of households listed in during the entire survey period. The World Bank the EA. team led the collaboration between ESS, UNHCR, and RRS and provided technical support to the ESS since the project’s inception. There are 36 EAs per domain. With 8 survey domains, and 12 households per EA, a total of 3,456 households The SESRE used a logistics plan similar to HoWStat. in the sample. Six ESS branches were responsible for administering the survey: the Asayita, Gondar, Jigjiga, Negele, (c) Replacement of Households Gambella, Assosa, and Addis Ababa branches. We implemented a two-layer replacement strategy: Twenty-four field teams carried out the fieldwork, First, in each EA, 12 additional households were each consisting of one statistician, one team sampled to serve as replacement households. We supervisor, and four enumerators. All field staff sampled the replacement households in the sample involved in the SERSE participated in the HoWStat allocation and sample size determination stage. As survey. Enumerators were knowledgeable about part of the data collection protocol, each household local cultures and languages and could detect needs to be visited at least three times before inconsistencies and misunderstandings during replacing a household from the list of replacements. interviews to ensure high-quality data. Supervisors The list of replacement households was only were additionally trained on how to troubleshoot provided to the enumerators upon demonstrating standard technical issues with tablets. The that three visits were attempted. In the case of in- supervisors conducted reinterviews, consistency, camp refugees, if the enumerators, together with the spot-checking, and data syncing to the head office. focal person from RRS and UNHCR, could not identify Also, the statisticians from ESS branch offices were the selected household within the camp, they were with the team all the time to support and monitor provided with a list of replacements. Second, in case the fieldwork. The data collection system consisted of missing to identify even the replaced sampled of encrypted Android devices for prolonged usage households, the enumerators were requested to go in the field equipped with the chosen survey back to the original sampled household and skip ten application and a GPS tracking application for EA households using a counterclockwise rule to find a delineations. Electronic data files were transferred new replacement household. daily to the ESS central office in Addis Ababa via the 95 Annexes secured link. The core team from ESS undertook institutions and development partners. This collaboration field supervision and was responsible for the day-to- allowed the sharing of experiences across institutions day field management. Also, the World Bank team and knowledge for ESS to implement such unique undertook field supervision, providing on-time and surveys in the future. The survey process, from on-the-spot guidance for the field teams whenever preparation to implementation, focused on ensuring and wherever they encountered a challenge. data quality for refugee data collection. During preparation, ESS translated the survey instrument (e) Challenges faced and lessons learned into different main languages and undertook an in- depth training of supervisors and enumerators for The SESRE served as a learning experience for enumerators to understand better the concepts of the including refugees in future rounds of the official questions related to the refugee context. Moreover, a household survey (HoWStat). Given the unique close follow-up and coordination in the field helped to feature of refugees compared to Ethiopians, the get better quality data and provided timely responses sampling methodology for the SESRE is unique for to challenges faced during data collection. sampling refugees. Therefore, SESRE successfully tested the feasibility of sampling refugees and their During the survey implementation period, the main hosts. To ensure the successful implementation of challenge was tracking sampled refugee households the sampling procedures, the ESS implemented a in all refugee domains. One of the Eritrean camps, pilot sampling methodology before data collection Alemwach Camp, was newly established at time started to ensure that all systems and processes of data collection. Tracing the originally sampled were functioning. The ESS and World Bank teams and backup households initially took a lot of work. conducted field visits for this pre-test in Afar and The issue of missing households in Asayita camp Addis Ababa. The field visits included discussions was severe during the second data collection with camp community leaders, including the phase. Moreover, some camps were very large; for refugee community leaders, about the upcoming example, there were more than 100,000 refugees survey to understand better any sensitivities that in one camp, creating challenges for field workers may arise. The field visits helped to understand the in tracing the sampled households. Refugees in camp administrative structure and environment Addis Ababa live in rented houses; the team faced of the teams facilitating the camp and to test the challenges in tracing some refugees due to changes accessibility of sampled refugee households inside in their residential locations. Challenges related the camp, in Addis Ababa, and the identification to identifying the eligible sample households of the host. ESS provided detailed feedback on were also observed due to outdated names of the the fieldwork procedures and adjustments made household heads in UNHCR lists, and UNHCR’s before the fieldwork began. For instance, in Afar registration of names which is not consistent with (Asayita), the visit helped to identify challenges in the Ethiopian context58. Thus, these challenges tracing sampled refugee households and to take the required additional effort by the team to ensure necessary corrective measures. Likewise, in Addis that sampled households and replacements Ababa, the visit assisted in designing an appropriate were traced, identified, and interviewed. Another strategy to select host communities. challenge was that some refugees were not willing to provide their current location due to personal The survey created a good opportunity for a security reasons, but these situations were resolved collaborative effort between different government by reaffirming the confidentiality of the survey. 58 UNHCR register names starting with last name, whereas, in Ethiopia names starts with first name. 96 Annexes Annex D: Descriptive Statistics and Regression Results Results on Sociodemographic Profile Table D.1: Demographic characteristics by survey domains Eritrean Somali South Sudanese Hosts Refugees Hosts Refugees Hosts Refugees Age group <15 39% 46% 54% 50% 47% 56% 15 to 24 19% 19% 17% 22% 21% 21% 25 to 44 26% 25% 21% 17% 21% 16% 45 to 64 13% 8% 7% 9% 8% 5% >=65 4% 2% 2% 2% 3% 1% Gender Male 48% 51% 50% 49% 48% 46% Female 52% 49% 50% 51% 52% 54% Marital Status Never Married 22% 27% 21% 30% 22% 25% Married 62% 56% 69% 52% 66% 57% Other 16% 17% 10% 18% 12% 18% Household characteristics Household size 4.3 4.7 5.7 6.0 5.4 6.5 Dependency ratio 0.8 1.1 1.5 1.4 1.1 1.7 Female-headed 38% 43% 43% 61% 51% 84% Head’s age 45.4 39.3 41.0 43.6 40.7 37.1 Source: World Bank Staff based on SESRE 2023. 97 Annexes Table D.2: Education outcomes by survey domains Eritrean Somali South Sudanese Hosts Refugees Hosts Refugees Hosts Refugees Education level No education 38% 49% 52% 56% 31% 43% Incomplete primary 20% 30% 13% 20% 24% 27% Complete primary 20% 17% 21% 18% 27% 25% Complete secondary 12% 4% 6% 4% 10% 3% Complete post-secondary 10% 0% 9% 2% 8% 2% Education level (youth -15 to 24 years) No education 4% 17% 20% 18% 4% 6% Incomplete primary 29% 59% 27% 41% 48% 62% Complete primary 50% 22% 48% 37% 41% 31% Complete secondary 11% 1% 4% 4% 6% 2% Complete post-secondary 5% 0% 2% 1% 1% 0% Children currently attending school (<=18 years) All 68% 51% 50% 51% 68% 70% Primary school 53% 43% 42% 44% 59% 59% Secondary school 11% 2% 7% 5% 5% 2% Primary school (7 to 14 years) 82% 64% 67% 61% 82% 81% Secondary school (15 to 18 years) 49% 15% 48% 30% 28% 18% Attending school above school age Primary school (15 to 18 years) 31% 49% 28% 49% 54% 73% Secondary school (19 to 24 years) 34% 11% 30% 32% 32% 38% Enrollment rates Primary NER 81% 65% 65% 62% 82% 80% Secondary NER 47% 14% 44% 32% 28% 17% Primary GER 101% 86% 79% 84% 114% 116% Secondary GER 54% 14% 47% 36% 31% 18% Source: World Bank Staff based on SESRE 2023. 98 Annexes Table D.3: Health outcomes by survey domains Eritrean Somali South Sudanese Hosts Refugees Hosts Refugees Hosts Refugees Faced any health problem 21% 20% 6% 7% 29% 25% Received medical assistance 86% 85% 85% 85% 85% 93% Child Nutrition Stunted 43% 52% 37% 47% 26% 26% Underweight 28% 37% 31% 38% 26% 19% Wasted 10% 10% 14% 19% 17% 12% Disability Seeing 2% 3% 1% 2% 3% 3% Hearing 1% 3% 1% 1% 2% 2% Walking or climbing steps 2% 2% 1% 2% 3% 2% Remembering or concentrating 1% 2% 1% 1% 1% 2% Difficulty with self-care 1% 1% 1% 1% 1% 2% Communicating 0% 1% 1% 1% 1% 1% Any disability 5% 8% 3% 4% 6% 5% Source: World Bank Staff based on SESRE 2023. Table D.4: Living conditions by survey domains Eritrean Somali South Sudanese Hosts Refugees Hosts Refugees Hosts Refugees Dwelling type Owned 57% 0% 77% 21% 72% 11% Rented 36% 2% 13% 0% 22% 2% UN/NGO temporary 0% 61% 0% 28% 1% 16% UN/NGO permanent 0% 36% 0% 50% 1% 71% Other 7% 0% 10% 1% 4% 0% Housing quality Overcrowded 22% 66% 44% 53% 42% 56% Improved wall 16% 7% 16% 5% 4% 0% Improved roof 72% 64% 75% 81% 58% 8% WASH Improved source of drinking water 80% 99% 97% 99% 64% 79% Improved bathing facilities 30% 28% 5% 9% 25% 14% Improved toilet facility 38% 39% 60% 59% 18% 34% Improved waste disposal method 24% 64% 30% 37% 10% 8% Source of lighting Electricity (meter) 74% 13% 25% 5% 29% 1% Electricity (meter, generator, solar) 89% 91% 38% 17% 34% 4% Source: World Bank Staff based on SESRE 2023. 99 Annexes Figure D.1: Age group by gender 100 90 80 70 Percent 60 50 40 30 20 10 0 < 15 15 to 25 to 45 to >= 65 < 15 15 to 25 to 45 to >= 65 < 15 15 to 25 to 45 to >= 65 < 15 15 to 25 to 45 to >= 65 24 44 64 24 44 64 24 44 64 24 44 64 Eritrean Somali South Sudanese Addis Ababa Male Female Source: World Bank Staff based on SESRE 2023. Figure D.2: Refugees’ education document 70 60 50 Percent 40 30 20 10 0 Eritrean Somali South Sudanese Addis refugees Availability of education document Able to verify education document Source: World Bank Staff based on SESRE 2023. Figure D.3: Share of school-age children in education per household 100 80 60 Percent 40 20 0 Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Eritrean Somali South Sudanese In camp Addis Ababa Total Average share of primary school age children in primary education Average share of secondary school age children in secondary education Source: World Bank Staff based on SESRE 2023. 100 Annexes Figure D.4: School-age children currently attending school by gender 100 80 60 Percent 40 20 0 Hosts Refugees Hosts Refugees Hosts Refugees In camp Addis Ababa Total Primary school (7 to 14 years) Boys Primary school (7 to 14 years) Girls Secondary school (15 to 18 years) Boys Secondary school (15 to 18 years) Girls Source: World Bank Staff based on SESRE 2023. Figure D.5: Reasons for not currently attending school by gender 15 to 18 years Refugees Boys Girls Hosts Refugees Hosts Girls 8 to 14 years Refugees Hosts Refugees Boys Hosts 0 10 20 30 40 50 60 70 80 90 100 Percent Need to work Unable to cover education expenses (fee and materials) School too far Too young Marriage or pregnancy Family not willing Sickness/injury or natural or human calamites Negative perception towards the benefit of education Other Source: World Bank Staff based on SESRE 2023. Figure D.6: Average annual household education expenditure (in ETB) 18.000 16,000 14,000 12,000 10,000 8,000 6,000 4,000 2,000 - Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Eritrean Somali South Sudanese In Camp Addis Ababa Total Average household expenditure on education (children in school) Average household expenditure on education per child (school age (4 to 18 years)) Source: World Bank Staff based on SESRE 2023. 101 Annexes Figure D.7: Type of health institutions 100 80 Percent 60 40 20 0 Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Eritrean Somali South Sudanese In camp Addis Ababa Total Government/public Private Mission/Religious/NGO Other Source: World Bank Staff based on SESRE 2023. Figure D.8: Problems faced in health institutions a. Faced any problem b. Types of problems faced 45 Refugees 40 In camp Addis Ababa Total Hosts 35 Refugees 30 Hosts Percent 25 Refugees 20 Hosts 15 10 0 10 20 30 40 50 60 70 80 90 100 Percent 5 Sanitation problem Long waiting time 0 Hosts Refugees Hosts Refugees Hosts Refugees Shortage of health professionals Too expensive In camp Addis Ababa Total Shortage/unavailability of medicines Unavailability of laboratory Shortage of medical equipment Lack of cooperativeness of the sta Source: World Bank Staff based on SESRE 2023. Figure D.9: Stunting by gender of children 60 50 40 Percent 30 20 10 0 Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Eritrean Somali South Sudanese In camp Addis Ababa Total Female Male Source: World Bank Staff based on SESRE 2023. 102 Annexes Figure D.10: Childbirth in health institutions (children under five years) 100 90 80 70 60 Percent 50 40 30 20 10 0 Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Eritrean Somali South Sudanese In camp Addis Ababa Total Health institutions (health center/hospital) Home Other Source: World Bank Staff based on SESRE 2023. Figure D.11: No birth evidence available (children under five years) 70 60 50 Percent 40 30 20 10 0 Eritrean Somali South Sudanese In camp Addis Ababa Total Hosts Refugees Source: World Bank Staff based on SESRE 2023. Figure D.12: Average annual per capita health expenditure 3,500 3,000 2,500 2,000 1,500 1,000 500 - Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Eritrean Somali South Sudanese In camp Addis Ababa Total Average household annual expenditure on health Average per capita annual expenditure on health Source: World Bank Staff based on SESRE 2023. 103 Annexes Figure D.13: Types of disability Refugees Total Hosts Refugees Addis Ababa Hosts Refugees In Camp Hosts Refugees South Sudanese Hosts Refugees Somali Hosts Refugees Eritrean Hosts 0 10 20 30 40 50 60 70 80 90 100 Percent Seeing Hearing Walking or climbing steps Remembering or concentrating Di iculty with selfcare Communicating Source: World Bank Staff based on SESRE 2023. Figure D.14: Rent expenditure (Refugees and hosts in Addis Ababa) 60,000 50,000 40,000 30,000 20,000 10,000 - Hosts Refugees Addis Ababa Total annual rent expenditure Per adult equivalent rent expenditure Source: World Bank Staff based on SESRE 2023. Figure D.15: Hand washing facility 100 80 60 Percent 40 20 0 Eritrean Somali South Addis Ababa Eritrean Somali South Addis Ababa Eritrean Somali South Addis Ababa Sudanese Sudanese Sudanese Has hand washing place/item Water available Detergent available Hosts Refugees Source: World Bank Staff based on SESRE 2023. 104 Annexes Table D.5: Labor force statistics by survey domains Eritrean Somali South Sudanese Hosts Refugees Hosts Refugees Hosts Refugees Participation (strict) 57% 38% 44% 35% 55% 28% Employment (strict) 89% 72% 94% 77% 96% 82% Unemployment (strict) 11% 28% 6% 23% 4% 18% Participation (relaxed) 62% 60% 51% 49% 57% 38% Employment (relaxed) 81% 45% 81% 55% 91% 60% Unemployment (relaxed) 19% 55% 19% 45% 9% 40% Source: World Bank Staff based on SESRE 2023. Table D.6: Determinants of refugee-host earnings gap (1) (2) (3) (4) Earnings Earnings Earnings Earnings -69.8*** -63.7*** -62.1*** -40.7*** Refugee (% difference from hosts) (0.077) (0.123) (0.122) (0.126) Control: Region Yes Yes Yes Yes Control: Demographics Yes Yes Yes Control: Occupation/Sector Yes Yes Sample: Working outside camp Yes Sample Size 743 742 742 572 Source: World Bank Staff based on SESRE 2023. Note: Monthly earnings are collected for employees only (including work for government, NGOs, and private households). Log earnings are winsorized at the 1st and 99th percentile within the domain. Regression coefficients are transformed to percent change interpretation using. Standard errors clustered at the EA level. * p < 0.10, ** p < 0.05, *** p < 0.01 105 Annexes Table D.7: Determinants of employment outcomes (1) (2) (3) (4) (5) (6) (7) (8) Hosts Refugees Working High-Skill Ln Income Working High-Skill Work Ln Income Ln Income Outside 0.235*** 0.030* 0.270*** 0.033 0.031* 0.172*** 0.506*** 0.876*** Male (0.021) (0.018) (0.061) (0.033) (0.017) (0.049) (0.101) (0.226) 0.067 *** 0.008 ** 0.064 *** 0.045 *** 0.004 -0.005 0.007 -0.043 Age (0.004) (0.004) (0.017) (0.005) (0.003) (0.013) (0.028) (0.046) -0.001*** -0.000* -0.001*** -0.001*** -0.000 0.000 -0.000 0.001 Age Sq. (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.001) - - - - - - - - Educ: < Primary -0.028 0.058*** 0.127 -0.042* 0.077** -0.104 -0.172 0.121 Educ: Primary (0.020) (0.021) (0.107) (0.022) (0.032) (0.069) (0.122) (0.361) 0.115*** 0.506*** 0.700*** -0.045 0.460*** -0.200** 0.098 0.849*** Educ: Secondary (0.031) (0.046) (0.100) (0.045) (0.107) (0.080) (0.212) (0.311) 0.231*** 0.716*** 0.812*** 0.004 0.493*** -0.073 0.222 0.528** Educ: Post-sec (0.024) (0.037) (0.078) (0.117) (0.121) (0.149) (0.367) (0.232) 0.009 *** 0.001 0.001 0.003 0.006 Years in Ethiopia (0.003) (0.002) (0.004) (0.012) (0.015) 0.352*** Work outside camp (0.128) Region fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Restrict to workers Yes outside camp N 3321 1626 494 3,069 830 975 238 73 Source: World Bank Staff based on SESRE 2023. Note: Columns 2 and 5-6 are restricted to working respondents. Columns 3 and 7-8 are restricted to employees. High-skill occupations include managers, professionals, and associate professionals. Ln Earnings is the log of monthly earnings winsorized at the 1st and 99th percentiles. All other models are linear probability models. Standard errors clustered at the EA level. * p < 0.10, ** p < 0.05, *** p < 0.01 106 Annexes Table D.8: Refugee Household Reliance on NGOs/Donations (1) (2) (3) (4) All Eritrea Somali South Sudan -0.007** -0.017** -0.011** -0.001 Years in Ethiopia (0.003) (0.006) (0.005) (0.005) -0.080** -0.335*** -0.180*** -0.001 Member works outside the camp (0.035) (0.099) (0.051) (0.041) Region Fixed Effects Yes No No No Demographic Controls Yes Yes Yes Yes N 1252 423 412 417 Source: World Bank Staff based on SESRE 2023. Note: Each column is a linear probability model where the outcome is a binary indicator of whether the household relies primarily on donations for income. Demographic controls include the household size and the share within each age and education group. Standard errors clustered at the EA level. * p < 0.10, ** p < 0.05, *** p < 0.01 Table D.9: Determinants of refugee-host earnings gap (1) (2) (3) Earnings Earnings Earnings -24.9** -22.4** -18.5* Refugee (% difference from hosts) (0.137) (0.115) (0.114) Control: Demographics Yes Yes Control: Occupation/Sector Yes N 503 503 503 Source: World Bank Staff based on SESRE 2023. Note: Monthly earnings are collected for employees only (including work for government, NGOs, and private households). Log earnings are winsorized at the 1st and 99th percentile within the domain. Regression coefficients are transformed to percent change interpretation using. Standard errors clustered at the EA level. * p < 0.10, ** p < 0.05, *** p < 0.01 107 Annexes Figure D.16: Top 3 difficulties with being a refugee by survey domains Figure D.17: Work status by survey domains 0.8 Refugees Sudanese 0.7 South 0.6 Hosts 0.5 Refugees Somali 0.4 0.3 Hosts 0.2 Refugees Eritrean 0.1 Hosts 0 Eritrean Somali South Sudanese 0 10 20 30 40 50 60 70 80 90 100 Lack of work/business opportunities High cost of living Percent Poor services or institutional support Lack of freedom or mobility Employed Unemployed, searching Unemployed, not searching Lack of community/family networks Insecurity or discrimination Inactive not in school Inactive in school Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. Figure D.18: Type of work by survey domains Figure D.19: Occupation by survey domains 100 100 90 80 80 70 60 Percent 60 Percent 50 40 40 30 20 20 0 10 Hosts Refugees Hosts Refugees Hosts Refugees 0 Hosts Refugees Hosts Refugees Hosts Refugees Eritrean Somali South Sudanese Eritrean Somali South Sudanese Elementary Occupations Machine Operators/Assemblers Cra /Related Trade Workers Skilled Agricultural Workers Self-employed Employee Public Service/Sales Workers Clerical Support Workers NGO/RRS Private household Unpaid Tech/Associate Professionals Managers/Professionals Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. Figure D.20: Work location by survey domains Figure D.21: Hours per week by survey domains 50 45 South Sudanese 40 35 30 Somali 25 20 15 Eritrean 10 5 0 10 20 30 50 40 60 70 80 90 100 0 Percent Hosts Refugees Hosts Refugees Hosts Refugees Inside the camp Outside the camp Eritrean Somali South Sudanese Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. 108 Annexes Figure D.22: Hourly earnings by survey domains Figure D.23: Household owns crops 70 0.6 60 0.5 50 0.4 40 0.3 30 0.2 20 0.1 10 0 0 Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Eritrean Somali South Sudanese Eritrean Somali South Sudanese Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. Figure D.24: Household owns livestock Figure D.25: Total value of livestock 0.6 80,000 70,000 0.5 60,000 0.4 50,000 0.3 40,000 30,000 0.2 20,000 0.1 10,000 0 0 Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Eritrean Somali South Sudanese Eritrean Somali South Sudanese Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. Figure D.26: Value per tropical livestock unit Figure D.27: Household has non-farm business 25,000 0.4 0.35 20,000 0.3 0.25 15,000 0.2 10,000 0.15 0.1 5,000 0.05 0 0 Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Refugees Eritrean Somali South Sudanese Eritrean Somali South Sudanese Source: World Bank Staff based on SESRE 2023. Source: World Bank Staff based on SESRE 2023. 109 Annexes Figure D.28: Value of productive assets in households with business 35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 Hosts Refugees Hosts Refugees Hosts Refugees Eritrean Somali South Sudanese Source: World Bank Staff based on SESRE 2023. Figure D.29: Primary source of income pre-post migration by survey domains 100 90 80 70 Percent 60 50 40 30 20 10 0 Hosts Refugees-Current Refugees-COB Hosts Refugees-Current Refugees-COB Hosts Refugees-Current Refugees-COB Eritrean Somali South Sudanese Salary (employment/casual labor) Crops/livestock Donations(NGO/gov) Remittances (local/international) Other (rental income, PSNP, pension) Source: World Bank Staff based on SESRE 2023. Figure D.30: Youth work status by survey domains Sudanese Refugees South Hosts Refugees Somali Hosts Refugees Eritrean Hosts 0 10 20 30 40 50 60 70 80 90 100 Percent Employed Unemployed, searching Unemployed, not searching Inactive not in school Inactive in school Source: World Bank Staff based on SESRE 2023. 110 Annexes Results on Refugees’ Aspirations Table D.10: Refugee intention to migrate abroad (1) (2) Camp-Based Refugees OCP Refugees 0.022 -0.001 Male (0.019) (0.008) - - Age Under 30 -0.000 -0.007 Age 30-44 (0.020) (0.008) -0.082*** -0.088** Age 45-64 (0.027) (0.036) Education: Primary incomplete - - -0.008 0.004 Education: Completed primary (0.036) (0.009) 0.016 0.006 Education: Completed secondary (0.064) (0.009) -0.016 -0.010 Education: Completed post-secondary (0.067) (0.034) -0.000 -0.000 Years in Ethiopia (0.004) (0.001) Region Fixed Effects Yes No N 3,069 830 Source: World Bank Staff based on SESRE 2023. Note: The outcome positively responds to the question, ‘Do you intend to migrate abroad?’. The sample includes all refugees aged 15 or over who were born abroad. Standard errors clustered at the EA level. * p < 0.10, ** p < 0.05, *** p < 0.01 111 Annexes Results on Welfare and Equity Table D.11: Poverty headcount rate by subgroups In-camp Addis Ababa Characteristics Subgroups Hosts Refugees Hosts Refugees Eritrean 16% 73% Somali 38% 76% Location or domain South Sudanese 36% 89% Addis Ababa 18% 7% Female 34% 87% 16% 10% Sex of head Male 30% 74% 19% 3% No education 38% 87% 35% 16% Primary incomplete 30% 81% 26% 13% Primary complete 29% 79% 14% 6% Head education Secondary incomplete 24% 65% 14% 5% Secondary complete 16% 63% 14% 3% Post-secondary 19% 68% 12% 0% Agriculture 44% 87% 76% Sector of head’s Industry 24% 75% 17% 0% employment Service 23% 84% 17% 8% Unemployed 35% 84% 18% 8% Salary 21% 62% 20% 6% Casual labor 45% 66% 37% 20% Crop/livestock farming 40% 87% Manufacturing 13% 65% 0% Main livelihood source Trade and services 19% 67% 15% 0% Safety nets or aid 33% 87% 27% 0% Remittances 40% 53% 0% 7% Others 66% 85% 8% 0% Low accessibility 34% 78% Market accessibility Medium accessibility 45% 92% High accessibility 16% 77% Nearest to zone 34% 93% Proximity to Nearest to woreda 24% 72% resource hubs Nearest to border 37% 81% Remote 33% 78% Source: World Bank Staff based on SESRE 2023. 112 Annexes Table D.12: Determinants of welfare (total expenditure per capita) (1) (2) (3) (4) In-camp refugees In-camp hosts OCP refugees OCP hosts 0.00 0.03 -0.01 0.02 Female headed (0.03) (0.03) (0.04) (0.05) 0.00** 0.00 -0.00 -0.00 Age of head (year) (0.00) (0.00) (0.00) (0.00) -0.12*** -0.09*** -0.20*** -0.16*** Household size (0.01) (0.01) (0.01) (0.02) 0.02*** 0.01*** 0.03*** 0.03*** Head years of schooling (0.00) (0.00) (0.01) (0.00) 0.10*** 0.19*** 0.39*** 0.17 Mobile phone (0.03) (0.03) (0.11) (0.15) 0.15*** 0.22*** 0.14 0.19 Household has electricity (0.04) (0.03) (0.27) (0.13) 0.04 0.03 -0.42* HH owns any livestock (0.03) (0.04) (0.25) 0.13*** 0.13*** 0.07 0.16 HH member has bank account (0.03) (0.03) (0.10) (0.26) -0.02 -0.08* -0.47* HH has agricultural holding (0.03) (0.04) (0.24) 0.07** -0.00 -0.05 0.26*** HH operates a nonfarm enterprise (0.03) (0.03) (0.10) (0.07) 0.35*** 0.57*** 0.14** 0.06 Share of employed members (0.08) (0.06) (0.06) (0.08) 0.07 -0.02 -0.04 -0.20 Health shock (0.06) (0.06) (0.10) (0.18) -0.01 -0.09*** -0.04 -0.19*** Market shock (0.02) (0.03) (0.04) (0.05) 0.01 -0.11 -0.06 -0.07 Employment shock (0.07) (0.11) (0.31) (0.18) -0.03 0.08 0.24 Drought shock (0.05) (0.09) (0.32) -0.05 -0.23* -0.23** -0.06 Political shock (0.05) (0.12) (0.11) (0.38) 10.05*** 10.22*** 10.58*** 10.49*** Constant (0.07) (0.09) (0.32) (0.36) Survey domain Yes Yes Yes Yes Survey time Yes Yes Yes Yes Observations 1286 1287 432 430 Source: World Bank Staff based on SESRE 2023. Note: Dependent variable is the log of total per adult equivalent consumption expenditure. All regressions include fixed effects for the survey domain and time (year and month) to account for locational and temporal differences. Standard errors in parentheses; * p < 0.10, ** p < 0.05, *** p < 0.01 113 Annexes The regression specification used is: where is the vector of control variables that include demographic characteristics (sex of household head, age of household head, family size, years of schooling completed by the head), assets and wealth (mobile phone ownership, livestock ownership in tropical livestock units, land ownership, bank account, non-farm business ownership, electricity access), employment (share of employed members), resource and market access (market accessibility and proximity to resource hubs), and shocks (health, market, employment, drought, political). The regression also controls for survey domain and survey time (month) fixed effects to account for the effects of location and time on welfare. Table D.13: Determinants of welfare for in-camp refugees (1) In-camp refugees 0.00 Head years since refugee status (from 2022) (0.00) 0.02 Head wants to go back to own/parents (0.03) -0.43*** Ration change (0.10) 0.05 Head: has any relative in own/parents’ COB (0.04) -0.00 HH received humanitarian food aid in past 12 months (0.04) -0.10*** Distance to woreda capital (log) (0.01) -0.03 Distance to border (log) (0.02) -0.02 Medium market accessibility (0.06) 0.29*** High market accessibility (0.05) 11.10*** Constant (0.34) Survey domain Yes Survey time Yes Observations 1,266 Source: World Bank Staff based on SESRE 2023. Note: Dependent variable is the log of total per capita consumption expenditure. All regressions include the controls in Table D.12. This table provides the results for the additional independent variables of interest. Standard errors in parentheses; * p < 0.10, ** p < 0.05, *** p < 0.01 114 Annexes Results on Markets and Opportunities Figure D.31: In-camp refugee locations by ecological Zone 2% 38% 60% Drought prone, lowland, pastoralist Humid moisture reliable, lowland Moisture reliable, highland-Cereal Source: World Bank Staff based on SESRE 2023 and Ethiopia Ecological Zone Classification from ESS. Figure D.32: Refugee’s labor market performance a. By distance to Zone city b. By distance to border 100 100 90 90 80 80 Share of individuals (%) Share of individuals (%) 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 Employment rate Unemployment rate LFP Employment rate Unemployment rate LFP 1-20km 20 -100km >100km 1-30km 30 -50km >50km Source: World Bank Staff based on SESRE 2023. Figure D.33: Economic sector a. By distance to Zone city b. By distance to border 1-20km 1- 30km 20 -100km 30 - 50km >100km >50km 0 20 40 60 80 100 0 20 40 60 80 100 Share of individuals (%) Share of individuals (%) Trade Industry Other service Agriculture Trade Industry Other service Agriculture Source: World Bank Staff based on SESRE 2023. 115 Annexes Table D.14: Variables used to estimate employment outcomes Indicators type Variables Data source Individual characteristics Age, sex, education, language skill, years in exile SESRE Gender of household head, household size, head Household characteristics education level, access to electricity, productive SESRE asset ownership, food insecurity experience Community Community economic Predominant land cover in the community SESRE characteristics development status Distance to towns and cities, distance to the Ethiopian shapefile and Remoteness nearest international border Refugee geospatial from ESS LFPR, unemployment rate, the share of wage Local labor market employment, the share of employment by LMS, 2021 Local Factors economic sector Ethiopia transport network Market access Market accessibility index layer, 2020 (ERA) & gridded population (GPWv4) Notes: For logistic regression, we assume local factors are exogenous in the model as refugees do not select their location. Since refugees’ residential location is not self-selected, they do not choose their respective camps to maximize their utility. Instead, they come across the border and are either assigned to camps close to where they crossed or a new camp is established. Our model compares refugee employment status by local factors in the hosting Zones and community: where subscripts denote : individual and : camp. refers to local factors; represents personal and presents household characteristics; refers to the characteristics of the community. The model includes control for refugee camps. 116 Annexes Table D.15: Factors determining the odds of obtaining a job for refugees: logit model Basic Local Market Proximity Market Variables Model access Model I Model II Model I Model II Individual feature 0.1372*** 0.1372*** 0.1372*** 0.1372*** 0.1291*** 0.1372*** Male (0.0283) (0.0283) (0.0283) (0.0283) (0.0269) (0.0283) 0.0521*** 0.0521*** 0.0521*** 0.0521*** 0.05459*** 0.0521*** Age (0.0075) (0.0075) (0.0075) (0.0075) (0.0069) (0.0075) -0.0006*** -0.0006*** -0.0006*** -0.0006*** -0.0007*** -0.0006*** Age squared (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) 0.1172** 0.1172** 0.1172** 0.1172** 0.08619* 0.1172** Some primary (0.0384) (0.0384) (0.0384) (0.0384) (0.0364) (0.0384) -0.0459 -0.0459 -0.0459 -0.0459 -0.04978 -0.0459 Speaks additional language (0.0343) (0.0343) (0.0343) (0.0343) (0.0327) (0.0343) 0.0857** 0.0857** 0.0857** 0.0857** 0.1048*** 0.0857** >15 years in exile (0.0317) (0.0317) (0.0317) (0.0317) (0.0307) (0.0317) Household feature -0.0458 -0.0458 -0.0458 -0.0458 -0.02736 -0.0458 Male head (0.0308) (0.0308) (0.0308) (0.0308) (0.0290) (0.0308) -0.0375 -0.0375 -0.0375 -0.0375 -0.001520 -0.0375 Head: primary education (0.0389) (0.0389) (0.0389) (0.0389) (0.0378) (0.0389) -0.0919 -0.0919 -0.0919 -0.0919 -0.04109 -0.0919 Head: secondary education (0.0628) (0.0628) (0.0628) (0.0628) (0.0595) (0.0628) 0.2333+ 0.2333+ 0.2333+ 0.2333+ 0.2705* 0.2333+ Head: post-secondary education (0.1318) (0.1318) (0.1318) (0.1318) (0.1115) (0.1318) -0.0233* -0.0233* -0.0233* -0.0233* -0.01886* -0.0233* HH size: member age [15,29] (0.0095) (0.0095) (0.0095) (0.0095) (0.0093) (0.0095) -0.0792*** -0.0792*** -0.0792*** -0.0792*** -0.07340*** -0.0792*** HH size: member age (30,44] (0.0220) (0.0220) (0.0220) (0.0220) (0.0220) (0.0220) -0.0363 -0.0363 -0.0363 -0.0363 -0.02057 -0.0363 HH size: member age (45,64] (0.0254) (0.0254) (0.0254) (0.0254) (0.0256) (0.0254) 0.0961+ 0.0961+ 0.0961+ 0.0961+ 0.1081* 0.0961+ HH access electricity (0.0550) (0.0550) (0.0550) (0.0550) (0.0551) (0.0550) 0.0953* 0.0953* 0.0953* 0.0953* 0.09562* 0.0953* HH own cart (0.0421) (0.0421) (0.0421) (0.0421) (0.0423) (0.0421) -0.0859** -0.0859** -0.0859** -0.0859** -0.09167** -0.0859** HH ran out of food a (0.0304) (0.0304) (0.0304) (0.0304) (0.0300) (0.0304) Local demography and economy activity -0.0798+ -0.0798+ -0.0798+ -0.0798+ -0.07447 -0.0798+ Most land cover: agricultureb (0.0475) (0.0475) (0.0475) (0.0475) (0.0482) (0.0475) -0.0400 -0.0400 -0.0400 -0.0400 -0.0400 Distance to bank (0.0250) (0.0250) (0.0250) (0.0250) (0.0250) -0.0876*** -0.0391* Internal migration: Rural-Urbanc (0.0262) (0.0157) Local labor market 117 Annexes Basic Local Market Proximity Market Variables Model access Model I Model II Model I Model II 0.0337* 0.0370*** LFPR (0.0139) (0.0091) 0.0427+ Share of wage employment (0.0239) -0.0290** Unemployment rate (0.0106) 0.0429** Share of employment in the service sector (0.0077) 0.0665*** Share of employment in the trade sector (0.0147) Share of employment in the 0.0269 manufacturing sector (0.0237) Share of employment in the 0.0206 agriculture sector (0.0131) Proximity and access to market -0.3414** Level two (0.1143) -0.3308** Level three (0.1127) -0.2098** Level four (0.0814) -0.0030* Distance to Zone city (Km) (0.0013) 0.4056** Market accessibility indicator (0.1381) Observations 2024 2024 2024 2024 2205 2024 Chi-square test 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 R2 0.1373 0.1373 0.1373 0.1373 0.1172 0.1373 Source: World Bank Staff based on SESRE 2023. Note: Average marginal effects are estimated. a refers whether a household ran out of food in the last 12 months. b refers if most of the land is covered in the community by agriculture activities i.e., less built-up and shops. c refers recent (5 years) internal migrants from rural to urban centers. Distance to the nearest bank variable is excluded from proximity model II, as it is captured by effect of distance to the nearest zone city. The left side for years in exile, Head education level, HH size, most land cover, and proximity level is less than 15 years, Head with no education, non-working age members, most land cover by built-up and shops, and nearest to Zone capital city, respectively. All estimates are controlled for refugee camps. Standard errors in parentheses. +p<0.10, * p<0.05, ** p<0.01, *** p<0.001 118 Annexes Table D.16: Proximity and market accessibility effects on engagement in agriculture activity: logit model Proximity Variables Market access Model I Model II Individual feature 0.1979*** 0.2016*** 0.1998*** Male (0.0289) (0.0286) (0.0297) -0.0307*** -0.0294*** -0.0296*** Age (0.0083) (0.0081) (0.0088) 0.0004*** 0.0004*** 0.0004** Age squared (0.0001) (0.0001) (0.0001) -0.1403*** -0.1436*** -0.1484*** Some primary (0.0305) (0.0298) (0.0307) Household feature -0.0157 -0.0171 -0.0186 HH size: member age [15,29] (0.0118) (0.0113) (0.0117) 0.0422* 0.0414* 0.0320+ HH size: member age (30,44] (0.0179) (0.0178) (0.0185) 0.0187 0.0278 0.0314 HH size: member age (45,64] (0.0221) (0.0226) (0.0237) -0.2980** -0.2655** -0.2658** HH access electricity (0.0929) (0.0884) (0.0974) Proximity and market access 0.2065*** Level two (0.0607) 0.2358** Level three (0.0788) 0.3559*** Level four (0.0769) 0.0022*** Distance to Zone city (Km) (0.0006) -0.0792* Market accessibility indicator (0.0330) Observations 742 737 742 Chi-square test 0.0000 0.0000 0.0000 R2 0.2517 0.2435 0.2128 Source: World Bank Staff based on SESRE 2023. Standard errors in parentheses. All estimates are controlled for regions. +p<0.10, * p<0.05, ** p<0.01, *** p<0.001 119 Annexes Table D.17: Proximity and market accessibility effects on engagement in service sector: logit model Proximity Variables Market access Model I Model II Individual feature -0.2151*** -0.2199*** -0.2153*** Male (0.0309) (0.0304) (0.0312) 0.0368*** 0.0331*** 0.0340*** Age (0.0089) (0.0089) (0.0096) -0.0005*** -0.0004*** -0.0004*** Age squared (0.0001) (0.0001) (0.0001) 0.1561*** 0.1636*** 0.1683*** Some primary (0.0333) (0.0334) (0.0344) Household feature 0.0217+ 0.0226+ 0.0248* HH size: member age [15,29] (0.0121) (0.0117) (0.0119) -0.0664*** -0.0657*** -0.0553** HH size: member age (30,44] (0.0197) (0.0195) (0.0205) -0.0146 -0.0232 -0.0248 HH size: member age (45,64] (0.0246) (0.0258) (0.0269) -0.0637 -0.0571 -0.1051+ HH access electricity (0.0489) (0.0499) (0.0559) Proximity and market access -0.3000*** Level two (0.0713) -0.3988*** Level three (0.0821) -0.4862*** Level four (0.0832) -0.002* Distance to Zone city (Km) (0.0008) 0.0897** Market accessibility indicator (0.0307) Observations 787 782 787 Chi-square test 0.0000 0.0000 0.0000 R2 0.2871 0.2738 0.2538 Source: World Bank Staff based on SESRE 2023. Standard errors in parentheses. All estimates are controlled for regions. +p<0.10, * p<0.05, ** p<0.01, *** p<0.001 120 Annexes Results on Social Cohesion Table D.18: Regression analysis of host and refugee attitudes (1) (2) (3) Hosts: Hosts: Refugees: Attitudes Index Trusts Refugees Trusts Hosts -0.021 -0.025 -0.029 Male (0.060) (0.036) (0.039) 0.004 0.003 -0.003 Age (0.009) (0.005) (0.006) -0.000 -0.000 0.000 Age Sq. (0.000) (0.000) (0.000) - - - Educ: Primary incomplete -0.094 -0.040 -0.009 Educ: Completed primary (0.066) (0.050) (0.052) -0.003 0.033 -0.083 Educ: Completed secondary (0.098) (0.055) (0.066) -0.074 -0.073 -0.055 Educ: Completed post-sec. (0.103) (0.050) (0.106) 0.408 *** 0.116 ** Agrees on improved local services (0.103) (0.055) 0.002 Years in Ethiopia (0.005) 0.215*** Agrees hosts culturally similar (0.055) Region Fixed Effects Yes Yes Yes N 1724 1666 1613 Source: World Bank Staff based on SESRE 2023. The Attitudes Index is the average of 10 questions regarding beliefs about refugees’ character, the rights they should receive, and their impact on the host community, standardized to a mean of 0 and SD of 1, where positive indicates better attitudes. Trusts Refugees is the binary response of hosts to “Do you think most refugees in Ethiopia are trustworthy?” and Trusts Hosts is the binary response of refugees to “Do you think most Ethiopians are trustworthy?”. Standard errors clustered at the EA level. * p < 0.10, ** p < 0.05, *** p < 0.01 121 Annexes Table D.19: Regression analysis of social integration outcomes (1) (2) (3) (4) (5) Has in Ethiopia: Easy to do: Family Friend Market Social Sharing Interactions Interactions Resources 0.007 0.067** 0.033 0.010 -0.076** Male (0.017) (0.033) (0.027) (0.036) (0.035) - - - - - Age Under 30 -0.030 0.033 0.007 0.004 -0.048* Age 30-44 (0.021) (0.025) (0.025) (0.029) (0.028) -0.041 0.024 0.002 0.011 0.012 Age 45-64 (0.025) (0.038) (0.038) (0.050) (0.041) -0.009 -0.062 0.100** 0.041 0.046 Age Over 64 (0.046) (0.058) (0.043) (0.089) (0.073) - - - - - Educ: Primary incomplete -0.004 0.121*** 0.008 0.028 0.063** Educ: Completed primary (0.027) (0.033) (0.035) (0.041) (0.032) -0.026 0.218*** -0.020 0.061 0.060 Educ: Completed secondary (0.028) (0.054) (0.051) (0.060) (0.057) 0.003 0.238 * 0.100 ** -0.020 0.238 Educ: Completed post-sec. (0.045) (0.131) (0.040) (0.095) (0.155) 0.003** 0.006* 0.008*** -0.001 0.006 Years in Ethiopia (0.002) (0.004) (0.003) (0.004) (0.004) 0.021 0.047 0.020 -0.010 -0.030 Agrees hosts culturally similar (0.016) (0.039) (0.037) (0.051) (0.036) Region Fixed Effects Yes Yes Yes Yes Yes N 1667 1667 1667 1667 1667 Source: World Bank Staff based on SESRE 2023. Standard errors clustered at the EA level. * p < 0.10, ** p < 0.05, *** p < 0.01 122 Annexes Table D.20: Regression analysis of social integration and labor market outcomes (1) (2) (3) (4) (5) (6) Working High-Skill Work Outside 0.069* 0.062* 0.046* 0.044* -0.085 -0.095 Has Ethiopian friend (0.037) (0.037) (0.024) (0.024) (0.066) (0.068) 0.051 0.041 0.026 0.024 0.149 0.139 Has Ethiopian family (0.057) (0.057) (0.035) (0.036) (0.120) (0.116) 0.053 0.010 0.050 Social interactions easy (0.033) (0.019) (0.073) 0.078* 0.079* 0.034** 0.034** 0.122 0.125 Agrees hosts culturally similar (0.046) (0.046) (0.017) (0.017) (0.079) (0.079) Region Fixed Effects Yes Yes Yes Yes Yes Yes Individual Controls Yes Yes Yes Yes Yes Yes N 1625 1625 524 524 445 445 Source: World Bank Staff based on SESRE 2023. Note: Columns 3-4 are restricted to workers, and Columns 5-6 are restricted to workers in camps. High-skill occupations include managers, professionals, and associate professionals (around 7% of refugees). Standard errors clustered at the EA level. * p < 0.10, ** p < 0.05, *** p < 0.01 123 Annexes Annex E: Robustness Checks of Refugees’ Consumption This Annex discusses assessing the disparity between refugee ration aid and reported consumption quantities. This is reported as a robustness check. As discussed earlier, the expenditure of in-camp refugees is almost half that of hosts despite sizeable food aid and significant investments made by the WFP and UNHCR in cash transfers (in selected camps). The significantly lower expenditures (food and non-food) among refugees compared to the host population led to higher poverty rates. The team cross-checked the food aid received by in-camp refugees based on administrative data from the UNHCR and WFP with food consumption data from SESRE. The analysis looks at both separately for information provided by UNHCR and WFP. While WFP is not responsible for distributing non-food items such as mattresses, cooking, feeding utensils, etc., UNHCR provides non-food items; maybe mattresses were distributed in Alemwach since it is a relatively new camp. WFP’s food and cash assistance targets all individuals in refugee households. The information received from UNHCR on food aid provided to refugees in each camp includes quantities per food item per month and cash transfers per person per month for each camp and period. The food items include cereal, wheat, maize, rice, sorghum, CSB/famex (CSB+), pulse, biscuit, date biscuit, dates, oil, vegetable oil, salt, and cash (Table E.4). We have computed the per person per month in-kind aid quantities into annual values using the household size and prices from SESRE and mapped them to the closest food item in SESRE (this was not straightforward as the items are different) considering food ration change periods. The food ration scaling factor is 50 percent vs. 84 percent. Table E.1: Food aid data/information received from UNHCR Item Remark Assumptions Cereal Not clear Other cereals Wheat Matched Maize Matched Rice Matched Sorghum Matched CSB/famex (CSB+) Not in SESRE Average of other cereals/pulses Pulse Not clear Peas Biscuit Matched Date biscuit Not in SESRE Merged with biscuits Dates Matched Oil Merged with edible oil Vegetable oil Merged with edible oil Salt Matched Cash - - Source: UNHCR Based on this information, we compare how the distribution list shared what refugees should have received to what they reported regarding food consumption. The results show that refugees reported quantities lower than UNHCR food aid admin data for every item except Biscuits. Refugee households still report lower quantities, even correcting for shares indicated as sold. 124 Annexes Table E.2: Food aid and consumption comparisons Quantity (per capita/year) Expenditure (per capita/year) Items SESRE UNHCR SESRE UNHCR Nonzero All Net of sold ration* Cereals/other cereals 23.2 186.1 175.5 493 12404 Wheat 60.5 133.2 116.6 1427 4710 Maize 39.6 125.5 118.3 312 4267 Rice 21.3 48.5 39.8 560 3392 Sorghum 12.7 132.8 125.7 4 5652 Pulses 7.1 18.9 17.5 248 1514 Vegetable oil/oil 3.7 9.7 8.9 627 1774 Salt 1.6 7.9 7.7 57 232 Biscuits 5.4 4.5 4.4 16 112 Dates 0.0 4.2 3.7 0 . CSB+ 19.3 15.0 14.0 36 532 Other food 565.9 - - 2558 Peas 5.7 - - 137 All cereals 78.2 - - 2994 All pulses 8.2 - - 385 Aggregate ration/month 46.7 - - Source: UNHCR and World Bank Staff based on SESRE 2023. Valuing food aid quantities with prices from SESRE suggests that if UNHCR food aid quantities were received/ reported by refugees, refugees’ food expenditures would be much more comparable to those of hosts (Based on this information, we compare how the distribution list shared what refugees should have received to what they reported regarding food consumption. The results show that refugees reported quantities lower than UNHCR food aid admin data for every item except Biscuits. Refugee households still report lower quantities, even correcting for shares indicated as sold (Table E.2). Table E.3: Aggregate food expenditures Value/year (per capita) Value/year (per adult) Food expenditure (all) [A] 11,412 13,898 Food expenditure (UNHCR items only) [B] 3,528 4,335 Food expenditure (UNHCR in-kind) [C] 11,313 13,933 Food expenditure (UNHCR in-kind + cash) [D] 16,179 19,965 Source: UNHCR and World Bank Staff based on SESRE 2023. Note: Food expenditure (all): aggregate food expenditure. Food expenditure (UNHCR items only): aggregate food expenditure from UNHCR items only Food expenditure (UNHCR in-kind): aggregate food expenditure from UNHCR items valued using SESRE prices Food expenditure (UNHCR in-kind + cash): aggregate food expenditure from UNHCR items valued using SESRE prices plus the cash equivalent transfers 125 Annexes Food aid data received from WFP Food aid information received information from the WFP includes five food items and their quantities distributed to refugees: cereal (mainly wheat but in some camps rice), pulses (mostly yellow split peas), CSB+, vegetable oil, and salt (Table E.5). The per person per month aid (in-kind) are converted into annual values and mapped them to closest food item in SESRE (this was not straightforward as the items are different). We assumed a 50 percent ration until November 2022 (for our Oct/Nov sample) and an 84 percent after Dec 2022 (Dec, Jan, and Feb sample). The WFP data are converted to annual values using household size and prices from SESRE. The data source is the “Revised cash transfer value from Oct 2022_refugee camps” file received from the WFP document that helps to get information regarding the changes in cereal cash equivalent – data on cereal cash equivalent for cash camps which is used to calculate cereals provided in those camps and cash transfer value per year. Based on this information, we compare how the list shared with us on what refugees should have received to what they reported regarding food consumption. Table E.4: Food aid data/information received from WFP Item Remark Assumptions Cereal Not clear Mapped to wheat* Pulse Not clear Mapped to peas* Vegetable oil Mapped to edible oil CSB/famex (CSB+) Not in SESRE Average of other cereals/pulses Salt Matched Cash - - Source: UNHCR Note: Rice was distributed for some months in Afar and Somali Dollo area camps, though wheat remained the main cereal distributed. YSP (Yellow Split Pea) was the main pulse distributed among refugees and their best-preferred pulse. CSB is a corn soya blend with added essential micronutrients and vitamins called super cereal. Based on this information, we compare how the distribution list shared what refugees should have received to what they reported regarding food consumption. The results show that refugees reported quantities lower than WFP food aid admin data for every item except for CSB+ and salt (Table E.6). Even when correcting for shares indicated as sold, refugee households still reported lower quantities. Using the WFP food aid information, we found a picture similar to UNHCR’s. 126 Annexes Table E.5: Food quantity and expenditure comparisons Quantity (per capita/year) Expenditure (per capita/year) Items SESRE WFP SESRE WFP Nonzero All Net of sold ration* Cereals 82.0 77.9 2,179 Wheat 60.5 1,427 Pulse 17.1 15.6 1,366 Peas 5.7 137 Vegetable oil 3.7 5.4 5.0 627 967 CSB+ 19.3 11.1 10.1 36 394 Salt 1.6 1.8 1.7 57 63 All cereals 76.9 - - 2,994 All pulses 8.3 - - 385 Aggregate ration/month 46.7 - Source: WFP and World Bank Staff based on SESRE 2023. Note: *Net of sold ration = quantity*share of ration sold (we asked the share of ration sold in SESRE) Table E.6: Aggregate food expenditures Value/year (per capita) Value/year (per adult) Food expenditure (all) [A] 11,414 13,898 Food expenditure (WFP items only) [B] 2,226 2,741 Food expenditure (WFP in-kind) [C] 4,968 6,165 Food expenditure (WFP in-kind + cash) [D] 7,653 9,440 Source: WFP and World Bank Staff based on SESRE 2023. Average food expenditures from SESRE (13,898 birr) are higher than valued in-kind food aid reported by WFP (6,165 birr). Valuing quantities of WFP food aid with prices from SESRE suggests that if food aid quantities were received/reported, refugees’ food expenditure would be 9,440 birr slightly above the values we get in SESRE of 13,898 birr, but still low compared to hosts at 28,324 birr. The level of disaggregation of food items matters. The more disaggregated, the higher the food aggregates. To summarize, food rations received are lower than admin data suggests, regardless of the data source. Possible explanations for lower food quantities include. First, SESRE only asks one aggregate question: “How much on average of your food ration do you sell in the market”? Second, food rations are only received once a month, which may not coincide with the interview date. Yet, SESRE asks about what food they consumed (not even a list of food items), not about food received as aid. Third, refugees might carry over food aid in the future. 127 Annexes Annex F: Comparison of Results from Skills Profile Survey and SESRE Table F.1: Results on common indicators from SPS 2017 and SESRE 2023 Skills Profile Survey 2017 SESRE 2023 (In camp) Hosts Refugees Hosts Refugees South Sudanese 23% 53% Somali 24% 30% Country of origin Eritrean 25% 5% Sudan 28% Female headed 35% 66% 44% 73% Children (0 to 14) 55% 61% 47% 53% Demographics Youth (15 to 24) 13% 17% 19% 21% Adults (25 to 64) 29% 22% 31% 24% Elderly (>=65) 2% 1% 3% 2% Net primary enrollment 74% 79% 75% 69% Education Net secondary enrollment 35% 13% 39% 22% Own a house 72% 5% 69% 14% Overcrowding 32% 59% 36% 56% Living conditions Improved sources of water 96% 98% 91% 100% Access to electricity (grid) 46% 8% 41% 3% Improved toilet facility (shared/not shared) 51% 69% 38% 43% Employed 61% 22% 48% 25% Unemployed 2% 6% 9% 19% Employment Inactive, not in school 23% 44% 20% 23% Inactive, in school 14% 27% 24% 33% US$1.9 per capita per day 27% 65% Poverty incidence National Poverty line 32% 84% High food insecurity 26% 67% Food security Food insecurity scale 4.0 8.1 Economic competition 33% 49% Social cohesion Increased insecurity 37% 39% Source: Pape et al. (2018) and World Bank Staff based on SESRE 2023. 128