Informing Durable Solutions for Internal Displacement In Nigeria, Somalia, South Sudan, and Sudan Volume B: Country Case Studies Informing Durable Solutions for Internal Displacement In Nigeria, Somalia, South Sudan, and Sudan Volume B: Country Case Studies Acknowledgments This report was led by Utz Johann Pape (TTL; Senior Economist, GPV01) and written together with Ambika Sharma (Consultant, GPV01). Gonzalo Nunez (Consultant, GPV01) contributed significantly to the analysis. The Introduction chapters were written by Taies Nezam (Consultant, GPV01) and Benjamin Petrini (Consultant, GPV01). The Nigeria case study was written by Menaal Ebrahim (Consultant, GPV01) and Jacob Udo-Udo (Consultant, GPV01) with significant contributions from Felix Appler (Consultant, GPV01) and Ambika Sharma. The Somali displacement study was written by Andrea Fitri Woodhouse (Senior Social Development Specialist, GSU03), Verena Phipps (Senior Social Development Specialist, GSU07) and Ambika Sharma. The study on South Sudan was written by Taies Nezam and Ambika Sharma, with inputs from Alexander Meckelburg (Consultant, GPV01). The Sudan case study was written by Felix Appler and Alexander Meckelburg with significant inputs from Simon Lange (Jr Professional Officer, GPV01). The refugee analysis on Ethiopia was prepared by Syedah Aroob Iqbal (ET Consultant, GED03) and Benjamin Petrini. The chapter on analysis for cross-cutting policy questions was written by Ambika Sharma. Taies Nezam and Benjamin Petrini contributed signifi- cantly to the Conclusions. Sean Lothrop (Consultant, GMTLC) and Abril Rios (Consultant, GPV01) helped edit the report. The team would also like to thank Pierella Paci (Practice Manager, GPV01) as well as the peer reviewers Joanna de Berry (Senior Social Development Specialist, GTFDR), Nandini Krishnan (Senior Economist, GPV06), Nadia Piffaretti (Senior Economist, GTFSA), and Quy-Toan Do (Senior Economist, DECPI) for guidance. The team would also like to thank, for the guidance and support received, Preeti Arora (Country Program Coordinator, AFCTZ), Elsa Araya (Senior Public Sector Specialist, GGOAE), Tom Bundervoet (Senior Economist, GPV01), Giorgia Demarchi (Social Scientist, GPV02), Pablo Fajn- zylber (Adviser, GGIVP), Xavier Furtado (Resident Representative, AFMBW), Qaiser Khan (Lead Economist, GSP01), Indira Konjhodzic (Country Program Coordinator, AFCNG), Nicole Klingen (Country Program Coordinator, AFCET), Rebecca Lacroix (Social Development Specialist, GTFOS), Lynne Sherburne-Benz (Director, GSJD1), Varalakshmi Vemuru (Lead Social Development Specialist, GSU07), and Tara Vishwanath (Lead Economist, GPV03). Data collection was implemented in close collaboration with partners. Therefore, the team would like to thank the Administration for Refugee and Returnee Affairs (ARRA) in Ethiopia; Altai Consulting; the Durable Solutions Working Group in Sudan; Forcier Consulting; IOM Nigeria; the Ministry of Planning, Investment and Economic Development of the Federal Government of Somalia; JIPS; the South Sudan National Bureau of Statistics; UNDP Sudan; and Zerihun Associates. The team would like to express their gratitude specifically to Atem Bul, Isaiah Chol, Anne-Elisabeth Costaf- rolaz, Abdirahman Omar Dahir, Tom Delrue, Abebual Demilew, Matthieu Dillais, Khadra Elmi, Jedediah Fix, Aude Galli, Henry Kwenin, Margaret Labanya, Eva Lescrauwaet, Riad Marrow, Kudzani Ndlovu, David Thiang, and Tewodros Zerihun. Funding for the surveys was received from DFID, the Somalia Knowledge for Results Trust Fund of the Multi-Partner Fund, the UN-World Bank Partnership Trust Fund, Humanitarian-Development-Peace Initiative (HDPI), and the World Bank’s Forced Displacement Trust Fund. ii Table of Contents INTRODUCTION AND OVERVIEW OF DISPLACEMENT IN SUB-SAHARAN AFRICA . . . . . . . . . . . . . . . . . . 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Overview of Conflict and Displacement in Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 CASE STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 IDPs in Nigeria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 IDPs in Somalia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 IDPs in South Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 IDPs in Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 Refugees from Somalia, South Sudan, and Sudan in Ethiopia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 CROSS-COUNTRY ANALYSIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191 Agricultural IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191 IDPs in Camps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197 Inequality and Social Perceptions in Host Communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200 Geographic Dispersion and Duration of Displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204 CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 Summary of Findings from Country Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 Summary of Findings from Cross-Country Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218 Synthesizing Comparative Findings from Country Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 Evidence to Inform Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230 APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233 Appendix A. Nigeria IDP Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233 Appendix B. Somali HFS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234 Appendix C. Crisis Recovery and High Frequency Survey South Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . 240 Appendix D. Sudan IDP Profiling Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 Appendix E. Skills Profile Survey Ethiopia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246 Appendix F. Typologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250 Appendix G. Targeting: Household Classification Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263 Appendix H. Glossary of Key Analysis Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266 iii iv  |  Informing Durable Solutions for Internal Displacement List of Figures Figure A.1 IDPs and refugees hosted in Sub-Saharan Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Figure B.1 Displacement dates of IDPs by region and type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Figure B.2 Number and locations of IDPs by state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Figure B.3 Population structure for IDPs and host communities, by sex and age . . . . . . . . . . . . . . . . . . . . . . . . 23 Figure B.4 Women-headed household for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Figure B.5 Household size and dependency ratio for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . 23 Figure B.6 Relation of separated member for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Figure B.7 Reason for separation for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Figure B.8 Religion composition for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Figure B.9 Tribal composition for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Figure B.10 Reasons for leaving original place of residence of IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Figure B.11 Reasons for arriving at current location for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Figure B.12 Place of origin for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Figure B.13 Number of times residence changed for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Figure B.14 Return intentions for IPDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Figure B.15 Reasons for wanting to return to original place of residence for IPDs in camp . . . . . . . . . . . . . . . . 29 Figure B.16 Reasons for wanting to stay at current location for IDPs in host communities . . . . . . . . . . . . . . . . 29 Figure B.17 Help needed to settle in preferred location for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Figure B.18 Information needed to settle in preferred location for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Figure B.19 Poverty headcount ratio for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Figure B.20 Poverty gap relative to US$1.90 PPP (2011) poverty line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Figure B.21 Food insecurity categories for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Figure B.22 Share of aid in food consumption for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . 31 Figure B.23 Type of dwelling for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Figure B.24 Household members per room for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Figure B.25 Tenure of dwelling for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Figure B.26 Number of years dwelling owned for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Figure B.27 Access to improved sanitation facilities for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . 33 Figure B.28 Shared household toilets for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Figure B.29 Access to improved water sources for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . 34 Figure B.30 Distance to basic services for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Figure B.31 Household member who collects water for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . 35 Figure B.32 Water collection obstacles for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Figure B.33 Place where last child was delivered for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . 35 Figure B.34 Person who helped deliver last child for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . 35 Figure B.35 Primary and secondary enrollment rates for IDPs and host communities . . . . . . . . . . . . . . . . . . . . 36 Figure B.36 Highest education level for working-age populations for IDPs and host communities . . . . . . . . . . . 36 Figure B.37 Gap in primary and secondary education for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Figure B.38 Reasons for not sending primary and secondary age children to school for IDPs . . . . . . . . . . . . . 37 Figure B.39 Labor force participation among the working age for IDPs and host communities . . . . . . . . . . . . . 38 Figure B.40 Main source of livelihood for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Volume B: Country Case Studies  | v Figure B.41 Access to agricultural land for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Figure B.42 Tenure of agricultural land for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Figure B.43 Reasons for being unemployed or inactive for IDPs and host communities . . . . . . . . . . . . . . . . . . 39 Figure B.44 Support required for securing employment for IDPs and host communities . . . . . . . . . . . . . . . . . . 39 Figure B.45 IDPs have good relations with neighbors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Figure B.46 IDPs receive sufficient aid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Figure B.47 Sources of credit at short notice for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Figure B.48 Ease of borrowing at short notice for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . 40 Figure B.49 Participation in public meetings for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Figure B.50 Interaction with community leader for IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . 41 Figure B.51 Vulnerable population by status of the household . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Figure B.52 Vulnerable IDP population by state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Figure B.53 Visualization of groups from the clustering analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Figure B.54 Displacement profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Figure B.55 Source of livelihood pre-displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Figure B.56 Current source of livelihood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Figure B.57 Current housing conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Figure B.58 Perceptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Figure B.59 Return intention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Figure B.60 Reasons for moving or staying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Figure B.61 Information required to decide whether to stay or move . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Figure B.62 Number of displacements occurring by month, January 2016–April 2018 . . . . . . . . . . . . . . . . . . . 53 Figure B.63 Regional distribution of IDPs, HFS sample, and UNHCR-PRMN data . . . . . . . . . . . . . . . . . . . . . . 56 Figure B.64 Population structure for IDPs and non-IDPs by sex and age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Figure B.65 IDP profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Figure B.66 Urban/rural composition of IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Figure B.69 Reason for leaving original location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Figure B.70 Reason for arriving at current location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Figure B.71 Years since displacement and arrival in current location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Figure B.72 Conflict events and dates of displacement of conflict-driven IDPs . . . . . . . . . . . . . . . . . . . . . . . . . 61 Figure B.73 Rainfall anomalies, Gu-Deyr seasons, and displacement dates of climate-driven IDPs . . . . . . . . . 62 Figure B.74 Return intentions of IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Figure B.75 Trends in revisiting the original residence location for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Figure B.76 Push factors for IDPs who don’t want to move . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 Figure B.77 Pull factors for IDPs who want to move . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 Figure B.78 Return timeline for households that intend to move . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Figure B.79 Legal identification and access to documentation restitution mechanisms . . . . . . . . . . . . . . . . . . . 64 Figure B.80 Poverty headcount ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Figure B.81 Poverty gap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Figure B.82 Hunger incidence in the last 4 weeks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Figure B.83 Percentage of population living in improved housing, now and before displacement . . . . . . . . . . . 69 Figure B.84 Access to improved drinking water for IDPs and residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 Figure B.85 Access to improved sanitation for IDPs and residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 vi  |  Informing Durable Solutions for Internal Displacement Figure B.86 Number of households sharing a toilet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Figure B.87 Households more than 30 minutes from services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Figure B.88 Access to electricity to charge mobile phone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Figure B.89 Under 15 minutes to network reception point . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Figure B.90 Births in health facilities for IDPs, hosts, and residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Figure B.91 Births attended by skilled health staff for IDPs, hosts, and residents . . . . . . . . . . . . . . . . . . . . . . . 72 Figure B.92 Adult literacy rate by sex, IDPs, and residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Figure B.93 School enrollment among the school age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Figure B.94 Labor force participation for IDPs and residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Figure B.95 Changes in employment activity after displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Figure B.96 Proportion of women perceived to be allowed to work outside the home . . . . . . . . . . . . . . . . . . . . 75 Figure B.97 Reasons for economic inactivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Figure B.98 Main employment activity for IDPs, hosts, and rural residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Figure B.99 Main source of household income for IDPs, hosts, and residents . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Figure B.100 Average remittances for IDPs, hosts, and residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Figure B.101 Perceptions of safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Figure B.102 Perceived relations with surrounding community . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Figure B.103 Vulnerable population by status of the household . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Figure B.104 Vulnerable IDP population by pre-war region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Figure B.105 Visualization of the two clusters from the clustering analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Figure B.106 Reasons for displacement from origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Figure B.107 Reasons for coming to current location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 Figure B.108 Closeness to basic facilities, agricultural land access and housing at origin . . . . . . . . . . . . . . . . . 81 Figure B.109 Primary source of household income at origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 Figure B.110 Poverty and current living conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Figure B.111 Primary source of household income, current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Figure B.112 Intention and timeline to move . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 Figure B.113 Factors guiding a decision to settle in the future . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 Figure B.114 HFS 2017 and CRS 2017 coverage (pre-war states) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Figure B.115 Population structure for IDPs and urban residents, by sex and age . . . . . . . . . . . . . . . . . . . . . . . 92 Figure B.116 Ethnic composition for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Figure B.117 Reasons for leaving original location for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Figure B.118 Reasons for arriving at current location for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Figure B.119 Conflict events and displacement dates for IDPs, January 2013–July 2017 . . . . . . . . . . . . . . . . . 94 Figure B.120 IDPs’ place of origin, by state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Figure B.121 IDPs’ place of origin vs. current location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Figure B.122 Reasons for separation of household members for IDPs and urban residents . . . . . . . . . . . . . . . 96 Figure B.123 Return intentions of IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Figure B.124 Reasons for staying in current location for IDPs who do not intend to relocate . . . . . . . . . . . . . . . 97 Figure B.125 Reasons for moving to new location for IDPs who intend to relocate . . . . . . . . . . . . . . . . . . . . . . 97 Figure B.126 Poverty headcount ratio for IDPs and residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Figure B.127 Poverty gap relative to US$1.90 PPP (2011) per day per capita poverty line for IDPs and residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Volume B: Country Case Studies  | vii Figure B.128 Frequency of facing hunger in the past four weeks for IDPs and urban residents . . . . . . . . . . . . 100 Figure B.129 Food aid and core food consumption, per capita per day for IDPs and urban residents . . . . . . . 100 Figure B.130 Access to improved housing, now and pre-displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Figure B.131 Trends in tenure of housing, now and pre-displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Figure B.132 Trends in access to improved water and sanitation for IDPs and urban residents . . . . . . . . . . . . 101 Figure B.133 Time (one way) to amenities for IDPs and urban residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Figure B.134 Access to improved sanitation accounting for sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Figure B.135 Crowding in dwellings for IDPs and urban residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Figure B.136 Literacy rates, 15 years and above for IDPs, urban, and rural residents . . . . . . . . . . . . . . . . . . . 103 Figure B.137 Adult educational attainment for IDPs and urban residents, by sex . . . . . . . . . . . . . . . . . . . . . . . 103 Figure B.138 Enrollment rates for school-age children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Figure B.139 Reasons for not attending secondary school . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Figure B.140 Labor activity status among the working age (15–64 years) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Figure B.141 Primary employment activity for IDPs and residents, now and before December 2013 . . . . . . . 106 Figure B.142 Main source of livelihood for IDPs and residents, currently and before December 2013 . . . . . . 106 Figure B.143 Holdings of agricultural land, currently and pre-December 2013 . . . . . . . . . . . . . . . . . . . . . . . . 107 Figure B.144 Holdings of livestock units, currently and pre-December 2013 . . . . . . . . . . . . . . . . . . . . . . . . . . 107 Figure B.145 Ownership of at least one productive asset, currently and pre-December 2013 . . . . . . . . . . . . . 107 Figure B.146 Trends in perceived safety for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 Figure B.147 Trends in perceived safety for IDPs and urban residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 Figure B.148 Trends in exposure to violence after December 2013 for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . 109 Figure B.149 Relations with neighbors within the camp for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 Figure B.150 Relations with host communities outside the camps for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 Figure B.151 Frequency of attending public meetings for IDPs and urban residents . . . . . . . . . . . . . . . . . . . . 110 Figure B.152 Vulnerable population by status of the household . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 Figure B.153 Vulnerable IDP population by (pre-war) state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 Figure B.154 Visualization of groups from the clustering analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Figure B.155 (Pre-war) state of origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Figure B.156 Current location relative to origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Figure B.157 Main source of livelihood pre-displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Figure B.158 Perceptions of current situation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 Figure B.159 Return intention and timing of moving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 Figure B.160 Information required to decide whether to stay or move . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 Figure B.161 Nominal wholesale prices in Khartoum and Al Fashir (SDG per kg), January 2016–July 2018 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 Figure B.162 Conflict events in Sudan since 2003, percentage by state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 Figure B.163 Camp comparison: dwelling made of mud . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 Figure B.164 Camp comparison: primary income is nonagricultural labor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 Figure B.165 Population structure for IDPs and urban populations, by sex and age . . . . . . . . . . . . . . . . . . . . 122 Figure B.166 First and second most important reasons for leaving original location for IDPs . . . . . . . . . . . . . 124 Figure B.167 First and second most important reasons for settling in camp for IDPs . . . . . . . . . . . . . . . . . . . . 125 Figure B.168 Conflict events and displacement dates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 Figure B.169 Trends in travelling to current location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 viii  |  Informing Durable Solutions for Internal Displacement Figure B.170 Places of origin of IDPs, North Darfur state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 Figure B.171 Years since displacement and arrival at current location for IDPs . . . . . . . . . . . . . . . . . . . . . . . . 126 Figure B.172 Trends in having returned since replacement for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Figure B.173 Trends in return intentions for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Figure B.174 Reasons for staying in current location for IDPs who do not want to relocate . . . . . . . . . . . . . . . 128 Figure B.175 Reasons for not moving to a new location for IDPs who do not want to relocate . . . . . . . . . . . . 128 Figure B.176 Reasons for moving to a new location for IDPs who want to relocate . . . . . . . . . . . . . . . . . . . . . 129 Figure B.177 Reasons for moving away from the camp for IDPs who want to relocate . . . . . . . . . . . . . . . . . . 129 Figure B.178 Poverty headcount ratio for IDP and host populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 Figure B.179 Poverty gap relative to US$1.90 PPP (2011) poverty line for IDP and host populations . . . . . . . 132 Figure B.180 Food insecurity for IDP and host populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 Figure B.181 Trends in coping strategies when hungry for IDP and host populations . . . . . . . . . . . . . . . . . . . 134 Figure B.182 Type of dwelling for IDP and host populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Figure B.183 Overcrowding: individuals per room, IDP and host populations . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Figure B.184 Ownership of dwelling for IDP and host populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Figure B.185 Source of lighting for IDP and host populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 Figure B.186 Access to improved water and sanitation, IDP and host populations . . . . . . . . . . . . . . . . . . . . . 136 Figure B.187 Time (one way) to amenities for IDP and host populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 Figure B.188 Literacy rates, 15 years and above for IDP and host populations . . . . . . . . . . . . . . . . . . . . . . . . 138 Figure B.189 Net attendance rates, primary and secondary school for IDP and host populations . . . . . . . . . . 138 Figure B.190 Adult educational attainment for IDP and host populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 Figure B.191 Labor force participation and employment status among working-age population . . . . . . . . . . . 139 Figure B.192 Primary employment activity for IDP and host populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 Figure B.193 Main sources of livelihoods for IDP and host populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 Figure B.194 Access to productive assets for IDP and host populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 Figure B.195 Perceptions of relationships between host and IDP communities for IDP and host populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 Figure B.196 Frequency of attending a public meeting for IDP and host populations . . . . . . . . . . . . . . . . . . . 143 Figure B.197 Frequency of interacting with a community leader for IDP and host populations . . . . . . . . . . . . 143 Figure B.198 Perceptions of safety, walking in the neighborhood for IDP and host populations . . . . . . . . . . . . 143 Figure B.199 Ownership of legal identification for IDP and host populations . . . . . . . . . . . . . . . . . . . . . . . . . . 144 Figure B.200 Reasons for not having legal identification for IDP and host populations . . . . . . . . . . . . . . . . . . 144 Figure B.201 Vulnerable population by status of the household . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 Figure B.202 Vulnerable IDP population by camp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 Figure B.203 Visualization of groups from the clustering analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 Figure B.204 District of origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 Figure B.205 Year and reason of displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 Figure B.206 Current source of livelihood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 Figure B.207 Current poverty and food insecurity status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 Figure B.208 Perception of current conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 Figure B.209 Return intention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Figure B.210 Reasons for moving or staying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Figure B.211 Timing of moving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Volume B: Country Case Studies  | ix Figure B.212 Refugee population in Ethiopia by country of origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154 Figure B.213 Map of Ethiopia, refugee camps and heat map of conflict events since 1997 . . . . . . . . . . . . . . 156 Figure B.214 Reasons for displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 Figure B.215 Reasons for settling in the current location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 Figure B.216 Demographics of refugee population in Ethiopia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Figure B.217 Sex of household head . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Figure B.218 Dependency ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Figure B.219 Percentage of refugee population separated from their household members . . . . . . . . . . . . . . . 162 Figure B.220 Reasons for separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162 Figure B.221 Demographics of separated population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 Figure B.222 Poverty incidence of refugee groups, countries of origin, and host communities . . . . . . . . . . . . 164 Figure B.223 Poverty gap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 Figure B.224 Food insecurity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Figure B.225 Aid as a share of food consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Figure B.226 Ownership of dwelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Figure B.227 Housing conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Figure B.228 Overcrowding (4 or more persons per room) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 Figure B.229 Source of lighting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168 Figure B.230 Access to improved sources of water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Figure B.231 Access to improved sanitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Figure B.232 Net primary and secondary enrollment rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Figure B.233 Mean time (minutes) taken (one way) to fetch water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 Figure B.234 Time (minutes) to walk one way to health facility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 Figure B.235 Time (minutes) to walk one way to primary school . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Figure B.236 Source of livelihood currently and before displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Figure B.237 Labor force participation and employment status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Figure B.238 Highest educational attainment for working-age population . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Figure B.239 Reasons for not participating in the labor force (refugees) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Figure B.240 Reasons for not securing employment (refugees) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Figure B.241 Access to productive assets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 Figure B.242 Primary employment activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 Figure B.243 Sector of employment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 Figure B.244 Interpersonal relations between refugees and host community . . . . . . . . . . . . . . . . . . . . . . . . . . 178 Figure B.245 Host community feelings: good relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 Figure B.246 Host community feelings: other measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 Figure B.247 Participation in public meetings, last 12 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 Figure B.248 Interaction with community leader, last 12 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 Figure B.249 Feelings of safety and security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 Figure B.250 Movement and return plans of refugees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Figure B.251 Reasons to stay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Figure B.252 Main support needed to settle in preferred location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Figure B.253 Vulnerable population by status and country of origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 Figure B.254 Visualization of groups from the clustering analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 x  |  Informing Durable Solutions for Internal Displacement Figure B.255 Country of origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 Figure B.256 Year and reason for displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 Figure B.257 Current assistance and food insecurity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 Figure B.258 Current poverty status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 Figure B.259 Current perceptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186 Figure B.260 Return intention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187 Figure B.261 Reasons for moving or staying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187 Figure B.262 Timing of moving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187 Figure C.1 Livelihoods of the displaced at origin, compared to host communities now . . . . . . . . . . . . . . . . . . 192 Figure C.2 Labor force participation status of working-age IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193 Figure C.3 Current livelihoods of IDPs and refugees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194 Figure C.4 Poverty headcount ratio for IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195 Figure C.5 IDPs’ perception of relations with surrounding host communities . . . . . . . . . . . . . . . . . . . . . . . . . . 195 Figure C.6 IDPs’ return intentions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196 List of Appendix Figures Figure F1 Clusters of households in 2D and 3D for Ethiopia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250 Figure F2 Housing conditions pre-displacement for Ethiopia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250 Figure F3 Source of livelihood pre-displacement for Ethiopia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251 Figure F4 Current access to key services for Ethiopia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252 Figure F5 Information required to decide whether to stay or move for Ethiopia . . . . . . . . . . . . . . . . . . . . . . . . 253 Figure F6 Clusters of households in 2D and 3D for Nigeria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254 Figure F7 Timing of moving for Nigeria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254 Figure F8 Clusters of households in 2D and 3D for Somalia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256 Figure F9 Clusters of households in 2D and 3D for South Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 Figure F10 Camp for South Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 Figure F11 Reasons for displacement from origin in South Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 Figure F12 Assets and housing pre-displacement in South Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258 Figure F13 Changes in water and sanitation conditions in South Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259 Figure F14 Clusters of households in 2D and 3D for Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260 Figure F15 Household characteristics pre-displacement in Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260 Figure F16 Current household characteristics in Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261 Figure F17 Information required to decide whether to stay or move in Sudan . . . . . . . . . . . . . . . . . . . . . . . . . 262 List of Tables Table A.1 Durable solutions analysis: criteria and themes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Table A.2 Location of displaced populations surveyed for this study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Table A.3 Number of forcibly displaced and host community households interviewed . . . . . . . . . . . . . . . . . . . . . 9 Table A.4 Twelve largest displacement situations globally . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Table A.5 Extent of forced displacement in concerned countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Table A.6 Overview of the study’s forced displacement situations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Volume B: Country Case Studies  | xi Table B.1 Sample size and comparison groups from the survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Table B.2 Determinants of return intentions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Table B.3 Current household characteristics and poverty status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Table B.4 Age dependency ratios and household size by sex of household head . . . . . . . . . . . . . . . . . . . . . . . 57 Table B.5 Return intention of IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Table B.6 Demographic attributes of poor households . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Table B.7 Demographic attributes of IDPs and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Table B.8 Household characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 Table B.9 Dependency ratio and household size, by sex of household head . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Table B.10 Trends in separation for IDPs and urban residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Table B.11 Return intention of IDPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Table B.12 Demographic attributes of poor households . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Table B.13 Demographic attributes of IDPs and urban residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Table B.14 Current household characteristics and poverty status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Table B.15 Summary of factors motivating mobility intentions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 Table B.16 Household composition for IDP and host populations, by sex and household head . . . . . . . . . . . . 124 Table B.17 Trends in separation for IDP and host populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Table B.18 Determinants of wealth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Table C.1 Differences in overall IDP outcomes with country fixed effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196 Table C.2 Outcomes of camp IDPs, non-camp IDPs, and hosts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 Table C.3 Links between host community characteristics and host community perceptions . . . . . . . . . . . . . . 203 Table C.4 Effect of displacement duration and geographic distance from origin on IDP outcomes . . . . . . . . . 206 List of Appendix Tables Table B1 Sample overview of Somali HFS 2017–18 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 Table C1 Heterogeneity among IDP households . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240 Table C2 Sample characteristics: CRS South Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 Table C3 Sample characteristics: high frequency survey South Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242 Table D1 Planned and listed interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244 Table E1 Number of refugees and host community households interviewed by stratum . . . . . . . . . . . . . . . . . 247 Table E2 Sampled population by country of nationality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248 Table E3 Refugee camps in sample frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248 Table F1 Size of each group of refugees in Ethiopia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250 Table F2 Main source of income of groups in Ethiopia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251 Table F3 Current household characteristics for Ethiopia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252 Table F4 Poverty and household characteristics of groups in Ethiopia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252 Table F5 Living conditions of groups in Ethiopia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253 Table F6 Size of each group of IDPs in Nigeria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254 Table F7 Main source of income of groups in Nigeria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255 Table F8 Poverty and household characteristics of groups in Nigeria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255 Table F9 Living conditions of groups in Nigeria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256 Table F10 Size of each group of IDPs in Somalia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256 xii  |  Informing Durable Solutions for Internal Displacement Table F11 Size of each group of IDPs in South Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 Table F12 Main source of income of groups in South Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258 Table F13 Poverty and household characteristics of groups in South Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . 259 Table F14 Living conditions of groups in South Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259 Table F15 Size of each group of IDPs in Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260 Table F16 Current household composition in Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261 Table F17 Main source of income of groups in Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261 Table F18 Poverty and household characteristics of groups in Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261 Table F19 Living conditions of groups in Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262 List of Boxes Box A.1 Data limitations in studying forced displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Box A.2 Definitions of key displacement terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Box B.1 Nigeria IDP survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Box B.2 Where are the IDPs? Timing of survey sampling and interpretation of spatial results . . . . . . . . . . . . . 55 Box B.3 Drivers of displacement in Somali regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Box B.4 The CRS collects rich micro-data on IDPs to complement the HFS 2017 . . . . . . . . . . . . . . . . . . . . . . . 91 Box B.5 The IDP Profiling Survey Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 Box B.6 The Abu Shouk and El Salam IDP camps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 Box B.7 The Darfur conflict since 2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130 Box B.8 SPS with refugees and host communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 Box B.9 GoE’s nine pledges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 Abbreviations ACLED Armed Conflict Location and Event Data ARRA Administration for Refugees and Returnee Affairs AUHIP African Union High-level Implementation Panel CBS Central Bureau of Statistics Sudan CPA Comprehensive Peace Agreement CRRF Comprehensive Refugee Response Framework CRS Crisis Recovery Survey DFID U.K. Department for International Development DLM Darfur Liberation Movement DTM Displacement Tracking Matrix EA Enumeration Area EB Enumeration Block FAO Food and Agriculture Organization (of the United Nations) FGS Federal Government of Somalia FMS Federal Member State GBV Gender-based Violence GDP Gross Domestic Product GoE Government of Ethiopia GoN Government of Nigeria GoS Government of Sudan GoSS Government of South Sudan GIS Geographic Information System HFSSS W4 High Frequency Survey Wave 4 IASC Inter-Agency Standing Committee ICC International Criminal Court IDP Internally Displaced Person IDMC Internal Displacement Monitoring Center IGAD Intergovernmental Authority on Development IOM International Organization for Migration IPC Integrated Phase Classification IPS Inter Press Service ISWAP West African Province of ISIS JEM Justice and Equality Movement LGA Local Government Area MCA Multiple Correspondence Analysis NCFRMI National Commission for Refugees, Migrants and Internally Displaced Persons NDMF National Disaster Management Framework xiii xiv  |  Informing Durable Solutions for Internal Displacement NEMA National Emergency Management Agency NGO Nongovernmental Organization ONLF Ogaden National Liberation Front PDF Popular Defense Forces PESS Population Estimation Survey 2014 PoC Protection of Civilian PPP Purchasing Power Parity R-ARCSS Revitalized Agreement on the Resolution of Conflict in South Sudan rCSI Reduced Coping Strategies Index SDG Sustainable Development Goal SEMA State Emergency Management Agencies SHFS Somali High Frequency Survey SLA/MM SLA faction under Minni Minawi SLM Sudan Liberation Movement SNNPR Southern Nations, Nationalities, and People’s Region SPLA/M Southern Peoples Liberation Army/Movement SPLM-IO Southern Peoples Liberation Movement—In Opposition SPS Skills Profile Survey UN United Nations UNFPA United Nations Population Fund UNHCR United Nations High Commissioner for Refugees UNICEF United Nations Children’s Fund UNMISS United Nations Mission in South Sudan UN OCHA United Nations Office for the Coordination of Humanitarian Affairs VAM Vulnerability Analysis Mapping Unit WASH Water, Sanitation, and Hygiene WFP World Food Programme Introduction and Overview of Displacement in Sub-Saharan Africa Introduction Background, Objectives, and Approach 1.  Internal displacement represents one of the most adverse symptoms of violent conflict affecting people’s lives. In parallel, displacement both deepens and cements vulnerability and poverty dynamics of those affected. Forced displacement includes two main categories of people: first, refugees and asylum seekers who cross an internationally recognized border; and, second, internally displaced persons (IDPs) resulting from conflict and violence. While people forcibly displaced due to either natural hazards or other environmental factors are present in some of the analyses (for example, Somalia and South Sudan), the report mostly concerns conflict-induced IDPs.1 The analysis also covers non-displaced host communities that live in the vicinity of IDPs and refugees for comparison. Recent approaches rec- ognize the critical importance of including host communities in development policies toward IDPs because of the mutual benefits that an integrated approach can bring to both groups. Including host communities also stems from the reality that displaced people are often in a protracted situation. Development-oriented policies for displaced peo- ple may benefit host communities and vice versa. 2.  This report analyzes socioeconomic micro-level data collected in four internal displacement situations in Sub-Saharan Africa. The quantitative study is based on household-level surveys conducted with both forced dis- placed populations and host communities. Geographically, the report covers a cluster of forced displacement situa- tions in Eastern African countries, including IDPs in Nigeria, Somalia, South Sudan, and Sudan. Refugees in Ethiopia from four nationalities including Somalia, South Sudan, and Sudan are also surveyed. These represent some of the worst displacement situations globally. Except for Ethiopia, the other four countries are some of the most conflict-affected states in Sub-Saharan Africa as well as globally, having experienced increasing numbers of war fatalities.2 All of them are among the top 12 countries for IDPs globally. Similarly, they are also at the top as either refugee-hosting countries (Ethiopia and Sudan) or countries of origin for refugees (South Sudan and Somalia).3 3.  Over the last decade, displacement has progressively become a priority for policy makers. The recognition that development solutions are needed beyond humanitarian assistance has been reinforced by the spike in the num- ber of displaced populations since 2010—a trend that has run in parallel to the dramatic increase in the number and intensity of violent conflict. Skyrocketing spending took place at the same time of increasing trends of both vio- lent conflict and forced displacement since 2010. The cost of humanitarian aid has dramatically increased in the last 15 years. From US$7.2 billion in 2000, the cost of humanitarian assistance tripled, reaching US$21.8 billion in 2015.4 To sustainably respond to emerging crises and protracted ones and to ease skyrocketing humanitarian costs, the 1. With the exception of Somali IDPs, many of whom are drought displaced. 2. United Nations and World Bank. 2018. “Pathways for Peace: Inclusive Approaches to Preventing Violent Conflict.” 3. IDMC and NRC. 2018. “Global Report on Internal Displacement 2018”; UNHCR (United Nations High Commissioner for Refugees). 2018b. “Global Trends: Forced Displacement in 2017.” 4. World Bank. 2017. “Forcibly Displaced: Toward a Development Approach Supporting Refugees, the Internally Displaced, and Their Hosts.” 1 2  |  Informing Durable Solutions for Internal Displacement international community has increasingly advocated for a development response to displacement situations beyond and in coordination with the humanitarian level. Protracted displacement situations, in particular, require development instruments to tackle challenges with a medium- to long-term horizon. There is increasing recognition that—while critical at the onset—humanitarian assistance is unable to address the socioeconomic dimensions of displacement, including access to livelihoods and employment. Economic self-reliance, as opposed to aid dependency, is the basis of the development approach to forced displacement, with the objective of ending displacement by finding durable solutions.5 4.  The unsustainability of humanitarian assistance and short-term support highlights the need for embrac- ing a developmental approach to support forcibly displaced populations worldwide. Most displacement crises are protracted and need a shift towards development policy interventions rather than maintaining a humanitarian approach. High-level international initiatives to address the development challenges of forced displacement include the universally adopted 2016 ‘New York Declaration for Refugees and Migrants’, which spearheaded the global com- pact on the displaced. The New York Declaration lays out a new comprehensive and integrated approach to address displacement crises, the Comprehensive Refugee Response Framework (CRRF).6 This approach commits to share inter- nationally the responsibility to address forced displacement and its causes, ease pressure on host countries, heighten self-reliance as a way to sustainably support forced displaced people, and promote practices that jointly benefit both displaced persons and host communities. 5.  Such increased attention on displaced populations has led to the production and analysis of data to better understand displacement—however, multidimensional data gaps prevent an assessment of socioeconomic conditions among displaced populations. Traditionally, forced displacement exclusively concerns the protection of IDPs in terms of humanitarian, legal, and security issues. The analysis of displaced people’s poverty levels and standards of living has been marginal—especially in Sub-Saharan Africa. Efforts to include IDPs’ needs in the Sustainable Develop- ment Goals (SDGs) agenda are still hampered by data gaps.7 It is only recently that policy makers have become inter- ested in addressing displaced people’s socioeconomic vulnerabilities, which are directly linked to their displacement condition. Evidence-based development policies to address socioeconomic vulnerabilities have the potential to unlock economic dependence of IDPs and refugees by strengthening displaced people’s agency, self-reliance, and synergies with host communities. Thus, addressing data gaps and limitations (Box A.1) is crucial to inform targeted policy inter- ventions for durable solutions. 5. While more efforts to include development policies are taking place, the fact that the cost of humanitarian aid has continued to grow in the last few years means that humanitarian instruments are still the most prevalent ones and are being used somewhat inefficiently to tackle both emerging and protracted situations. 6. Global Digital Portal. 2018. “Comprehensive Refugee Response Framework (CRRF)”; UN General Assembly. 2016. “New York Declaration for Refugees and Migrants.” 7. International Peace Institute. 2018. “Reaching Internally Displaced Persons to Achieve the 2030 Agenda.” Volume B: Country Case Studies  | 3  BOX A.1    Data limitations in studying forced displacement A critical part of the agenda on forced displacement is the availability of comprehensive, reliable, and comparable data. Data inform the development of policy and programming to address the protection and assistance needs of displaced populations as well as public opinion and political discourse. However, the data that are currently available have many limita- tions, especially with respect to IDPs.8 Gathering accurate information on IDPs is challenging. The Internal Displacement Monitoring Center (IDMC) compiles, ana- lyzes, and disseminates data on IDPs collected by governments, international organizations (UNHCR, Office for the Coordination of Humanitarian Affairs [OCHA], and International Organization of Migration (IOM), nongovernmental organizations (NGOs), and the media. UNHCR collects, analyzes, and disseminates data on refugees. It also relies on data provided by host governments. The definitions and methods used by these groups vary across countries and over time. This usually means that aggregate num- bers are inconsistent and disaggregated numbers are not available, particularly for IDPs.9 There is incomplete data for most countries. Fluid population movements combine with security and other concerns (logis- tic, financial, and so on) to make primary data collection difficult in many areas. This often means that data are collected only for stable and accessible areas such as camps with a large number of displaced people (South Sudan). This also means that ‘invisible’ refugees and IDPs like those living in private accommodations among the host community are not captured (Sudan).10 There is also insufficient data on inflows (those who are counted as displaced) and outflows (those who are no longer counted as displaced). Inflows include new displacements and those born in displacement. Outflows include those who have returned, integrated locally, or settled elsewhere (that is, achieved durable solutions) as well as those who have crossed borders and become refugees (in the case of IDPs). This process is made more difficult because there is no clear and measurable ‘end’ to most displacement situations, especially for IDPs. Outflows also include those who have died in displacement.11 There are other gaps that are of particular concern to development actors. There is little information on returnees and host communities. There is also little information on the socioeconomic needs of displaced populations and impacts of displace- ment on host communities. These include needs that are particular to displaced populations, as well as those that are shared with the host communities.12 This study tries to fill some of these gaps by collecting socioeconomic data on IDPs, refugees, and their hosts. 6.  The present study helps close important socioeconomic data gaps, while addressing the issue of data reli- ability by reducing response inaccuracy and misreporting through behavioral science. While data on displace- ment suffer from fragmentation and lack of standardized collection methods, the report focuses on addressing some of the practical challenges in collecting quantitative data in displacement and fragile contexts. These challenges stem from the unstable and volatile nature of displaced people’s camps. Not only is there a scarcity of micro-level quantita- tive studies that hamper efforts to build evidence to inform policies, but also the traditional data collection methods are insufficient when working in these fragile and displacement-affected situations.13 Thus, methodologically, this study 8. World Bank. 2017. “Forcibly Displaced: Toward a Development Approach Supporting Refugees, the Internally Displaced, and Their Hosts.” 9. Sarzin. 2017. “Stocktaking of Global Forced Displacement Data.” 10. Ibid. 11. IDMC and NRC. 2018. “Global Report on Internal Displacement 2018.” 12. Sarzin. 2017. “Stocktaking of Global Forced Displacement Data.” 13. World Bank. 2017. “Forcibly Displaced: Toward a Development Approach Supporting Refugees, the Internally Displaced, and Their Hosts.” 4  |  Informing Durable Solutions for Internal Displacement proposes and adopts innovative behavioral nudges in survey design that can reduce underreporting of consumption patterns, and ultimately mitigate potential bias. 7.  Ultimately, the study’s objective is to inform policy and programs for durable solutions to forced displace- ment through the analysis of micro-level quantitative data collected for both displaced and non-displaced populations. Durable solutions are an array of sustainable and positive outcomes that permanently put an end to a displacement situation.14 There are three durable solutions: return to the country or area of origin, locally integrate into the country or area of displacement, and resettle to a third country or area. There is also a need for strong programming to overcome displacement-related vulnerability, even before a durable solution is reached. The study aims to fill both the data and the knowledge gap on the socioeconomic conditions of displaced populations and contribute to policies that can help overcome displacement-related vulnerabilities and thus be conducive to durable solutions. It does so through the collection of quantitative micro-data through household-level surveys with displaced and non-displaced communities. Due to the scarcity of quantitative socioeconomic data on IDPs, the knowledge generated constitutes a unique evidence base to inform policies. The household survey and ensuing analysis provide a comprehensive snap- shot of the demographic characteristics, standards of living, and perceptions on intentions, future prospects, and ties between displaced and non-displaced persons. 8.  The present study compares the welfare level of selected displaced populations with host communities or local residents. Through parallel household-level data on socioeconomic indicators for both displaced populations and host communities, the study subscribes to an integrated approach to displacement situations. While the study compares the differences in scope and scale between the needs of displaced people and those of host communities, both are articulated in a joint manner that emphasizes common needs, dynamics, and potential policies, especially on service delivery. Because displaced people and especially refugees are often hosted in geographically marginalized and lagging areas of the country, a comprehensive approach to displacement situations is also one that highlights the development needs of host communities. In the displaced people–host community joint dynamics, it is critical to examine the potential negative impacts that the displaced may create on the local labor market, the housing market, and service delivery. Thus, comprehensive approaches should strengthen potential synergies and mutual benefits in the socioeconomic space. 9.  IDPs are poorer and more vulnerable than host communities. In the case studies analyzed for this report, pov- erty incidence is usually higher for IDPs compared to host communities. Insights on poverty drivers and living con- ditions contribute to both a deeper understanding of displacement dynamics in general and to specific potential policy solutions that target both groups. Household-level data also reveal the intentions, priorities, and preferences of displaced groups, which, in turn, provides critical insights on specific vulnerabilities and on how the displaced think about durable solutions. Ultimately, such an evidence base provides the rationale for development actors’ involvement in these areas. 14. A durable solution is achieved when the displaced person no longer has vulnerabilities associated with his/her displacement (see below for an exhaustive discussion on the dimensions of durable solutions). Volume B: Country Case Studies  | 5   BOX A.2    Definitions of key displacement terms The Guiding Principles on Internal Displacement define IDPs as those “who have been forced or obliged to flee or to leave their homes or places of habitual residence, in particular as a result of or in order to avoid the effects of armed conflict, situations of generalized violence, violations of human rights or natural or human-made disasters, and who have not crossed an internation- ally recognized state border.”15 Host communities are people whose welfare is affected by the presence of displaced populations, namely refugees, IDPs, and returnees.16 Returnees are former refugees who returned to their country of origin or former IDPs who returned to their area of origin or habitual residence.17 The 1951 Convention Relating to the Status of Refugees (1951 Convention) and the 1967 Protocol define a refugee as a person who “owing to well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his [or her] nationality and is unable or, owing to such fear, is unwilling to avail him [or her] self of the protection of that country.”18 People who flee poverty, famine, or natural disasters are not considered refugees. Durable Solutions 10.  According to the Inter-Agency Standing Committee (IASC) Framework on Durable Solutions for Internally Displaced Persons (IASC Framework), IDPs achieve durable solutions when they no longer have “assistance and protection needs that are linked to their displacement and can enjoy their human rights without discrim- ination on account of their displacement.” This can be achieved through (a) sustainable reintegration at the place of origin (return), (b) sustainable integration at the place of refuge (local integration), and (c) sustainable integration in another party of the country (settlement elsewhere).19 The IASC Framework extends the Guiding Principles on Internal Displacement and builds on the work of UNHCR and others on durable solutions for refugees. 11.  The World Bank defines the ‘end point’ of displacement as the moment when the displaced no longer needs dedicated development assistance. Previous to the ‘end point’, they need assistance because they have vulnerabil- ities linked to their displacement. These vulnerabilities make it difficult for them to seize development opportunities. When they are able to overcome these vulnerabilities and take advantage of broader poverty reduction programs, they have attained a durable solution. Hence, for development actors, the focus is not on where people live. The socioeco- nomic approach complements the traditional legal and rights frameworks favored by UNHCR and the IASC.20 12.  Achieving a durable solution involves not only socioeconomic rehabilitation but also a conducive security, policy, and legal environment; the report informs the former but does not analyze the latter in detail. The anal- ysis in this report draws socioeconomic profiles of IDPs and refugees at the household and individual levels. However, a 15. OCHA. 1998. “Guiding Principles on Internal Displacement.” 16. World Bank. 2017. “Forcibly Displaced: Toward a Development Approach Supporting Refugees, the Internally Displaced, and Their Hosts.” 17. UNHCR. 2017. “Statistical Yearbook 2016.” 18. UNHCR. 2011. “Convention and Protocol Relating to the Status of Refugees.” 19. IASC. 2010. “IASC Framework on Durable Solutions for Internally Displaced Persons.” 20. World Bank. 2017. “Forcibly Displaced: Toward a Development Approach Supporting Refugees, the Internally Displaced, and Their Hosts.” 6  |  Informing Durable Solutions for Internal Displacement detailed analysis of the policy and legal contexts that influence durable solutions is outside the scope of the report. The case studies cover a country context that also touches upon legal or policy frameworks to help embed the IDPs’ and refugees’ socioeconomic profiles in the larger displacement narrative of the country. Analytical Framework 13.  The analytical framework developed for this study is based on the definitions, principles, and criteria in the IASC Framework on Durable Solutions for Internally Displaced Persons.21 The IASC Framework has recently been operationalized with the Analysis Guide and Durable Solutions Indicator Library.22 Durable solution analysis is a systematic process of measuring progress toward durable solutions to internal displacement and identifying priorities for action. It has three essential components: (a) displaced persons’ durable solution preferences, (b) eight durable solu- tion criteria, and (c) a demographic profile of the displaced population. 14.  Although the physical location of the displaced person is not a solution in itself, it does form an important part of the larger analysis of their preferences and priorities. The three settlement options (return, local inte- gration, and settlement elsewhere) are not static or definitive. Mobility can be an important coping mechanism for displaced people or form a crucial part of the solution (where some members of the household return to the place of origin while others stay in the place of refuge). Where the preferred solution is not possible, durable solution analysis can inform activities designed to reduce the adverse consequences of displacement. 15.  The analysis is guided by the durable solutions indicator framework, while the policy insights focus on overcoming displacement-induced vulnerabilities. The IASC Framework helps measure whether a durable solution has been reached. The core criteria for durable solutions are (a) long-term safety, security, and freedom of movement; (b) an adequate standard of living, including access to adequate food, water, housing, health care, and basic education; (c) access to employment and livelihoods; and (d) access to effective mechanisms that restore housing, land, and prop- erty or provide compensation for lost assets. These core criteria are analyzed to draw IDP profiles in the four countries. The core criteria are complemented by measures on social capital, and factors influencing IDPs’ return intentions are also analyzed. The Indicator Library operationalizes these criteria into measurable indicators (Table A.1). When analyzed in comparison to non-displaced populations, these indicators identify vulnerabilities linked to displacement. 16.  Hosts and other non-displaced communities are included in the IASC Framework–based analysis to pro- vide a comparison group for understanding IDPs’ vulnerabilities. While the IASC Framework is designed to address specific needs and challenges of IDPs, the micro-data provides rich socioeconomic information about host and other resident communities. Thus, socioeconomic profiles of these non-displaced populations are also drawn using the IASC analytical indicators to provide a comparison for IDPs. For the sake of comparison, refugee profiles are also drawn using the IASC indicators, enhancing and tailoring the analysis as required. 21. IASC. 2010. “IASC Framework on Durable Solutions for Internally Displaced Persons.” 22. JIPS (Joint IDP Profiling Service). 2018. “Durable Solutions Analysis Guide: A Tool To Measure Progress Toward Durable Solutions to Internal Displacement.” Volume B: Country Case Studies  | 7  TABLE A.1    Durable solutions analysis: criteria and themes Criteria Themes used to develop indicators 1.  Safety, security, and freedom of • Threats to safety movement • Security incidents • Freedom of movement (especially relevant for refugees) • Social cohesion 2.  Adequate standard of living • Poverty • Food security • Housing conditions (including crowding) and tenure security • Water and sanitation • Health care • Education (attendance/enrollment and literacy/attainment) 3.  Access to livelihoods and • Employment (before and after displacement) employment • Livelihoods (before and after displacement) • Access to land, livestock, and productive assets (before and after displacement) • Legal or administrative obstacles to employment (especially relevant for refugees) 4.  Mechanisms to restore housing, • Ownership or tenancy (before displacement) land, and property • Documents to prove ownership or tenancy (before displacement) • Access to restitution or compensation mechanisms 5.  Access to documentation • Access to documentation (before and after displacement) • Access to replacement mechanisms 6.  Family reunification • Separated household members • Access to reunification mechanisms 7.  Participation in public affairs • Community, social, or political organizations • Reconciliation and confidence-building initiatives 8.  Access to remedies and justice • Access to informal or formal justice • Reparations Source: Durable Solutions Analysis Guide. 17.  Basic demographic indicators are also critical for understanding the impacts of displacement on different groups. At a minimum, data need to be disaggregated by sex, age, and location. In many situations, other diversity criteria such as ethnicity, language, or area of origin are necessary to understand differences within the population, as well as challenges and opportunities for durable solutions. Data on displacement history are also useful. This includes (a) date of initial displacement, (b) causes of displacement, (c) number of times displaced, and (d) main reasons for choosing current location. 18.  While a durable solution is achieved when internal displacement ends, there is much developmental investment to be made in the space preceding a durable solution. Policy and programs around challenges of inclusion, integration, self-reliance, socioeconomic opportunities, and parity in living conditions and opportunities between hosts and displaced persons are relevant even before an end to displacement is achieved. Developmental responses are thus deployed to address these displacement-related vulnerabilities. The report presents insights for key interventions, programs, and policy, which can move displaced populations toward better developmental outcomes. Scope, Structure, and Methodology 19.  The quantitative study is based on household surveys and is divided into three volumes structured around country case studies and technical aspects of surveys. Volume A of the study provides the Executive Summary and Overview. Volume B contains the case studies with results and policy implications for displaced populations in the four 8  |  Informing Durable Solutions for Internal Displacement countries surveyed, as well as a case study with refugees from Somalia, South Sudan, and Sudan (among others). After a background chapter on conflict and forced displacement in Africa, the case studies explain results on the displaced and corresponding non-displaced comparison groups, and provide policy insights. The case studies are based on household-level surveys that were purposefully designed and implemented for the study. A concluding chapter summarizes the main findings from the case studies, and, through a durable solutions approach, proposes policy implications. Volume C of the study covers the technical aspects of micro-data collection in displacement contexts, problematizing some of the challenges and providing experimental insights into survey methodology. Volume C also details the analysis methodology employed in drawing ‘typologies’, or distinct groups, of displaced populations. 20.  The case studies are stand-alone displacement profiles that depict the socioeconomic conditions of IDPs and the non-displaced communities. The surveys’ questionnaires were relatively harmonized across the case studies to allow some cross comparisons. All case studies have the following components—though there is some flexibility on the sequence (Table A.2). First, a displacement and demographic profile is provided. This includes the causes of displacement as reported by displaced people themselves, and the reasons for choosing the current location as a temporary settlement. Further, demographics characteristics disaggregated by age and sex, underlining the patterns of family separation. Second, core poverty indicators and standard of living measures are provided, including poverty incidence, housing situations, food security, and service delivery. These indicators are analyzed for both displaced and host community members to draw a comparison between the two. Third, the analysis of livelihood sources and labor market participation provides a measure of the degree of self-reliance of the displaced people vs. aid dependency. The parallel assessment of a host communities’ livelihoods complements the analysis by providing the knowledge around potential economic opportunities and the structure of the local economy. Included in this part is also a comparison between the present situation and the pre-displacement situation of displaced people in terms of assets and livelihood. The fourth part assesses social cohesion through perceptions by displaced and host populations on displaced-host community ties. It also includes the degree of public participation and the perceptions on and experiences of safety and security by both host and displaced populations. The fifth part explains the results of typology analysis, which iden- tifies patterns in the data that point to distinct profiles or groups among the displaced. Understanding the differences in these profiles explains the displacement trajectories of the populations, allowing more targeted interventions. The last part concludes with the present and future intentions by displaced groups in terms of movement, resettlement, or return. The analysis highlights the potential push and pull factors for deciding to either stay in or move from the current location, or to return to the country/area of origin.   TABLE A.2    Location of displaced populations surveyed for this study Country Camp or non-camp Somalia Both camp and non-camp South Sudan Camp Nigeria Both camp and non-camp Sudan Camp Ethiopia Camp Source: Crisis Recovery Survey (CRS), High Frequency Survey Wave 4 (HFSSS W4), Skills Profile Survey (SPS), Nigeria IDPS, Somali High Frequency Survey Wave 2 (SHFS W2), and Sudan Inter Press Service (IPS). Volume B: Country Case Studies  | 9 Volume B 21.  ‘Overview of Conflict and Displacement in Africa’. This background chapter gives an overview of violent conflict and forced displacement trends in Africa. While noting the surge in conflict and displacement since 2010, the chapter explores the structural dynamics of violent conflict in the region. In particular, the relationship between conflict and displacement in the region is examined.   TABLE A.3    Number of forcibly displaced and host community households interviewed Ethiopia Nigeria Somalia South Sudan Sudan Displaced people23 3,600 1,500 1,000 2,400 2,000 Host community 1,700 1,500 500 950 1,000 Source: CRS, HFSSS W4, SPS, Nigeria IDPS, SHFS W2, and Sudan IPS. 22.  Case Study on IDPs in Nigeria. The Nigeria IDP Survey 2018 was conducted in six states of northeast Nigeria (Adamawa, Bauchi, Borno, Gombe, Taraba, and Yobe) where most of the IDPs are currently living. The survey collects micro-data on consumption, health, education, living conditions, and displacement-related questions. Within each state, an equal number of IDP and host community households were interviewed (Table A.3), where IDPs consisted of either those in camps or in host community settings. IDPs in camps make up 40 percent of the IDPs, while IDPs in host communities make up 60 percent. Nearly 2 million people are currently displaced in northeast Nigeria, most of them due to armed conflict. IDPs are largely women and children who live either in host communities or in camps. Almost all IDPs are poor, food insecure, and doing poorly on a range of basic living outcomes. Though slightly better off than IDPs, host communities face widespread poverty and severe living conditions. IDPs in camps generally have less favorable outcomes than IDPs living among the host community and are more keen to return home. A durable solution for Nige- ria’s IDPs will require security for all and a consideration of the needs of different groups. IDPs are provided water and sanitation facilities, are close to many basic amenities, and have good relations with their host communities. 23.  Case Study on IDPs in Somalia. Household-level data were collected through the Somali High Frequency Survey Wave 2 (SHFS W2) in 2017. The sample of IDPs and host communities is representative of the entire Somali population living in secure areas. Urban households represent 40 percent of the total population (10 percent from Mogadishu and 30 percent in other urban areas), followed by the nomads with 25 percent, rural households and IDPs in settle- ments with 20 percent and 15 percent, respectively (Table A.3). In Somalia, in addition to conflict and instability, recent drought conditions have exacerbated existing rates of acute and protracted displacement. IDPs, like the rest of the Somali population, are overwhelmingly young: over 50 percent of IDPs are under 15 years. IDPs face greater poverty and worse living conditions than residents. While almost 70 percent of Somali residents are poor, IDPs form an espe- cially marginalized group, with over 75 percent of IDPs living on less than US$1.90 per day, and more than 50 percent of IDP households facing hunger. Along with worse living conditions, IDPs also have lower human capital. In terms of livelihood, agricultural income has been squeezed out over the course of displacement, and many IDPs are employed in helping with businesses. The persistent and cyclical nature of the drivers of migration and conflict contribute to entrenched conditions, which require a developmental, resilience-based approach to help affected populations cope with these shocks and stresses, combined with continuing humanitarian assistance to shore up basic needs. 23. The survey and case study for Nigeria, Somalia, South Sudan, and Sudan concern only IDPs. In Ethiopia, they exclusively cover refugees. 10  |  Informing Durable Solutions for Internal Displacement 24.  Case Study on IDPs in South Sudan. Two household surveys were conducted in 2017 in urban settings: (a) the Crisis Recovery Survey (CRS) represents four of the largest urban-based IDP camps and (b) the High Frequency Survey South Sudan Wave 4 (HFSSS W4) is representative of the non-displaced urban population (Table A.3). The chapter pro- vides findings on several measures of welfare at the household level, comparing urban-based IDPs and non-displaced communities in urban settings. Both groups have been negatively affected by the civil war that broke out in late 2013. The conflict displaced nearly 4.5 million people, a third of the entire South Sudanese population. About 1.9 million of the displaced are IDPs, 85 percent of whom are outside camps. Predictably, the household-level data highlight that IDPs are significantly poorer and more likely to be unemployed than urban residents. They mainly rely on assistance and live in tents provided by humanitarian assistance. They suffer considerably more from overcrowding than urban residents, which contributes to psychological distress, tensions with other camp dwellers, and disease outbreaks. IDPs tend not to feel safe in the camps or enjoy freedom of movement. Finally, they have tense relations with the host community, which is a critical barrier to local integration. Future prospects for displaced people are completely uncertain, but secu- rity is the greatest priority. 25.  Case Study on IDPs in Sudan. In the Sudan IDP Profiling Survey 2018, IDP households from the two camps of Abu Shouk and El Salam, as well as households from the immediately neighboring city of Al Fashir, were interviewed (Table A.3). As both camps are in the immediate neighborhood of the city, this setup allows for good comparability between IDPs and the host population. At the same time, this setup limits the representativeness of the data to IDPs in camps around Al Fashir. Most IDPs in Sudan’s surveyed camps were displaced at the height of the Darfur conflict in 2003–04. IDPs and hosts are staggeringly poor, requiring urgent employment and livelihood opportunities. Most IDPs live in permanent structures provided by the humanitarian community, but do not depend on aid for food or income. While IDPs enjoy access to many services they value, food availability and access to electricity are often lacking. Both IDPs and hosts describe their relations as good, and they generally feel safe. Improving security and economic oppor- tunities in camps and return areas, as well as an upgrading of camp infrastructure, would be key elements in bringing about a durable solution. 26.  Comparative Refugee Profiles from Somalia, South Sudan, and Sudan in Ethiopia. To complement the IDP results, data on refugees from three of the IDP case studies are presented in a section on refugees in Ethiopia. Data on refugees and host communities in Ethiopia come from the Skilld Profile Survey (SPS) 2017, a household survey that was conducted with host community members and with refugees from South Sudan, Somalia, Eritrea, and Sudan living in camps. The four main regions that host refugees are: Tigray and Afar (hosting mostly Eritre- ans), Gambella (hosting South Sudanese), Benishangul Gumuz (hosting mostly Sudanese, but also South Sudanese), and Somalia (hosting Somalis). Given this variety, displacement situations in the country are remarkably diverse and result from a combination of protracted conflicts in neighboring countries (Somalia, Eritrea, and Sudan), more recent crises (South Sudan, Yemen), and endemic internal ethnic unrest in some peripheral regions (Oromia, Somali/ Ogaden, and Afar). As a result, Ethiopia has been one of the most important refugee-hosting countries for decades. Refugees are worse off in terms of standard of living compared to host communities, although they have compa- rable access to services. Among refugee groups, Eritreans are the ones that enjoy more rights compared to others and display a higher standard of living and lower poverty rates. On the other hand, South Sudanese are the poorest group on many indicators, including food security, housing, labor force participation, and ties to a host community. At the policy level, important changes are underway to ease limited socioeconomic rights of refugees and decrease camp reliance (that is, implementation of ‘nine pledges’). Volume B: Country Case Studies  | 11 27.  Cross-country analysis. This section pools and synchronizes the data across the five countries to jointly analyze key cross-cutting policy questions. It explores the livelihood shifts that IDPs and refugees experience as they move to a new labor market environment, investigates whether camp-based IDPs truly enjoy better living standards as is often perceived, analyzes how inequality in the host community can affect host households and their perceptions of IDPs, and maps how the geographic dispersion and longer displacement durations affect IDPs’ socioeconomic outcomes and return intentions. Results show that agricultural IDPs displaced into urban centers face a starkly different labor market environment and higher poverty. IDPs based in camps are poorer, face lower service access, and are more aid dependent than hosts and IDPs outside camps. Inequality in host communities is linked to worse perceptions of IDPs and signals the poor socioeconomic conditions that many members of host communities face. Finally, IDPs displaced farther from the original residence are more nonagricultural, have been displaced longer, and prefer to return. 28.  The timing of the surveys in the different countries can have a bearing on the results. The surveys for the five countries were conducted over the course of a few weeks to a few months. Further, they were conducted at different times of the year, in different countries. Thus, seasonal effects were likely captured in several indicators; for instance, employment rates might vary in other seasons. This is a caveat for the analysis. Volume C 29.  Volume C of the study provides experimental insights on survey methodology and details on typology analysis. Providing evidence for durable solutions targeting IDPs is often neglected despite their large numbers and the advantage of their status as citizens. While steps have been taken to make data collection more comprehensive and standardized, some methodological questions remain unanswered. Volume C addresses key survey methodology questions in the displacement context, and also proposes clustering approaches to derive typologies of IDPs. 30.  Evidence suggests that IDPs underreport consumption. An experiment in South Sudan investigates whether ‘honesty primes’ can increase honest reporting of consumption. Insights from behavioral science have often been used as a policy tool to increase honesty and discourage anti-social behavior. Building on this literature, a set of ‘honesty primes’ was experimentally administered to one-half of the survey respondents in IDP camps and in urban areas across South Sudan. The results suggest that the ‘honesty primes’ had a positive impact on the consumption reported by the poorest and more vulnerable respondents. 31.  The report also proposes clustering approaches to derive typologies of IDPs, to inform the required speci- ficity of programs to find durable solutions. Among the displaced, different groups can have different trajectories in displacement. Initial circumstances of displacement can translate into different needs and solutions depending on the displacement trajectory, which is pertinent for policy. A clustering analysis helps identify the different typologies of the displaced. The aim of the analysis is to exploit the socioeconomic micro-level data to identify different groups or profiles of displaced households across the countries considered in Sub-Saharan Africa. These typologies are drawn using data on the causes of displacement, the current needs of displaced people, and the potential solution to end displacement. 32. The typology analysis identifies different groups in the data and checks how these groups differ in policy-relevant indicators. A series of indicators ranging from conditions of displacement and pre-displacement out- comes (cause-based lens), current socioeconomic conditions (needs-based lens), and future intentions and support required to settle anew (solutions-based lens) are used as inputs in a multiple correspondence analysis (MCA). The 12  |  Informing Durable Solutions for Internal Displacement MCA aggregates the inputted indicators to identify which IDP and refugee households are similar to each other, and which ones have less resemblance to each other. This results in the formation of distinct groups, or typologies, of the displaced. Once these groups are derived, they are compared on a series of policy-relevant indicators (along the cause- needs-solutions–based approach) to highlight how they differ. Understanding the differences among the groups can allow better targeting of policy and programs. Clusters identified might not always align with ethnic groups, language groups, or locations. Thus, the scope of this analysis is to illustrate the relevance of the full displacement trajectory and explore how this can point to tailored solutions. However, a policy maker would likely need to rely on targeting mech- anisms that help identify individual households. Overview of Conflict and Displacement in Africa 33.  Globally, the number of IDPs, and forcibly displaced populations in general, has reached historic highs. In June 2019, there were 70 million forcibly displaced people, of which approximately 29 million are refugees and asylum seekers and 41 million are IDPs. The number of conflict-induced IDPs reached historic highs since 2015. In 2017 alone, there were 11.8 million new IDPs, which is almost double the number of new IDPs in 2016. About 88 percent of new IDPs come from just 10 countries, including Ethiopia, Somalia, and South Sudan. Of the 41 million IDPs, over 80 percent are hosted in just 12 countries (Table A.4). More strikingly, 68 percent of all refugees come from only five countries (including South Sudan and Somalia). Thus, forced displacement revolves around a limited number of highly destruc- tive violent conflicts in these five countries.24  TABLE A.4    Twelve largest displacement situations globally25 Number Number Number (end-2017), Origin country (end-2017), Host country (end-2017), Host country (IDPs) in millions (refugees) in millions (refugees) in millions 1 Syrian Arab Republic 6.8 Syrian Arab Republic 6.31 Turkey 3.48 2 Colombia 6.5 Afghanistan 2.62 Pakistan 1.39 3 Democratic Republic 4.5 South Sudan 2.44 Uganda 1.35 of Congo 4 Iraq 2.6 Myanmar 1.16 Lebanon 1.00 5 Sudan 2.1 Somalia 0.99 Islamic Republic of Iran 0.98 6 Republic of Yemen 2.0 Sudan 0.69 Germany 0.97 7 South Sudan 1.9 Democratic Republic of 0.62 Bangladesh 0.93 Congo 8 Nigeria 1.7 Central African Republic 0.55 Sudan 0.91 9 Afghanistan 1.3 Eritrea 0.49 Ethiopia 0.89 10 Turkey 1.1 Burundi 0.44 Jordan 0.69 11 Ethiopia 1.1 Iraq 0.36 Democratic Republic of 0.54 Congo 12 Somalia 0.8 Vietnam 0.33 Kenya 0.43 Source: IDMC and NRC 2018; UNHCR 2018b. 24. IDMC and NRC. 2018. “Global Report on Internal Displacement 2018”; UNHCR. 2018b. “Global Trends: Forced Displacement in 2017;” UNHCR. 2019. “Figures at a a Glance. Statistical Yearbooks.” 25. Refugee and IDP numbers can vary between sources, and estimates from international agencies and governments sometimes differ. The ranking of the 12 displacement situations presented here refers to the data sources mentioned at the bottom of the table. Volume B: Country Case Studies  | 13 34.  Globally, conflict has been on the rise. Since 2010, the number of violent conflicts and associated fatalities glob- ally has dramatically increased. In 2016, 49 armed conflicts caused over 100,000 direct deaths, which represent a slight decline since the 2014 peak, but overall are some of the highest levels since the end of the Cold War. During the same year, more countries were involved in some form of violent conflict than at any time in the previous 30 years, resulting in the highest number of internationalized violent conflicts since the Second World War. Today, conflicts spilling over state borders and creating instability in neighboring countries are the norm.26 35.  Sub-Saharan Africa remains the most conflict-affected region in the world, with one-third of the total number of conflicts taking place in the region. In 2015, Africa recorded the highest number of armed conflicts (21) since 1945. Compounding dynamics of power contestation, intercommunal conflicts, and the spread of vio- lent extremism currently characterize violent conflict in Africa. Large-scale civil wars are slightly declining and being replaced by more diffuse (and more numerous) forms of political violence (riots, social violence, bombings, and so on). Sub-Saharan Africa has also been experiencing a proliferation of conflicts between non-state armed groups, where the state is not involved. In 2016, 63 percent of all conflicts in the region were non-state based: one in four took place in Somalia. During the same year, four of the top ten deadliest conflicts globally took place in Sub-Saharan Africa— namely in the four countries that are part of this study: Nigeria, Somalia, South Sudan, and Sudan. Similarly, these four countries plus the Democratic Republic of Congo (DRC) experienced the highest rates of civilian fatalities between 2002 and 2016, with a peak in 2012–2016.27 36.  Sub-Saharan African conflict and internal displacement go hand in hand. The increase in the number of conflicts and related fatalities in Africa has been accompanied by an even higher targeting of civilians: 45 percent of all incidents of political violence in 2016 were directed against civilian populations. Thus, direct violence against civilians causes them to flee. Trends in the intensity of violent conflict and scale of forced displacement move in parallel, as the latter is a consequence of the former. Peaks in war deaths are often matched by corresponding peaks of waves of displaced people.28 As Sub-Saharan Africa is the most conflict-affected region globally, it is also highly prone to forced displacement. While the region accounts for 14 percent of the global population, its rate of new displacement in 2017 was 46 percent, as almost one in two newly displaced persons globally came from this region.29 Conflict and forced displacement in Africa are also interdependent as far as solutions are concerned: while resolving internal displacement can contribute to peace processes, the long-term fate of both refugees and IDPs often depends on the resolution of conflict. 37.  Internal displacement in Africa is characterized by overlapping drivers of conflict and environmental con- ditions. In terms of dynamics, there is an overlap between regional- and local-level crises that produce forced dis- placement, particularly in the Horn of Africa. The drivers of displacement in Africa are a complex juxtaposition of social, economic, political, and environmental factors, which are hard to disaggregate. While ‘refugees’ clearly define those who are subject to violence or the threat of it and who cross an international boundary, the classification of internal 26. United Nations and World Bank. 2018. “Pathways for Peace: Inclusive Approaches to Preventing Violent Conflict”; Dupuy et al. 2017. “Trends in Armed Conflict, 1946–2016.” 27. United Nations and World Bank. 2018. “Pathways for Peace: Inclusive Approaches to Preventing Violent Conflict”; Williams. 2017. “Continuity and Change in War and Conflict in Africa”; ACLED (Armed Conflict Location and Event Data). 2016; “ACLED-Africa, v.7”; IDMC. 2017. “Africa Report on Internal Displacement.” 28. World Bank. 2017. “Forcibly Displaced: Toward a Development Approach Supporting Refugees, the Internally Displaced, and Their Hosts.” 29. IDMC. 2017. “Africa Report on Internal Displacement.” 14  |  Informing Durable Solutions for Internal Displacement displacement suffers from more gray areas. The distinction in the literature between disaster- and conflict-induced IDPs is in many cases arbitrary, and in many instances the two factors combined produce internal displacement. The social and economic effects on resources, livelihoods, and governance that slow onset hazards (including droughts, land degradation, coastal erosion, and desertification) may create can hardly be singled out without factoring in those environmental causes. Ethiopia, Sudan, Somalia, and Nigeria are particularly prone to multidimensional displacement. In East Africa, land degradation and droughts are simultaneously causes and consequences of poverty and resource depletion, and are considered major threats to growth, food security, political stability, and social peace. In Somalia, for example, the 2011 famine resulted from the combination of environmental degradation (that is, drought) and political and economic factors that led to widespread displacement, among others.30 38.  Displacement trends in Sub-Saharan Africa have been dramatically upward in the last five years, and inter- nal displacement is the most pressing crisis. After the Cold War and up until 2013, the share of refugees hosted in Sub-Saharan Africa remained relatively stable, between 2 and 3.5 million (Figure A.1). This situation reverted in the last five years: by the end of 2017, Africa hosted nearly 6.3 million refugees, or about 31 percent of the total number of refu- gees in the world. On the other hand, internal displacement has been greater in scale than cross-border displacement every year since 2001. IDP trends have fluctuated more than refugee trends in the last decade, with a recent dramatic surge. Since 2011, IDPs in Sub-Saharan Africa almost doubled, from 7 million to 13.7 million. In 2016 alone, conflict and violence produced 2.8 million new IDPs.31   FIGURE A.1    IDPs and refugees hosted in Sub-Saharan Africa32 16 14 12 10 Millions 8 6 4 2 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Refugees IDPs Source: Authors’ calculation based on UNHCR Population Statistics and IDMC data, 2000–2017. 30. World Bank. 2017. “Forcibly Displaced: Toward a Development Approach Supporting Refugees, the Internally Displaced, and Their Hosts”; IDMC. 2017. “Africa Report on Internal Displacement”; Ferris. 2012. “Internal Displacement in Africa: An Overview of Trends and Opportunities.” 31. IDMC. 2017. “Africa Report on Internal Displacement”; UNHCR.2017. “Statistical Yearbook 2016.” 32. Except for 2017, ‘refugees’ include asylum-seekers. Volume B: Country Case Studies  | 15 39.  North and East Africa are the most displacement-affected subregions in Africa.33 North and East Africa, including the Horn of Africa are the subregions that have led the growth in the number of both IDPs and refugees. The share of IDPs that reside in the subregion was 54 percent in 2016, or 6.7 million IDPs. In 2017, there was a dramatic increase in the number of IDPs in Ethiopia to over 1 million, but also a more conservative estimate in the number of IDPs in Sudan, which lowered from 3.3 to 2.1 million. Thus, the total IDP stock in East Africa lowered to 6.1 million in 2017, or 44 percent of the total number of IDPs in Africa.34 Similarly, the share of refugees in East Africa is 22 percent of the total number of refugees globally. Approximately 4.3 million refugees were hosted in East Africa at the end of 2017 (they were 3.3 million at the end of 2016).35 East Africa also recorded the least progress in terms of conflict termination and/or resolution when compared to West and Southern Africa, which have witnessed important achievements in solv- ing conflict and decreasing the number of displaced people (for example in Angola, Mozambique, Liberia, and Sierra Leone). Compared to two decades ago, the Horn of Africa has progressively become a hot spot for originating refugees and IDPs. Refugees originating from Somalia, Sudan, South Sudan, and Eritrea have all substantially increased. By the same token, Kenya and Ethiopia have boomed as refugee-hosting countries—a trend which confirms that most of the forcibly displaced find refuge and settle in neighboring countries.36 40.  The countries included in this study are some of the most displacement affected in Sub-Saharan Africa and globally. As of end 2017, Sudan, South Sudan, Nigeria, Ethiopia, and Somalia rank fifth, seventh, eighth, eleventh, and twelfth, respectively, in the top 12 countries globally for the number of conflict-induced IDPs (Table A4). After the Dem- ocratic Republic of Congo (DRC), which tops the ranking at the regional level with over 4 million IDPs, these five coun- tries have the highest number of IDPs in Sub-Saharan Africa, ranging from 800,000 (Somalia) to over 2 million (Sudan) (Table A.5). In 2017 alone, 5.5 million new IDPs were added in Africa: nearly 2.3 million of them came from the five countries combined (857,000 in South Sudan, 725,000 in Ethiopia, 388,000 in Somalia, 279,000 in Nigeria, and 17,000 in Sudan). In terms of refugees, Sudan and Ethiopia are among the top 12 refugee-hosting countries globally, ranking second and third in Africa after Uganda. South Sudan, Somalia, and Sudan rank third, fifth, and sixth, respectively, in the top 12 refugee countries of origin.37 41.  Forcibly displaced people in Sub-Saharan Africa are mostly located in camps (40 percent) or in rural set- tings, while a minority (25 percent) reside in urban environments. East Africa is the most camp-reliant subregion in Africa: in 2013, 76 percent of refugees lived in camps (Table A.6). In fact, the surveys conducted for this study were exclusively administered in camps. This situation is in stark contrast with the rest of the world, where 94 percent of ref- ugees and IDPs do not reside in camps. Globally, 58 percent of refugees are urban based, a percentage that rises to an estimated 80–90 percent in the Middle East and North Africa region.38 33. UNHCR and IDMC have slightly different lists of countries that define ‘East Africa’. Both lists include the four countries of this study (Ethiopia, Somalia, South Sudan, and Sudan), not including Nigeria, obviously. When talking about refugees in East Africa, the following paragraph includes the UNHCR list with the following countries: Chad, Djibouti, Eritrea, Ethiopia, Kenya, Somalia, South Sudan, Sudan, and Uganda. When talking about IDPs in East Africa, the following paragraph includes the IDMC list with the following countries: Comoros, Djibouti, Eritrea, Ethiopia, Kenya, Madagascar, Mauritius, Rwanda, Seychelles, Somalia, South Sudan, Sudan, Tanzania, and Uganda. 34. Verwimp and Maystadt. 2015. “Forced Displacement and Refugees in Sub-Saharan Africa: An Economic Inquiry”; UNHCR. 2018b. “Global Trends: Forced Displacement in 2017”; IDMC. 2017. “Africa Report on Internal Displacement.” 35. Verwimp and Maystadt. 2015. “Forced Displacement and Refugees in Sub-Saharan Africa: An Economic Inquiry”; UNHCR. 2018b. “Global Trends: Forced Displacement in 2017.” 36. Verwimp and Maystadt. 2015. “Forced Displacement and Refugees in Sub-Saharan Africa: An Economic Inquiry.” 37. IDMC and NRC, 2018. “Global Report on Internal Displacement 2018”; UNHCR. 2018b. “Global Trends: Forced Displacement in 2017.” 38. Verwimp and Maystadt. 2015. “Forced Displacement and Refugees in Sub-Saharan Africa: An Economic Inquiry”; World Bank. 2017. “Cities of Refuge in the Middle East. Bringing an Urban Lens to the Forced Displacement Challenge”; UNHCR. 2018b. “Global Trends: Forced Displacement in 2017.” 16  |  Informing Durable Solutions for Internal Displacement   TABLE A.5    Extent of forced displacement in concerned countries IDPs (conflict-induced) Number of refugees hosted by Number of refugees displaced Country (end-2017) (end-2017) from (end-2017) Ethiopia 1,078,000 889,000    87,000 Nigeria 1,707,000   1,900   239,000 Somalia   825,000  14,500   986,000 South Sudan 1,899,000 283,000 2,439,000 Sudan 2,072,000 906,500   694,000 Source: UNHCR. 2018b. Global Trends: Forced Displacement in 2017; IDMC data.   TABLE A.6    Overview of the study’s forced displacement situations Displacement Month, year situation Population Overview of data Source IDPs in Nigeria 1,707,000 Ongoing insurgency by Boko Haram has created December, 2017 IDMC data 2.2 million refugees and IDPs since 2014. Nearly 300,000 new IDPs in 2017. IDPs in Somalia   825,000 Protracted displacement patterns result from December, 2017 IDMC data long-standing mix of insecurity, weak government control, and environmental factors. Nearly 400,000 new IDPs in 2017. IDPs in South Sudan 1,899,000 Civil war broke out in late 2013, and is currently December, 2017 IDMC data experiencing fractionalization of violence. Displacement is ongoing despite ceasefire declared in late 2017. IDPs in Sudan 2,072,000 Internal displacement results from prolonged both December, 2017 IDMC data state and non-state based fighting in Darfur, South Kordofan, and Blue Nile states. Violence decreased in 2017. Somali refugees in   256,000 Located in camps (Somali/Ogaden region), April, 2018 UNHCR 2018a Ethiopia refugees fled due to environmental issues and insecurity. Protracted situation. Eritrean refugees in   168,000 Most integrated refugee group; they fled April, 2018 UNHCR 2018a Ethiopia persecution and economic hardship. Located in urban areas and camps in Tigray and Afar, and Addis Ababa. Sudanese refugees    44,000 Some of the poorest refugees, fled continuous April, 2018 UNHCR 2018a in Ethiopia waves of localized violence. Settled in Benishangul-Gumuz. South Sudanese   440,000 Highest poverty incidence, refugees fled civil war April, 2018 UNHCR 2018a refugees in Ethiopia during last 4 years. Located in Gambella. Ongoing humanitarian crisis. Source: UNHCR 2018a, IDMC data. Case Studies IDPs in Nigeria Introduction and Country Context 42.  Nigeria is Africa’s most populous country and a key player in the West African subregion. Strategically located on the Gulf of Guinea, Nigeria shares 4,047 km of borders with Benin to the west, Niger and Chad to the north, and Cameroon to the east. It has a coastline to the south on the Atlantic Ocean of approximately 853 k. With a pop- ulation of approximately 184 million, Nigeria is the most populous country in Africa, and the seventh most populous country in the world. Nigeria will be the third most populous country in the world by 2050, overtaking Pakistan, Brazil, Indonesia, and the United States.39 Nigeria is Africa’s largest economy and biggest oil exporter. It has the largest natural gas reserves in Africa and the ninth most proved natural gas reserves in the world. Apart from petroleum and gas, Nige- ria also has many other natural resources, including bauxite, coal, gold, iron ore, lead, limestone, niobium, tin, and zinc. 43.  Since its independence in 1960 from British colonial rule, Nigeria has had a difficult political history, includ- ing many violent conflicts. Soon after independence the country was plunged into a fratricidal war that claimed more than 1 million lives in the southeastern region of the country. The Nigeria-Biafra civil war, which ended in 1970 with the defeat of the secessionist eastern region by federal troops, forced the displacement of over 1 million people and created tense relations between the mainly Muslim north and the Christian southeast. The strains of the war are still evident in Nigeria’s challenge to maintain a federal system of government while providing relative autonomy to the 36 states of the federation. Recent calls by political elites for restructuring and more local control of natural resources, as well as the resurgence of secessionist groups, are indicative of Nigeria’s continuous efforts to evolve a political structure that fulfills the yearnings of all its people. 44.  Employment and other opportunities for personal development for Nigeria’s teeming youth remain very limited. Corruption, poor infrastructure, and poor access to health care and education continue to create stressed living conditions for the majority of Nigeria’s young population. Furthermore, increasing inequality has deepened the gap between the rich and the poor. Tense relationships between Muslims and Christians have led to more radicalized religious identities, while also pushing religious communities deeper into their silos. 45.  Climate change has added a layer of complexity to an already vulnerable and stressed political, social, and economic ecosystem in northeastern Nigeria. The shrinking of Lake Chad affects hundreds of thousands of fishermen, farmers, and pastoralists that live around the lake and the surrounding islands and depend on its water for food and livelihoods. Climate change has combined with a range of complex ethnic and political factors to push Fulani herdsmen deeper into the south in search for pasture. This has led to a series of violent clashes and reprisal attacks between the pastoralists and local farmers, particularly in Adamawa, Benue, and Plateau states, and further displacements. The military has been deployed in the most affected areas. Additionally, increasing banditry and armed 39. United Nations. 2015. “World Population Prospectus: The 2015 Revision.” 17 18  |  Informing Durable Solutions for Internal Displacement violence in the broader semiarid Sahel region has resulted in military deployments in Zamfara state and other parts of northwestern Nigeria, thus further stretching the country’s military. 46.  Boko Haram started as a relatively minor security threat in Maiduguri, Borno state, and has evolved into a complex terrorist network that has morphed with global jihadist revivalist movements to threaten not only Nigeria but the entire Lake Chad region. After embarking on one of the most violent military campaigns by a terrorist group in modern history, Jama’atu Ahlis Sunna Lidda’awati Wal-Jihad (also known as Boko Haram) pledged allegiance in March 2015 to the Islamic State and officially became the West African Province of ISIS (ISWAP). However, in August 2016 the Islamic State supported a split from the Abubakar Shekau–led Jama’atu Ahlis Sunna Lidda’awati Wal-Jihad. Shekau’s rival Abu Musab al-Barnawi was appointed the new governor of ISWAP. The split within the group has not reduced the violence; rather, Boko Haram violence and its areas of operational influence have expanded more recently. 47.  The Boko Haram insurgency has contributed significantly to deaths from political, ethnic, and religious violence in Nigeria. More than 53,000 deaths from political violence have been reported in Nigeria between 2011 and the second quarter of 2018.40 More than 60 percent of the deaths have been in the northeast states of Borno, Adamawa, and Yobe, with Borno state alone accounting for more than 25,000 deaths during the period.41 48.  More than half of the displaced population in Nigeria’s northeast region fled their homes between 2013 and 2015 when the insurgents seized control of vast territories. At the height of the crisis—between mid-2013 and 2015—Boko Haram captured and occupied over 30,000 square kilometers of towns and villages mainly in Borno, Adamawa, and Yobe states, committing grave atrocities against the local populations. This led to hundreds of thou- sands of people fleeing their homes, particularly in the three states, with many fleeing to the safer Gombe state (Fig- ure B.1). Counterinsurgency operations by Nigerian security forces in the surrounding towns and villages also forced several other hundreds of thousands to flee and seek refuge in safer communities and locations where they could receive humanitarian assistance.   FIGURE B.1    Displacement dates of IDPs by region and type 100 80 % of IDPs 60 40 20 0 1998 2002 2003 2006 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Adamawa Bauchi Borno Gombe Taraba Yobe Hosted Camp Source: Authors’ calculations using Nigeria IDP Survey, 2018. 40. Africa Program at the Council on Foreign Relations. n.d. “Nigeria Security Tracker.” 41. Ibid. Volume B: Country Case Studies  | 19 49.  The number of IDPs in the northeast has recorded a steady upward trend in recent months. As of June 2018, there were 1,918,508 displaced individuals and 364,323 displaced households.42 This represents a 2 percent rise from April 2018 and a 5 percent rise from June 2017, which was 1,825,321. Though marginal, current trends indicate a steady rise in the number of displaced persons since the December 2017 dip of 1,702,680.43 The rise is attributable to the resurging attacks by Boko Haram and ISWAP as well as pastoralist-farmer clashes in Taraba and Gombe states and the consequent military operations in the region. 50.  Many of the displaced persons are located in Borno state, the epicenter of the Boko Haram insurgency. People internally displaced by the Boko Haram insurgency in the northeast are spread across at least 13 states. Of these, the vast majority of IDPs are currently located in the three states of Borno, Adama, and Yobe, where active conflict is still ongoing at the time of writing. There were 1,439,953 IDPs in Borno as of June 2018, while Adamawa and Yobe states host 178,977 and 136,662 IDPs, respectively.44 Borno state is the birthplace of Boko Haram. Mohammed Yusuf, the slain founder of the group, started his movement in Maiduguri, the Borno state capital, when he founded the Ibn Taymiyyah mosque near the Maiduguri railway station.45 Of the displaced population, about 1.1 million are in 1,933 host community locations, with the majority living with host families, while more than 660,000 live in the nearly 240 highly congested camps and camp-like settlements such as government land or public buildings and schools.46 Because of the congested camps and poor and unhygienic living conditions, IDPs, particularly children, are highly vulnerable to outbreaks of diseases and infections. Secondary displacement is quite common. More than 70 percent of IDPs have reported that they moved twice or more since they first left home.47 51.  A vast majority of IDPs outside of Borno state live in host communities. About 61 percent of all IDPs across the six northeast states were living in host communities as of June 2018, while 39 percent were in camps.48 Borno state, which has 75 percent of all IDPs, is the only state where the percentage of IDPs residing in host communities are the same as those residing in displacement sites (Figure B.2).49 There are 282 displacement sites in the northeast, of which 59 percent are collective settlements, 40 percent are camps, and 1 percent are transitional centers. About 93 percent of IDPs in Adamawa live in host communities. Yobe state has 91 percent of IDPs in host communities and 9 percent in displacement sites.50 Taraba and Bauchi both have significantly higher numbers of IDPs in host communities, while all IDPs in Gombe live in host communities. A majority of displaced persons stay in host communities mainly because of family, religious, and cultural relationships with members of host communities. Although this puts huge pressures on families and communities, it reduces the pressure on displacement camps. This prevents IDPs from the complicated process of documentation at displacement camps. 42. IOM (International Organization for Migration). 2018. “Nigeria Displacement Tracking Matrix, (DTM) Round 23, June 2018.” 43. IOM. 2017. “Nigeria Displacement Tracking Matrix, (DTM) Round XX, December 2017.” 44. IOM. 2018. “Nigeria Displacement Tracking Matrix, (DTM) Round 23, June 2018.” 45. Matfess. 2017. “Women and the War on Boko Haram: Wives, Weapons, Witnesses.” 46. IOM. 2018. “Nigeria Displacement Tracking Matrix, (DTM) Round 23, June 2018.” 47. Ibid. 48. Ibid. 49. Ibid. 50. Ibid. 20  |  Informing Durable Solutions for Internal Displacement   FIGURE B.2    Number and locations of IDPs by state Source: IOM 2018. 52.  A large number of people displaced within Nigeria have returned to their states of origin. As of April 2018, there were more than 1.33 million reported IDP returns in Nigeria as resources in their respective areas of refuge had dwindled. The durability of these returns is questionable, as the majority of people face considerable humanitarian needs upon return, effectively remaining in a situation of internal displacement. 53.  Nigeria has fairly robust institutional provisions as well as legal and policy frameworks to guide the pro- tection and assistance of IDPs. The National Commission for Refugees, Migrants, and Internally Displaced Persons (NCFRMI) coordinates Nigeria’s national action and response on IDPs, refugees, returnees, and other persons of concern to the commission. The NCFRMI was established by the National Commission for Refugees Act (1989), now Cap. N21 Laws of the Federation of Nigeria 2004 to safeguard the interest and treatment of refugees. The NCFRMI Act incor- porated the 1951 United Nations (UN) Convention relating to the status of refugees, its 1967 Protocol, and the 1969 Organization of African Unity Convention governing specific aspects of refugee problems in Africa. The functions of the commission as contained in the enabling act include laying general guidelines and policy on issues relating to refugees and persons seeking asylum in Nigeria and advising the government on relevant policies. The mandate of the commis- sion was expanded in 2002 and 2009 to include resettlement of IDPs and the coordination of migration, respectively. The commission has taken the responsibility of providing protection and assistance to IDPs and refugees. One of the core responsibilities of the commission is to protect IDPs against discrimination, acts of rape, assault, and other forms of violence. Volume B: Country Case Studies  | 21   BOX B.1    Nigeria IDP survey The Nigeria IDP survey represents IDPs and host community households across northeast Nigeria. The Nigeria IDP survey was conducted in six states of northeast Nigeria (Adamawa, Bauchi, Borno, Gombe, Taraba, and Yobe) where most of the IDPs are currently living. The survey collects micro-data on consumption, health, education, living conditions, and displacement-related questions. About 1,400 IDP and 1,400 host community households were interviewed; households were interviewed proportional to IDP population numbers in each state (Table B.1). Within each state, an equal number of IDP and host community households were interviewed, where IDPs consisted of either those in camps or in host community settings. IDPs in camps make up 40 percent of the IDPs, while IDPs in host communities make up 60 percent. Borno rep- resents 80 percent of the total sample, as most of the IDPs are currently living there.51   TABLE B.1    Sample size and comparison groups from the survey IDP and host community sample Comparison group N % of sample IDP 1,437 48.8 Host community 1,510 51.2 Total 2,947 100 IDP sample Comparison group N % of IDP sample Hosted 874 61 In-camp 563 39 Male head 934 65 Female head 503 35 Poorest quintile 503 35 Q2 321 22 Q3 196 14 Q4 192 13 Richest quintile 196 14 Total 1,437 100 Source: Authors’ calculations using Nigeria IDP Survey, 2018. 54.  Nigeria’s National Emergency Management Agency (NEMA) has been responsible for setting up and man- aging IDP camps in northeast Nigeria. Established via Act 12 as amended by Act 50 of 1999, NEMA is mandated to manage responses to disasters, formulate policy on disaster management, and coordinate relief programs. NEMA’s operations are guided by the National Contingency Plan, Search and Rescue and Epidemic Evacuation Plan, National Disaster Management Framework (NDMF), and Emergency Response Standard Operating Procedures. One of the core thematic areas of the NDMF is Institutional Capacity for Disaster Management, where it has the responsibility to pro- vide relief materials to IDPs and assist in their rehabilitation. This thematic area mandates the establishment of disaster management structures at the federal, state, and local government levels. The work of NEMA overlaps in many aspects with that of NCFRMI. 51. As the majority of the survey was conducted in Borno (2,200 out of 2,800 interviews), the analysis does not compare outcomes across different states due to the low number of observations from the other states. 22  |  Informing Durable Solutions for Internal Displacement 55.  Nigeria’s National Policy on Internal Displacement, which proposes best practices for the management of IDPs, is still in draft form and yet to be adopted by the legislature. Drafted by a Presidential Committee in 2006, the policy was revised in 2009 and 2012 to respond to the provisions of the African Union (Kampala) Convention for the Protection and Assistance of IDPs in Africa, of which Nigeria is a signatory. However, the national policy has not been given legislative framework. The process for the adoption of the national policy began in 2012, but it is yet to be adopted by the legislature as a statute. In the absence of a clear policy framework on internal displacement in Nige- ria, despite the high number of IDPs, the national response to IDPs has remained largely fragmented and ad hoc. The consequence is that the national response to the root causes of internal displacement, which the policy targets, has remained generally feeble. 56.  Nigeria is a signatory to the African Union (Kampala) Convention for the Protection and Assistance of IDPs in Africa but it is yet to be domesticated. The 2009 Kampala Convention was adopted based on the UN Guiding Principles on Displacement. It entered into force in 2012 and thus became the first legally binding regional instrument to guarantee and protect the rights of IDPs. Although Nigeria signed and ratified the Convention since 2012, there is yet to be a full domestication of the convention. Section 12 (1) of the Nigerian 1999 Constitution requires that treaties between the federation and any other country “shall have the force of law except to the extent to which any such treaty has been enacted into law by the National Assembly.”52 Nigerian legislators and humanitarian activists have already accomplished remarkable efforts toward achieving national assembly approval and presidential assent. 57.  By signing the Kampala Convention, the Government of Nigeria (GoN) demonstrated that it has the pri- mary duty and responsibility to assist and protect IDPs in its territory, and it has taken important steps at federal, state, and local government levels to respond to the needs of IDPs. Although the Kampala Convention is yet to be incorporated into domestic laws in Nigeria, its ratification does impose binding obligations on the GoN. In response to these obligations, the GoN has worked collaboratively with the humanitarian community to provide lifesaving assistance that has helped stabilize living conditions of millions of displaced persons. Nigeria’s humanitarian response plan targets 6.1 million people in 2018 out of an estimated 7.7 million people in need of lifesaving assistance in the most directly affected states of Borno, Adamawa, and Yobe.53 The GoN has also worked to avert the risk of famine and has contained a major cholera outbreak in the region. To help bring essential services closer to host communities, the GoN established five humanitarian hubs in Maidugrui, Gwoza, Bama, Ngala, and Dikwa. The Office of Nigeria’s Vice President in 2018 oversaw the establishment of a humanitarian technology hub in Yola, the Adamawa state capital, in collaboration with the American University of Nigeria, to help boost economic recovery in the region while supporting humanitarian recovery efforts. Demographic Profile 58.  A significant proportion of IDPs in northeast Nigeria are children. About 57 percent of the IDPs are children under 15 years and less than 2 percent are elderly (Figure B.3). There are slightly fewer children among the host com- munity, where about 52 percent are under the age of 15 years. Within IDPs, those living in host communities have more children than those in camps (58 percent vs. 55 percent, p < 0.01) but IDPs in camps have more women (51 percent vs. 52. “Constitution of the Federal Republic of Nigeria.” 1999. 53. OCHA. 2018. “Northeast Nigeria Humanitarian Situation Update, July 2018 Edition.” Volume B: Country Case Studies  | 23 45 percent, p < 0.01). Children tend to be more vulnerable to abuses from insurgents and even from security operatives, while women tend to be more at risk of gender-based violence and trafficking.   FIGURE B.3    Population structure for IDPs and host communities, by sex and age 60 0.9 0.8 1.0 0.6 0.6 50 0.7 1.0 0.7 12.8 12.4 15.2 14.4 14.1 % of population 40 14.4 15.7 7.8 7.7 14.4 30 8.9 7.9 7.6 6.7 6.5 6.5 20 32.2 33.9 24.7 24.0 27.8 26.7 28.1 10 23.6 0 Men Women Men Women Men Women Men Women IDP Host community Hosted Camp Children (0–14) Youth (15–24) Adults (25–64) Elderly (>65) Source: Authors’ calculations using Nigeria IDP Survey, 2018. 59.  Households headed by women have smaller household sizes but higher dependency ratios. Woman-headed households account for nearly 40 percent of both IDP and host community households (Figure B.4). The widespread violence in the northeast contributes to death and family separation, possibly for both IDPs and hosts. Within IDPs, there are slightly fewer woman-headed households in camps than in host communities (p < 0.01). Household sizes and dependency ratios are similar for IDPs in host communities and in camps (Figure B.5). However, woman-headed households on average have one less member than man-headed households (5.6 compared to 6.4, p < 0.01), but have a significantly higher dependency ratio than man-headed households (2.3 compared to 1.4, p < 0.01). This is likely due to woman-headed IDP households losing adult males in the household from the conflict.   FIGURE B.4    Women-headed household for   FIGURE B.5    Household size and dependency ratio for IDPs IDPs and host communities and host communities 60 8 Dependency Ratio/ Household size % of households 6 40 4 2 20 0 IDP Host community Man headed Woman headed Hosted Camp 0 IDP Host Hosted Camp community Overall Source: Authors’ calculations using Nigeria IDP Survey, 2018. Overall IDP Dependency ratio Household size Source: Authors’ calculations using Nigeria IDP Survey, 2018. 60.  Several woman-headed households lost their husbands and children during displacement, who in some cases died or went missing because of the conflict. About 12 percent of both woman- and man-headed households lost a member of their household during displacement (Figure B.6). About 20 percent of the separated members for woman-headed households were husbands, and 40 percent were children. Similarly, about 60 percent of the separated 24  |  Informing Durable Solutions for Internal Displacement members for man-headed households were spouses and children. Reasons for separation are, however, grimmer for woman-headed households. About 60 percent of the separated members in woman-headed households died or are missing, compared to about 25 percent of the separated members in man-headed households (Figure B.7). It is possible that women who lost their husbands recently and took on the responsibility of the households are also suffering from the trauma of dead and missing household members.  FIGURE B.6    Relation of separated member for IDPs   FIGURE B.7    Reason for separation for IDPs 100 100 90 90 80 80 household members household members % of IDPs who lost % of IDPs who lost 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 Male head Female head Male head Female head Spouse Child Deceased Missing Niece or nephew Sibling Displaced to another location Stayed behind Other relative Other Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018. 61.  Most IDPs are Muslims from the Kanuri tribe, which is also a sect that Boko Haram represents. Over 95 per- cent of both IDPs and host community households are Muslim, while 5 percent are Christian (Figure B.18). While the country has about an equal composition of both Muslims and Christians, the northeast region is mostly inhabited by Muslims.54 While the religious composition of IDPs is largely uniform, they represent various ethnic tribes. Most of them identify with the Kanuri tribe, and many are from Hausa and other smaller tribes (Figure B.9). Although the leader of Boko Haram is from the Kanuri tribe, and whose multimedia messaging in the Kanuri language has focused on the perceived marginalization of the Muslim and Kanuri people in the northeast, they also make up the majority of Boko Haram’s victims.55 54. United Nations. 2015. “World Population Prospectus: The 2015 Revision.” 55. Agbiboa. 2013. “Why Boko Haram Exists: The Relative Deprivation Perspective.” Volume B: Country Case Studies  | 25   FIGURE B.8    Religion composition for IDPs and host   FIGURE B.9    Tribal composition for IDPs communities 100 80 % of population 60 40 20 0 IDP Host Hosted Camp community Christianity Islam Hausa Kanuri Fulani Marghi Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018. Displacement Profile 62.  Armed conflict is the main driving force of displacement in northeast Nigeria. Over 60 percent of IDPs in northeast Nigeria have been displaced due to armed conflict in their villages or surrounding areas (Figure B.10). Just over 15 percent of IDPs were also displaced by a lack of access to basic services. These factors appear to be equally relevant for IDPs in host communities and IDPs in camps. The reasons for choosing the current location also reflect that security is an overriding concern. Almost 50 percent of the IDP households reported better security as being the major reason they chose their current location, while for 20 percent, it was better access to services (Figure B.11). More than 15 percent chose the current location due to family reasons, possibly having existing family ties in the area.   FIGURE B.10    Reasons for leaving original place of   FIGURE B.11    Reasons for arriving at current location residence for IDPs for IDPs 100 100 80 80 % of IDPs % of IDPs 60 60 40 40 20 20 0 0 Hosted Camp Hosted Camp Armed conflict Better security Communal clashes Better access to assets/services Increased crime violence Better access to livelihood opportunities Lack of access to assets/services Family reasons Other reasons Access to humanitarian aid Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018. 26  |  Informing Durable Solutions for Internal Displacement 63.  Most IDPs have relocated within their state of origin and have not been repeatedly displaced. Over 95 per- cent of IDPs have relocated within their state of origin, but outside of their local government area (LGA), an adminis- trative subdivision of the state. This proportion is slightly higher for IDPs living in camps (p < 0.05) (Figure B.12). About 15 percent of the IDPs in camps come from the same LGA, as opposed to 11 percent of IDPs in host communities. A combination of security factors, family ties, and a desire to remain within a culturally familiar terrain might explain why many households do not relocate to other states. A majority of IDP households have also not been repeatedly displaced, with only about one in five reported changing their residence more than once after the initial displacement (Figure B.13).   FIGURE B.12    Place of origin for IDPs   FIGURE B.13    Number of times residence changed for IDPs 100 100 80 80 % of IDPs 60 % of IDPs 60 40 40 20 20 0 Host Camp 0 Hosted Camp Same ward Same local government area Same state Different state Once Twice Thrice or more Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018. 64.  Most IDPs in camps want to return to their original residence, while IDPs in host communities want to stay where they are. Nearly 60 percent of IDPs prefer to stay at their current location while 40 percent prefer to return to their homes, and almost none express a desire to move to a new area (Figure B.14). These preferences are similar for men- and women-headed households and poor and non-poor households, but differ drastically for IDPs in camps and those in host communities. About 70 percent of IDPs in host communities want to stay at their current location, whereas only about 20 percent of IDPs in camps intend to do so. This difference also holds after controlling for negative camp conditions and other personal and communal push and pull factors that might affect the decision (Table B.2). 65.  IDPs who want to return to their original residence are motivated by a lack of basic services, while those who want to stay are motivated by security concerns. Lack of access to services and opportunities is the main reason for 45 percent of IDPs in camps who want to return to their original residence (Figure B.15). On the other hand, better security is the main reason for about 50 percent of the IDPs in host communities who want to stay at their cur- rent location (Figure B.16). IDPs hosted in camps and displacement sites are often living in congested shelters or iso- lated in insecure or inhospitable areas, making them vulnerable to exploitation and abuse.56 Additionally, humanitarian 56. OCHA. 2015. “Nigeria: Northeast Crisis Situation.” Volume B: Country Case Studies  | 27 assistance has fallen short and not everyone has had access to basic services in camps.57 Indeed, having a good rela- tionship with the host community is an important determinant of the wish to stay, even after controlling for the effect of other living conditions that influence the decision (Table B.2). Other key factors for wanting to return are having a missing family member and having had agricultural land before displacement.  FIGURE B.14    Return intentions for IDPs 100 90 80 70 60 % of IDPs 50 40 30 20 10 0 p 40 ll d ad 60 am ra ad te ve he p os he m C To O H tto an an Bo M om W Stay here Return Move to new area Source: Authors’ calculations using Nigeria IDP Survey, 2018. 66.  Addressing security concerns is imperative to allow IDPs to settle in their preferred locations. Over 40 per- cent of IDPs report security as the biggest factor to be able to settle in their new or current location (Figure B.17). In addition, IDPs also cite access to land and livelihood development to be able to settle. IDPs in camps are keener to get land and shelter than those in host communities (38 vs. 27 percent, respectively), indicating that housing is perceived as much more temporary by the IDPs in camps. Even after controlling for living in a camp and other circumstances, owning the dwelling is an important determinant for wanting to stay in the current location (Table B.2). IDPs in host communities are more likely than those in camps to cite livelihood development (8 percent for camp IDPs and 16 per- cent for those in host communities). Security updates constitute the most pressing information need for IDPs, with over 60 percent stating that it is the main information they require to make an informed decision about where to settle (Figure B.18). Hence, while access to shelter and livelihood development are major points of interventions, safety and security in IDP’s current and original places of residence are critical in helping IDPs decide where to settle. 57. IDMC. 2018. “Global Report on Internal Displacement.” 28  |  Informing Durable Solutions for Internal Displacement   TABLE B.2    Determinants of return intentions Dependent variable: intention to return vs. stay (reference) Specification 1 (only demographic Specification 2 Specification 3 Specification 4 Specification 5 controls and (with camp feature (with other pull (with social (with origin FE Independent variables region FE*) controls) factors) relations controls) and ethnicity FE) Camp IDP 1.372*** 0.993*** 0.998*** 0.999*** 0.915*** Year of displacement −0.072 −0.083 −0.047 −0.048 −0.019 Sex of household head −0.042 0.020 −0.034 −0.040 −0.041 Age of household head −0.026 −0.024 −0.015 −0.015 −0.009 Household size 0.031 0.016 0.016 0.017 0.028 Dependency ratio 0.005 −0.025 −0.052 −0.059 −0.087 Years of education of 0.009 0.019* 0.022** 0.021* 0.022* households Consumption per person −0.000 0.000 −0.000 −0.000 −0.000 Concrete/brick house −0.098 −0.180 −0.208 −0.210 Own the dwelling −0.506** −0.516** −0.515** −0.584** Live in squatting 0.704*** 0.615*** 0.613*** 0.717*** Number of people/room 0.036 0.014 0.011 0.007 Feel safe in current home −0.809*** −0.835*** −0.530 −0.531 Access to agricultural 0.220 0.044 0.040 0.035 land Separated household 0.926*** 0.929*** 1.016*** member In contact with people at 0.344** 0.319** 0.326** origin Had access to agricultural 0.610*** 0.630*** 0.627*** land before displacement Good relations with −0.215 −0.279 neighborhood Good relations with hosts −0.667*** −0.724*** Region FE Yes Yes Yes Yes Yes Origin FE No No No No Yes Ethnicity FE No No No No Yes Pseudo R-squared 0.0843 0.107 0.131 0.139 0.155 Observations 1,217 1,210 1,195 1,193 1,180 Source: Authors’ calculations using Nigeria IDP Survey, 2018. Note: *FE = fixed effects. Volume B: Country Case Studies  | 29   FIGURE B.15    Reasons for wanting to return to original   FIGURE B.16    Reasons for wanting to stay at cur- place of residence for IDPs in camps rent location for IDPs in host communities Family reasons Family close by Lack of aid Better access to home/land Lack of access to services and opportunities Better access to livelihood Increased crime Better access to services Communal clashes Better security Armed conflict 0 20 40 60 0 10 20 30 40 50 60 % of IDPs in camps that want to return % of IDP in host communities that want to stay Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018.   FIGURE B.17    Help needed to settle in preferred   FIGURE B.18    Information needed to settle in preferred location for IDPs location for IDPs Access to basic services Help with returning Land/property/housing Livelihood development Livelihood opportunities Agricultural tools Basic services Access to land/shelter Security Security Political situation 0 10 20 30 40 50 0 10 20 30 40 50 60 70 % of IDPs % of IDPs Camp Hosted Camp Hosted Source: Author’s calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018. Standard of Living 67.  Almost all IDPs and host community members are living in poverty. A staggering 87 percent of IDPs are living below the international poverty line of US$1.90 2011 Purchasing Power Parity (PPP) per day per capita (Figure B.19). Host communities have a similar rate of poverty. IDPs in camps and host communities have similar poverty rates, and so do woman-headed and man-headed households. These poverty rates are steeply higher than the national poverty rate of 46 percent. The conflict-affected northeast region of the country has poverty rates ranging from 60 percent to 90 percent.58 Additionally, as of June 2018, Nigeria overtook India as the country with the highest number of people living in extreme poverty, with about 50 percent of the population living under US$1.25 per day.59 The host communi- ties that receive IDPs are themselves in severely resource-constrained conditions, urging a developmental investment in the area for the benefit of not only IDPs but also the host communities. 58. Khan and Cheri. 2016. “An Examination of Poverty as the Foundation of Crisis in Northern Nigeria.” 59. World Data Lab. 2018. “World Poverty Clock.” 30  |  Informing Durable Solutions for Internal Displacement   FIGURE B.19    Poverty headcount ratio for IDPs and host communities 100 80 %of population 60 40 20 0 IDP Host community Man headed Woman headed Man headed Woman headed Man headed Woman headed Overall Man headed Woman headed Overall Overall Host IDP Hosted Camp community Poverty incidence National average, 2010 North East average, 2010 Source: Authors’ calculations using Nigeria IDP Survey, 2018. 68.  Poverty is deeper for IDPs than for host communities. IDPs who are poor consume less than 30 percent of the US$1.90 PPP (2011) per day per capita threshold, while host communities consume about 37 percent of this threshold (p < 0.01) (Figure B.20). IDPs in camps and in host communities have similar poverty gaps, which is the consumption shortfall of the poor relative to the poverty line (P line). In the IDP camps, woman-headed households have deeper poverty gaps than man-headed households, indicating an added vulnerability in the camp setting.   FIGURE B.20    Poverty gap relative to US$1.90 PPP (2011) poverty line 350 Poverty gap in Naira (mean consumption 300 shortfall relative to PPP P line) 250 200 63% 71% 70% 70% 70% 70% 73% 70% 71% 79% 150 100 50 0 IDP Host community Man headed Woman headed Overall Man headed Woman headed Overall Man headed Woman headed Overall Hosted Camp IDP Mean income Poverty line Source: Authors’ calculations using Nigeria IDP Survey, 2018. Volume B: Country Case Studies  | 31 69.  While IDPs and host communities have similar poverty rates, IDPs are much more food insecure. About 61 percent of IDPs are highly food insecure compared to 48 percent of host community households (p < 0.05) (Fig- ure B.21). IDPs living in host communities also face these high rates of food insecurity, though they live among the host community households. Food insecurity is more prevalent among woman-headed households, with 70 percent being highly food insecure compared to 58 percent of man-headed IDP households. Along with the trend in poverty gaps, food insecurity also affects the woman-headed households more. Food aid is not a significant part of IDP food con- sumption; however, IDPs in camps are still getting ample food and nonfood assistance. On average, food aid only makes up about 1 percent of food consumption for IDPs and almost 0 for host communities (Figure B.22). The aid shares are driven by camps, where about 4 percent of food consumption comes from food aid, compared to less than 1 percent for IDPs in host communities.   FIGURE B.21    Food insecurity categories for IDPs and   FIGURE B.22    Share of aid in food consumption for host communities IDPs and host communities 100 15 80 12 % of population food consumption % of food aid in 60 9 40 6 20 3 0 0 IDP Host community Hosted Camp Man head Woman head Poorest quintile Q2 Q3 Q4 Richest quintile IDP Host community Hosted Camp Man head Woman head Poorest quintile Q2 Q3 Q4 Richest quintile Overall IDP Overall IDP Source: Authors’ calculations using Nigeria IDP Survey, 2018. High food insecurity Medium food insecurity Low food insecurity Source: Authors’ calculations using Nigeria IDP Survey, 2018. 70.  Housing conditions of IDPs have deteriorated, with the poorest IDPs and those in camps having the worst conditions. About one in four IDP households are living without a house, or in tents or tukuls (huts made with sticks) (Figure B.23). Before displacement, no IDPs were without a house or lived in tents. Conditions are much worse for IDPs in camps, half of whom live in poor housing arrangements. The poorest two quintiles of IDPs also have significantly worse housing conditions than the richer quintiles. Overcrowding also increased sharply for IDPs, with almost half of them living in households with more than four people per room, compared to one in five households before displacement (Figure B.24). In terms of both physical structure and overcrowding, IDPs had similar housing conditions before dis- placement compared to host community households’ current housing conditions. While IDPs in camps are only slightly more overcrowded than those in host communities, their inferior physical housing structures can make overcrowding a more complex challenge. 32  |  Informing Durable Solutions for Internal Displacement   FIGURE B.23    Type of dwelling for IDPs and host   FIGURE B.24    Household members per room for IDPs communities and host communities 100 100 % of population 80 80 % of population 60 60 40 40 20 20 0 0 IDP—origin IDP—current Host community Hosted Camp Poorest quintile Q2 Q3 Q4 Richest quintile IDP—origin IDP—current Host community Hosted Camp Poorest quintile Q2 Q3 Q4 Richest quintile Overall IDP Overall IDP No house Tent More than 4 2–4 1–2 Less than 1 Tukul/gottiya Mud/wood hut Source: Authors’ calculations using Nigeria IDP Survey, 2018. Concrete/brick house Source: Authors’ calculations using Nigeria IDP Survey, 2018. 71.  Most IDPs lost homes they had owned for many years before being displaced. About 70 percent of IDPs owned their homes, while others rented their homes before displacement (Figure B.25). IDPs were better off than host community households are currently, as about 50 percent of the host community households own their homes. Cur- rently, only 10 percent of the IDPs own their homes (Figure B.26). However, most IDPs had owned their homes for many years, with 70 percent of them owning them for between 4 to 20 years before displacement.   FIGURE B.25    Tenure of dwelling for IDPs and host   FIGURE B.26    Number of years dwelling owned for communities IDPs and host communities 100 100 80 80 % of population % of population 60 60 40 40 20 20 0 0 IDP—origin IDP—current Host community IDP—origin IDP—current Host community Owned Rented Less than 1 year 1 to 3 years Relatives/friends Squatting 4 to 10 years 10 to 20 years Other Many generations Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018. Volume B: Country Case Studies  | 33 72.  IDPs use better sanitation facilities than before; however, IDPs in camps suffer from overcrowding in toi- lets. Nearly 80 percent of IDPs use improved sanitation facilities compared to 70 percent before displacement, with no difference between IDPs in camps and those in host communities (Figure B.27). Almost 80 percent of host community households also use improved sanitation facilities. IDPs make up 57 percent of those households that share toilets with others (compared to 43 percent for host communities, p < 0.01) (Figure B.28). Among the IDP households that share toilets, almost 40 percent do so with more than four households. The situation is much worse in camps, where nearly 70 percent share a toilet with more than four households, and over 40 percent share with more than ten households. In contrast, only 30 percent of IDPs in host community settings share toilets with more than four households. IDPs in camps are therefore suffering from overcrowding in rooms and toilets.   FIGURE B.27    Access to improved sanitation facilities   FIGURE B.28    Shared household toilets for IDPs and for IDPs and host communities host communities 100 100 % of population sharing toilet 80 80 IDP % of population 60 60 40 40 20 20 HC 0 0 Share toilet IDP Host community Hosted Camp in nt ity d p ad ad te am ig rre un he he os or C cu m P— H an an om P— M om ID tc ID W os H Source: Authors’ calculations using Nigeria IDP Survey, 2018. Overall IDP 3 or less households 4 to 9 households 10 or more households Source: Authors’ calculations using Nigeria IDP Survey, 2018. 73.  IDPs are using improved sources of water and have similar access to services as host community house- holds. About 90 percent of IDPs are using improved sources of water, compared to 80 percent before displacement (Figure B.29). IDPs’ current water sources are also slightly better than host community households, as 90 percent use improved water sources (compared to 83 percent for host community households, p < 0.01). Distance to basic ame- nities, such as the nearest water source, health center, school, market, and police, are similar for both IDPs and host community households (Figure B.30). IDPs are therefore comparable with host communities in terms of access to basic services. 34  |  Informing Durable Solutions for Internal Displacement   FIGURE B.29    Access to improved water   FIGURE B.30    Distance to basic services for IDPs and host sources for IDPs and host communities communities 100 60 Mean time (minutes) taken 80 % of population 40 60 40 20 20 0 0 Hosted Hosted Hosted Hosted Hosted Camp Camp Camp Camp Camp IDP Host community IDP Host community IDP Host community IDP Host community IDP Host community in ity p ad am ig un he or C m P— an om om ID tc W os H Source: Authors’ calculations using Nigeria IDP Survey, 2018. Water Health Education Market Police Source: Authors’ calculations using Nigeria IDP Survey, 2018. 74.  IDP women are often burdened with water collection duties, and half of IDP households face obstacles in collecting water. Adult and young women make up about 60 percent of IDPs who are tasked to collect water (compared to 43 percent for host communities, p < 0.01) (Figure B.31). Within IDPs, this number is similar across IDPs in host communities and camps and man- and woman-headed households. Additionally, while about 48 percent of IDPs report no obstacles for collecting water (compared to 63 percent of host community households), 30 percent of IDPs also report long queuing times (Figure B.32). Hence, women in IDP households likely spend a substantial amount of time collecting water, which could have educational, employment, or other opportunity costs. Additionally, given the insecure environment IDPs live in, it can also increase the risk of being victims of sexual assault and other forms of gender-based violence (GBV). 75.  IDP households, particularly those in camps or headed by men, use less favorable child delivery services. About 46 percent of IDP women who gave birth in the last two years delivered their child at home, as opposed to a clinic or hospital. This is twice as many as host community households (Figure B.33). However, about 60 percent of IDPs in camps delivered their child at home compared to about 42 percent of IDPs in host communities. Interestingly, this proportion is much lower for households headed by women, where about 38 percent gave birth at home, compared to 50 percent of the households headed by men. Similar trends are reflected in birth delivery assistance. IDPs were twice as likely as host communities to use traditional birth attendants or relatives, as opposed to doctors (p < 0.01). Using traditional attendants and relatives is also more common for IDPs in camps and those headed by men (60 percent and 50 percent, respectively) (Figure B.34). IDPs may be adopting these practices due to a lack of options, with health cen- ters being less accessible as a result of the displacement crisis. Volume B: Country Case Studies  | 35   FIGURE B.31    Household member who collects water   FIGURE B.32    Water collection obstacles for IDPs for IDPs and host communities and host communities 100 100% 80 80% that collects water % of population % of population 60 60% 40 40% 20 20% 0 0% IDP Host community IDP Host community Hosted Camp Man headed Woman headed No obstacle Insufficient water Source doesn't work Affordability Source: Authors’ calculations using Nigeria IDP Survey, 2018. Overall IDP Female child Adult woman Male child Adult man Source: Authors’ calculations using Nigeria IDP Survey, 2018.   FIGURE B.33    Place where last child was delivered for   FIGURE B.34    Person who helped deliver last child for IDPs and host communities IDPs and host communities 100 100 % of populaton that delivered % of populaton that delivered 80 80 a child in last two years a child in last two years 60 60 40 40 20 20 0 0 P ity d p d d P ity ed p om ead ad te am a a ID am ID un he he un t he os os h C m C m H an an H an an m om M om co M tc W t W os os H H Hospital Maternity clinic At home Doctor/clinical officer Nurse/midwife Source: Authors’ calculations using Nigeria IDP Survey, 2018. Patient attendant Traditional birth attendant Relative/friend No one Source: Authors’ calculations using Nigeria IDP Survey, 2018. 36  |  Informing Durable Solutions for Internal Displacement 76.  IDPs have lower primary school enrollment than the host community, and girls have worse educational outcomes in both groups. About 33 percent of both primary and secondary school age IDP children are currently enrolled in school (Figure B.35). For host communities, around 50 percent of primary school age but only 27 percent of secondary school age children are enrolled in school. Girls have lower enrollment rates in both groups. In IDP house- holds, similar proportion of girls and boys are enrolled in primary school, whereas twice as many boys are enrolled in secondary school (p < 0.01). For host community households, boys have higher enrollment rates in both in primary and secondary school. This pattern is also reflected among the working-age population in these households. Girls are more likely to either not be educated or only have religious training, in both IDP and host community households (Figure B.36). Hence, while host community households may be sending more of their children to school, both types of households prioritize education for boys slightly over girls.   FIGURE B.35    Primary and secondary enrollment   FIGURE B.36    Highest education level for work- rates for IDPs and host communities ing-age populations for IDPs and host communities 80 100 % of primary and secondary 70 80 60 % of population age children 50 60 40 30 40 20 20 10 0 0 IDP Host community Hosted Camp Boys Girls Boys Girls IDP Host Community Hosted Camp Boys Girls Boys Girls Overall IDP Host Overall IDP Host community community Net primary enrollment No education Religious Net secondary enrollment Others Primary & intermediate Secondary University Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018. 77.  A majority of IDP children have been out of school for more than three years, reportedly due to affordabil- ity. More than half of primary and secondary age IDP children have been out of school for over three years (Figure B.37). As many IDPs were initially displaced about three to four years ago (in 2014–2015), it is likely that most children were unable to resume their education after displacement (Figure B.1). About 50 percent of the children in camp settings have been out of school for over three years compared to only 16 percent of children in host community settings. Interestingly, over 60 percent of IDPs state lack of financial resources as the main reason they are unable to send their children to school (Figure B.38). However, primary school education is free in northeast Nigeria, so lack of information is also a likely barrier in this case.60 60. Centre for Public Impact. 2017. “Universal Basic Education in Nigeria.” Volume B: Country Case Studies  | 37  FIGURE B.37    Gap in primary and secondary   FIGURE B.38    Reasons for not sending primary and sec- education for IDPs ondary age children to school for IDPs 80 % of primary and secondary age IDP children not attending secondary age IDP 70 % of primary and children not attending school 60 100 50 school 80 40 30 60 20 40 10 0 20 ng ol st k k es or or ho re rc u w w yo te sc ou se 0 r in fo o es d ou to o Overall Boys Girls Hosted Camp he d lr N H de ill is a St ci ee fin an d/ N 6 months or less More than half a year fin ol o of 1 year 2 years To ck La 3 years More than 3 years Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018. Livelihoods and Labor 78.  About 65 percent of IDPs are employed, though discrepancies exist within IDPs. About 46 percent of working-age IDPs are employed, and an additional 19 percent are employed and enrolled in education. Nearly 20 per- cent of the working-age IDPs are unemployed or inactive in the labor force, which is similar to the inactivity rates in host communities (Figure B.39). Inactivity in the labor market is more common among women and in camps. About 30 percent of the camp labor force is inactive or unemployed, compared to about 16 percent of IDPs in host commu- nities. Further, women are about three times as likely as men to be inactive in the labor force or unemployed. Because maintaining sustainable livelihoods is key to achieving a durable solution, these differences in employment status indicate a need to address the labor market barriers that specific groups face. 79.  Most IDP households get their livelihood from agriculture. Almost 70 percent of IDP households today get their livelihood primarily from agriculture, compared to 50 percent before displacement (Figure B.40). IDP livelihoods have shifted into agriculture and away from salaries—only about 16 percent of households rely on salaries today as opposed to 33 per- cent before displacement. The switch away from salaries and into agriculture has led the IDP livelihood structure to become similar to that of host communities, which could indicate an adaptation into the new labor market. IDPs living among host communities are much more likely to be in agriculture—75 percent as opposed to 54 percent of IDPs in camps. 80.  IDPs have lost their agricultural land, but renting land from host communities could be allowing them to maintain agricultural livelihoods. Over 45 percent of IDP households had access to agricultural land before dis- placement, while only 26 percent do now (Figure B.41). Further, most IDPs who had access to land before displacement were owners (64 percent), while only 25 percent of IDPs own their agricultural land today (Figure B.42). IDPs in camps had greater rates of access to land before displacement, compared to those living in host communities (60 percent and 41 percent, respectively; p < 0.01). However, today both groups have similar rates of access. IDPs have lost a lot of agricultural land over the course of displacement, along with housing. Most IDPs (56 percent) who have access to land currently are renting it. Host communities have been assisting IDPs by renting them plots of agricultural land, and that could explain why many of them have switched to agricultural jobs.61 Before displacement, IDPs may have also been renting out the land they owned, since their livelihoods were not as agriculture dependent. 61. World Bank Group. 2016. “North-East Nigeria. Recovery and Peace Building Assessment.” 38  |  Informing Durable Solutions for Internal Displacement   FIGURE B.39    Labor force participation among the   FIGURE B.40    Main source of livelihood for IDPs and working age for IDPs and host communities host communities 100 100 % of working age population 80 80 % of households 60 60 40 40 20 20 0 0 P ity d p en en P— igin om nt ity d p ed ed te am ID te am rre un M om un ad ad os os or C C m cu m he he H W P— H om an an ID tc tc M om ID os os W H H Unemployed Agriculture Inactive and not enrolled in school/college Manufacturing Inactive and enrolled in school/college Retail Employed and enrolled in school/college Wages and salaries Employed and not enrolled in school/college Other Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018.  FIGURE B.41    Access to agricultural land for IDPs and   FIGURE B.42    Tenure of agricultural land for IDPs and host communities host communities 80 100 70 90 % of population with % of population with 80 60 access to land access to land 70 50 60 40 50 30 40 20 30 10 20 0 10 IDP—origin IDP—current Host community Hosted—before Hosted—current Camp—before Camp—current 0 IDP—origin IDP—current Host community Owned Rented Provided for free by relatives Overall IDP Community land Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018. Volume B: Country Case Studies  | 39 81.  IDPs who are inactive cite a lack of opportunities, skills, and capital, while host communities primarily cite a lack of opportunities.62 About 35 percent of IDPs are unable to find a job due to lack of opportunities, 27 percent due to lack of skills, and 20 percent due to lack of capital for business (Figure B.43). In contrast, 80 percent of host com- munity members cite lack of opportunities as the main reason they are unable to secure a job or participate in the work- force. IDPs are nearly three times more likely than host communities to cite a lack of skills (27 percent and 10 percent, respectively), indicating a need to adjust to the new labor market. Both IDPs and host communities primarily ask for monetary investment or micro-loans to support them in securing employment (Figure B.44). About 30 percent of IDPs also demand vocational training while about 20 percent of host community households would like to get connected to potential employers. Host community households are more likely to cite lack of work opportunities, while IDPs are more likely to cite both lack of work opportunities and lack of skills—even if IDPs and host community individuals have similar levels of education (Figure B.36). This could be because IDPs engaged in different livelihood activities before displacement, than the activities and opportunities in their current locations.   FIGURE B.43    Reasons for being unemployed or   FIGURE B.44    Support required for securing employ- inactive for IDPs and host communities ment for IDPs and host communities 100 100 80 80 inactive population inactive population % of unemployed/ % of unemployed/ 60 60 40 40 20 20 0 0 IDP Host community IDP Host community Lack of skills Monetary investment Lack of informatiom Micro loan Lack of work opportunities Business start-up training Lack of capital to start business Vocational training Other Securing contacts Source: Authors’ calculations using Nigeria IDP Survey, 2018. Social and Public Capital 82.  Both IDPs and host communities agree that they enjoy good relations, with the latter feeling that IDPs do not get enough aid. Almost 90 percent of IDPs and 80 percent of host community households at least slightly agree that IDPs have good relations with locals (Figure B.45). Perceptions of good relations with the host community are common to both IDPs in camps and those living among hosts. IDPs seem to be well integrated, and poor neighbor- hood relations are unlikely to be the reason why IDPs in camps want to return (Figure B.14). About 60 percent of IDPs agree that they get enough support or aid, while only 50 percent of host community households agree (Figure B.46). Among IDPs, 10 percent more IDPs in camps agree that they get enough support than IDPs in host communities (p < 0.05), indicating a stronger assistance and outreach in camps. Host communities appear sympathetic, believing more strongly than the IDPs themselves that they do not get sufficient aid, and possibly renting them land to perform agricultural work (Figure B.42). 62. The population here consists of working-age IDPs and host community members who are either unemployed or inactive in the labor force. 40  |  Informing Durable Solutions for Internal Displacement  FIGURE B.45    IDPs have good relations with   FIGURE B.46    IDPs receive sufficient aid neighbors 100 100 80 % of population 80 % of population 60 60 40 40 20 20 0 IDP Host community Hosted Camp 0 IDP Host community Hosted Camp Overall IDP Overall IDP Strongly agree Slightly agree Strongly agree Slightly agree Neutral Slightly disagree Neutral Slightly disagree Strongly disagree Strongly disagree Source: Authors’ calculations using Nigeria IDP survey 2018. 83.  Both IDPs and host community households rely on their social networks for credit, which they perceive as difficult to tap. When asked whom they would approach if they needed 50,000 Naira in a short time, over 60 percent of both IDPs and host community households report that they would borrow the money from their relatives in case of an emergency, while about 20 percent would borrow from friends (Figure B.47).63 Additionally, almost 70 percent of both groups also believe that borrowing this money from their friends or relatives would be difficult or very difficult (Figure B.48). The reliance on social networks for emergency money can indicate a lack of formal safety nets. It is likely that IDPs and host communities do not have much access or information about banks or other institutions, prompting them to turn to relatives and friends even though they believe it would be very difficult to borrow the money.   FIGURE B.47    Sources of credit at short notice for  FIGURE B.48    Ease of borrowing at short notice for IDPs and host communities IDPs and host communities 100 100 80 80 % of population % of population 60 60 40 40 20 20 0 0 IDP Host community IDP Host community Relatives Friends Very easy Easy Community Landlord Neither easy nor difficult Difficult Microfinance/banks Nobody Very difficult Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018. 63. 50,000 Nigeria Naira is equivalent to roughly US$140, as of September 2018. Volume B: Country Case Studies  | 41 84.  IDPs are less likely than host communities to participate in public meetings or meet community leaders. Almost 70 percent of IDPs have not attended any public meetings in the past year, compared to 57 percent of host community households (Figure B.49). In addition, 60 percent of IDPs have not met a community leader in the past year either, whereas 47 percent of host community households have not done so (Figure B.50). IDPs in camps and woman-headed households are less likely to have met a community leader in the past compared to IDPs in host com- munities and man-headed IDP households.   FIGURE B.49    Participation in public meetings for   FIGURE B.50    Interaction with community leader for IDPs and host communities IDPs and host communities 100 100 80 80 % of population % of population 60 60 40 40 20 20 0 0 IDP Host community Hosted Camp Man headed Woman headed IDP Host community Hosted Camp Man head Woman head Overall IDP Overall IDP Never 1 to 4 times 5 or more times Never 1 to 4 times 5 or more times Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018. Targeting Analysis 85.  From the targeting perspective, IDPs living among host communities and in Borno face the most seri- ous gaps to income generation. Most IDP households in Nigeria are productive but poor (71 percent), followed by support-dependent (19 percent) and self-reliant (10 percent). Though host communities have a slightly larger share of self-reliant households than IDPs, most households across both groups are productive but poor. IDPs living in host communities are more likely to be support-dependent than the host community and the camp IDPs. Further, among the IDPs, support-dependent households are extremely concentrated in Borno state (Figure B.51; Figure B.52). 86.  Substantial investment will be necessary to improve living conditions among host communities and sustain their ability to accommodate disadvantaged and vulnerable IDP groups. From the targeting per- spective, hosts are in extremely dire living conditions themselves, and any durable solution will require build- ing the living standards and opportunities for the hosts so that they can accommodate IDPs. From a sustainable income-generation perspective, IDPs in Borno state are at a larger gap to a durable solution than other areas. A policy response would need to be three-pronged: developing the entire region, rehabilitating camp IDPs to more permanent locations, and supporting services and income generation for IDPs in more permanent living arrange- ments (among hosts or otherwise). 42  |  Informing Durable Solutions for Internal Displacement   FIGURE B.51    Vulnerable population by status of the   FIGURE B.52    Vulnerable IDP population by state household 100 100 80 % of households 80 % of households 60 60 40 40 20 20 0 Adamawa Bauchi Borno 0 Host IDPs in IDPs in host community settlements community Self-reliant Self-reliant Productive but poor Productive but poor Support-dependent Support-dependent Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018. Typology of IDPs 87.  Two distinct groups of IDPs are identified in the analysis for Nigeria. Group 1 represents 74 percent of the IDP population and Group 2 around 26 percent (Figure B.53) The Boko Haram conflict has led to a large number of deaths and displacements. The place of origin is similar for both groups of IDPs. Even though most households were displaced by armed conflict, Group 1 is slightly more likely to cite this reason compared to Group 2. Before displacement both groups had similar living conditions, yet Group 2 was more inclined to an agricultural livelihood and Group 1 was more likely to rely on wages, salaries, and their own business. Currently, households in Group 1 have less members, have a higher dependency ratio, and are more likely to be headed by an unemployed woman. IDPs in Group 2 are more likely to have an agricultural livelihood and to receive assistance, although both groups are equally poor and food insecure. The differences in housing conditions and access to services between groups are determined by their current location, as Group 1 is more likely to live in host communities, Group 2 in settlements or camps. IDPs in Group 2 were more satisfied before displacement, are more dissatisfied today, are less likely to feel safe, and are more pessimistic about the future. Households in Group 1 prefer to stay in their current location motivated by security reasons, while IDPs in Group 2 intend to return to their place of origin guided by access to land, services, and employment. Cause Profile 88.  Most IDPs in Group 2 live in camps, are slightly less likely to be displaced by armed conflict, and more likely to have members separated from the household, compared to IDPs in Group 1. The Boko Haram insurgency con- tributed significantly to deaths from political, ethnic, and religious violence in Nigeria. Most IDPs have relocated within their state of origin, indicating the need for finding a camp close to their place of origin and the risks associated with prolonged travel. The place of origin is similar for both groups of IDPs; around 1 out of 10 households come from the same LGA, 8 out of 10 households from the same state, and only 1 out of 10 from another state or country. However, only 15 percent of households from Group 1 live in IDP settlements, compared to 50 percent of households in Group 2 Volume B: Country Case Studies  | 43 (Figure B.54). Even though most households were displaced by armed conflict, IDPs in Group 1 are slightly more likely to cite this reason (92 percent) compared to the population in Group 2 (77 percent). Nevertheless, 20 percent of house- holds from Group 2 had a member living in the household who was separated during displacement, against 2 percent of IDPs from Group 1.  FIGURE B.53    Visualization of groups from the clustering analysis Source: Authors’ calculations using Nigeria IDP Survey, 2018. Note: Group 1 is represented by the blue circles and Group 2 by the black triangles. 89.  Before displacement both groups had similar living conditions, yet Group 2 was more inclined to an agri- cultural livelihood and Group 1 was more likely to rely on wages, salaries, and their own business. Housing conditions were similar for both groups of IDPs before displacement. Most households owned or rented their dwelling and were close to main services. Also, the majority had improved water (8 out of 10 households) and sanitation (7 out of 10 households). Access to electricity was different since nearly half of Group 2 had electricity, compared to 33 percent of IDPs in Group 1. The results are robust after controlling for region effect and other household characteristics. Moreover, 77 percent of households from Group 2 had access to agricultural land before displacement, compared to 34 percent of IDPs in Group 1.64 As a result, more than 5 out 10 households in Group 2 had an agricultural livelihood before displace- ment, compared to 3 out of 10 from Group 1 (Figure B.55). Contrary to this, IDPs in Group 1 were 1.6 times more likely to obtain income from wages, a salary, and their own business. Around 1 out of 10 households in every group relied on aid, remittances, or other livelihood before displacement. The difference in main source of income between both groups is robust after controlling for region effect and other household characteristics. 64. Having access to land does not necessarily imply ownership. 44  |  Informing Durable Solutions for Internal Displacement   FIGURE B.54    Displacement profile  FIGURE B.55    Source of livelihood pre-displacement 100 80 90 70 80 60 70 % of households % of households 60 50 50 40 40 30 30 20 20 10 10 0 0 In IDP Displaced by With members Agriculture Wage, salary, Aid, settlement armed conflict separated from and own remittances, the household business and other Group 1 Group 2 Group 1 Group 2 Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018. Needs Profile 90.  Households in Group 1 have less members, a higher dependency ratio, and are more likely to be headed by an unemployed woman compared to IDPs in Group 2. Households in Group 1 have 5.9 household members and an age-dependency ratio of 1.9 against 6.7 members and a dependency ratio of 1.5 among IDPs in Group 2 (Table B.3).65 A larger dependency ratio comes from differences in the number of working-age members between both groups, since they have a similar share of children and elderly in the household. IDPs in Group 2 are mostly headed by men (76 percent), whereas 45 percent of households in Group 1 are headed by women. Sex differences in the head of the household are robust after controlling for region effects and other characteristics. Furthermore, 8 out 10 household heads in Group 2 are employed, compared to 7 out of 10 from Group 1. The current household composition is affected by the displacement circumstances. Many IDPs lost men and children during displacement, and as a result women took the responsibility of the household. IDPs in Group 1 were more likely to be displaced by armed conflict, which possibly explains having less members and more women as heads of the household compared to those in Group 2.   TABLE B.3    Current household characteristics and poverty status   Group 1 Group 2 Household size  5.9  6.7 Share of children in the household 17.1 16.5 Age dependency ratio  1.9  1.5 Share of households headed by women (%) 45.0 23.6 Share of employed household heads (%) 68.0 80.0 Poverty incidence (% of population) 88.0 84.3 Poverty gap (% of the poverty line) 73.4 64.2 Source: Authors’ calculations using Nigeria IDP Survey, 2018. 65. The age dependency ratio is defined as the proportion of children and old age dependents of working-age population (15–64). Volume B: Country Case Studies  | 45   FIGURE B.56    Current source of livelihood   FIGURE B.57    Current housing conditions 80 70 Shared toilet 60 % of households 50 Improved sanitation 40 30 Improved water sources 20 10 Rent/own the dwelling 0 0 10 20 30 40 50 60 70 80 90 100 Agriculture Wage, salary, and Received own business assistance % of households Group 1 Group 2 Group 1 Group 2 Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018. 91.  IDPs in Group 2 are more likely to have an agricultural livelihood and to receive assistance compared to Group 1, although both groups are equally poor and food insecure. The majority of IDPs are poor and food insecure. Around 60 percent of households face high food insecurity in each of the two groups identified. In addition, poverty incidence is similar for households in both groups (Table B.3)66 However, poor households in Group 1 are fur- ther away from the poverty line (73 percent) compared to poor IDPs in Group 2 (64 percent). In their current location, 15 percent of IDPs in Group 1 have access to agricultural land, and 13 percent have an agricultural livelihood, compared to 47 percent and 25 percent, respectively, of IDPs from Group 2 (Figure B.56). IDPs in Group 2 lost their agricultural land, but renting land from host communities could allow them to maintain an agricultural livelihood. Moreover, IDPs in Group 2 are almost three times more likely to receive assistance from development partners, NGOs, or the government. This group seems to have greater access to safety nets and networks, possibly because they are more likely to live in IDP settlements where assistance is focalized. 92.  Households from Group 2 are less likely to own or rent their dwelling, to have improved sanitation, and to share the toilet, but more likely to have improved water sources, compared to Group 1. The differences in housing conditions and access to services between groups of IDPs are determined by their current location, as Group 1 is more likely to live in host communities, whereas Group 2 in camps. About 84 percent of households in Group 1 own or rent their dwelling against 47 percent of IDPs in Group 2 (Figure B.57). Despite relatively high levels of access to improved water sources and sanitation, Group 1 is more likely to have access to improved sanitation (83 percent vs. 72 percent) while Group 2 to improved water sources (95 percent vs. 89 percent). Also, sharing a toilet is more common among households in Group 1 despite having less members in the household. The differences in housing conditions between both groups are significant after controlling for region effects and other household characteristics (Appendix F). 66. The poverty line corresponds to a daily value of US$1.90 PPP per day. 46  |  Informing Durable Solutions for Internal Displacement   FIGURE B.58   Perceptions 100 90 80 70 % of households 60 50 40 30 20 10 0 Satisfied with living Satisfied with Feel safe in Expect better conditions before current living current conditions in the displacement conditions location future Group 1 Group 2 Source: Authors’ calculations using Nigeria IDP Survey, 2018. 93.  Group 2 households were more satisfied before displacement, are more dissatisfied today, are less likely to feel safe, and are more pessimistic about the future. A similar share of households from Group 1 was satisfied with their living conditions before displacement (73 percent) and is satisfied with their current conditions (66 percent). Contrary to this, 90 percent of households from Group 2 were satisfied before displacement, yet currently only 50 percent are satisfied (Figure B.58). Besides, Group 1 is more likely to feel safe from crime and violence (95 percent vs. 87 percent), and they are also more optimistic about the future (85 percent vs. 77 percent). Safety concerns could be driven by exposure to armed conflict, since one out of five households in Group 2 were separated from a household member. A worse perception about the future in Group 2 relative to Group 1 is probably associated with living in IDP settlements and having to rely on aid, as well as a worse perception of safety. Solutions Profile 94.  Households in Group 1 prefer to stay in their current location motivated by security reasons, while IDPs in Group 2 intend to return to their place of origin guided by access to land, services, and employment. The profile of both groups of IDPs is different in terms of return intention and durable solution. About 78 percent of households in Group 1 want to stay in their current location, compared to only 3 percent of those in Group 2 (Figure B.59). Among those who want to relocate, the majority of households in both groups (96 percent) would like to return to their place of origin, and only some (4 percent) would like to move somewhere else (Figure B.60). Security is the main reason behind the decision to stay for IDPs in Group 1 (87 percent), whereas IDPs in Group 2 looking to return to their place of origin are motivated by access to land, services, and employment (49 percent), as well as other reasons (32 percent). The timeline for moving is similar for both groups, only 20 percent of households in each group plan to move in the next 12 months, and most (around 80 percent) do not have a clear timeline. Insecurity was less likely to be the cause of displacement for households in Group 2, and thus they are less likely now to consider insecurity as a key factor. Furthermore, a desire to stay seems to be more common among households that live outside camps—most households in Group 1—who are more satisfied with their current conditions and more optimistic about the future. Volume B: Country Case Studies  | 47   FIGURE B.59    Return intention   FIGURE B.60    Reasons for moving or staying 100 100 90 90 80 80 70 70 % of households % of households 60 60 50 50 40 40 30 30 20 20 10 10 0 0 Stay in current Return to place Move Security Access to land, Other location of origin somewhere reasons services, and reasons else employment Group 1 Group 2 Group 1 Group 2 Source: Authors’ calculations using Nigeria IDP Survey, 2018. Source: Authors’ calculations using Nigeria IDP Survey, 2018. 95.  Group 1 is more likely to have all the information required to make a decision about where to live, while Group 2 prefers to obtain information about employment opportunities in other ways (radio, news- papers, through community leaders). Most households reported they have all the information they need to decide whether to move or stay in their current location. IDPs in Group 1 are slightly more likely to have all the information (89 percent vs. 83 percent). In terms of the information needed, both groups require different informa- tion. IDPs in Group 1 are nearly six times more likely to look for information about the political situation compared to households in Group 2. Meanwhile, IDPs in Group 2 want information about the availability of employment (17 percent vs. 1 percent of households in Group 1), since their decision is mainly motivated by access to land, services, and employment (Figure B.61). A higher exposure to armed conflict for IDPs in Group 1 might explain the need for more information about the political situation. Group 2 might be at a disadvantage since they seek information on employment and other conditions, which might be harder to get, and for that reason they are twice as likely to report they prefer to obtain information in other ways (such as via radio, newspapers, or through community leaders) compared to households in Group 1. Policy Implications of the Typologies 96.  The two groups of IDPs can be identified from their location and return intentions. IDPs in Group 1 are living mainly in host communities (85 percent) and not in camps in the state of Borno (89 percent), whereas half of house- holds in Group 2 are in camps located in the states of Borno (65 percent) and Taraba (24 percent). In addition, most households in Group 1 want to stay in their current location, while most of those in Group 2 want to return to their place of origin. These observable differences could prove useful when targeting policy responses to these groups of IDPs in Nigeria, given their different needs and durable solutions. 48  |  Informing Durable Solutions for Internal Displacement   FIGURE B.61    Information required to decide whether to stay or move 100 90 80 70 % of households 60 50 40 30 20 10 0 Have all the Need info about Need info on Prefer to obtain information the political the availability info in other needed situation of employment ways Group 1 Group 2 Source: Authors’ calculations using Nigeria IDP Survey, 2018. 97.  Households in Group 1 require increasing their access to safety nets and gender-responsive programs. IDPs in Group 1 seem to have less access to safety nets because most of them are not located in IDP settlements. Nige- ria’s NEMA and the State Emergency Management Agencies (SEMA) should support IDPs living in host communities that are currently left behind because policy efforts tend to concentrate on IDP camps with structured systems. This is especially relevant for the population in Group 1 as they want to stay in their current location and require better condi- tions to ultimately bring a durable solution to their displacement. In addition, almost half of the households in Group 1 are headed by women. Policy efforts should address gender-based vulnerabilities, as well as GBV and discrimination against IDP women. 98.  IDPs in Group 2 need more access to agricultural land and to improving their skills to diversify their sources of income. Skills and human capital are relevant due to the loss of physical capital during displacement and since eco- nomic inactivity makes it harder for IDPs to find employment. Before and after displacement, households from Group 2 were more likely to have an agricultural livelihood. Besides, their decision to stay or move is guided by access to land, services, and employment. Therefore, policy efforts for IDPs in Group 2 should aim to provide greater access to agricul- tural land before they return to their place of origin, as well as to upgrade their skills to avoid a dependency on aid and to support an income diversification strategy. 99.  The two groups of IDPs have different information needs and durable solutions. Most IDPs have been dis- placed by armed conflict, lost household members, and witnessed trauma. Cognitive and non-cognitive approaches that can help overcome conflict-induced and displacement-induced trauma are particularly relevant for both groups. Durable solutions for IDPs in Nigeria must address various dimensions such as security, living conditions, employment, and food insecurity. IDPs in Group 1 want to stay in their current location and thus living standards have an important bearing on the prospect of a durable solution. Better conditions are required to facilitate the integration of households in their communities, besides improvements in living conditions and access to services. Households in Group 2 want to return to their place of origin, and even though most IDPs have all the information they need to make this decision Volume B: Country Case Studies  | 49 they do not have a clear timeline, as returning might not be feasible due to insecurity and violence. Hence, addressing security concerns is key to allowing households in Group 2 to return to their place of origin. Furthermore, developing a humanitarian response plan to find a durable solution requires providing regular and useful information about the security conditions and income-generating opportunities. Conclusions Informing Durable Solutions 100.  Durable solutions for IDPs must target an array of dimensions, such as food insecurity, housing, educa- tion, and employment, and ensure a safe environment. A majority of the IDPs fall in the extreme poverty and high food insecurity category. They have poor housing conditions, education enrollment rates, and employment options. IDPs in camps are worse off than IDPs in host communities in many respects including housing, health, and education. Additionally, women and girls in IDP households face several challenges. However, almost all IDPs demand support that will ensure their safety. Aid needs to be scaled up, but many targeted durable solutions are also needed to successfully resettle and improve the lives of IDPs. 101.  The presence of IDPs in host communities has had far-reaching social effects on host communities, fur- ther weakening access to education, and health and social services, while also exacerbating strained infra- structure, poverty, and food insecurity. The influx of IDPs into host communities in northeast Nigeria has rapidly increased the population of the host villages and towns. The consequence has been a huge strain on already weak social services and basic infrastructural facilities. Education, health care services, and local markets have particularly felt the strain. There have also been negative impacts on the state of water, sanitation, and hygiene in host communities. Essentially, the presence of IDPs has huge consequences on the living conditions of host communities in northeast Nigeria. 102.  Cultural and religious barriers, and not financial barriers, may be responsible for the low enrollment of displaced children in primary schools. In most parts of northeastern Nigeria, even before the insurgency, secular education was seen as being of lesser value than Islamic education. The very concept of secular or Western education was seen locally as fraudulent and contrary to Islam. This is the core foundation upon which Boko Haram’s founding philosophy is constructed. ‘Almajiri’ or Itinerant Qur’anic education is generally more popular than secular education among most of the Kanuri population, which make up the highest number of displaced persons. Although the major- ity of displaced persons attribute the non-enrollment of their children in schools to financial difficulties, there may be other reasons outside of finances. Primary education is free in public schools in Nigeria. Some schools do charge very nominal dues, for example Parents Teachers Association dues, but usually such dues are very low and rarely a limitation for families that are keen to send their children to school. Moreover, most of the state governments have been very successful in ensuring that there are no financial barriers whatsoever to accessing primary-level education. More can be done to challenge religious and cultural barriers to education in northeast Nigeria as part of a broader humanitarian response plan. 103.  The Boko Haram violence and consequent population displacement has enormous negative impacts on the already precarious education and health facilities in northeast Nigeria. The violence has led to the destruc- tion and suspension of many schools, particularly in Borno, Adamawa, and Yobe states, thereby further disrupting the already fragile educational foundations of the region. Several health facilities and schools have been damaged. This has caused health workers, teachers, and civil servants to flee, thus affecting essential services. Where services are still 50  |  Informing Durable Solutions for Internal Displacement available, providers are overburdened with the increased pressure on services from both host communities and inter- nally displaced families. In addition to loss of livelihoods and poor access to food and water, IDPs also face poor health conditions as well as acute protection risks. 104.  Dependable and accurate information is an important resource for IDPs and should therefore form part of a broader humanitarian response plan. During violent conflicts, information becomes a humanitarian need for displaced persons. It could be as important as food itself. A majority of IDPs yearn for more information on the security situation of their host community and of their homeland. For displaced persons, the hunger for truthful and depend- able information is not out of mere curiosity but out of a fundamental need to know their vulnerabilities so they can negotiate their resilience. They want to know if it is safe to go out or to remain at home, to move to the next town or to remain. They also want to know if it is safe to return to their place of origin. IDPs regularly seek information on available opportunities for their personal development and restoration of their livelihoods either locally or in neighboring towns. In developing a humanitarian response plan, therefore, a key element that deserves greater consideration is how dis- placed persons can be provided with regular, up-to-date, and useful information about their situation. 105. The displacement experience has significant impacts on all aspects of social life and the emotional well-being of IDPs. While IDPs generally face several difficulties, ranging from access to food, safe drinking water, shel- ter, land, and basic services such as health and education, the displacement experience itself exacts a huge emotional toll on them. This is more severe with men, particularly in IDP camps who feel not only a stronger sense of personal loss but also strong personal humiliation and indignity. This humiliation, though often ignored, is a form of trauma and thus attracts its own type of suffering and mourning.67 Displacement has cognitive and emotional meanings and is manifested in different forms in IDP camps and in host communities. The meanings that displaced persons make of their displacement experience is of strong significance and can in many ways influence their coping strategies and their long-term mental stability, thus, this is an important subject for further study. 106.  IDPs in host communities tend to have stronger social and empathetic connections that help them cope with the emotional trauma of displacement and loss. Host communities, despite the strains from IDPs, provide the closest semblance possible of normal social living for IDPs, which is unlike the institutional and deindividualizing nature of the camps. IDPs in host communities in most cases live with distant family relations, friends, and fellow religious believers where they find spaces for sharing and empathetic connections. These spaces help alleviate the emotional toll of displacement, while also providing a personal validation of their individual displacement story and their longing for restoration—not simply of their homes and livelihoods but of their personal dignity. 107.  Farming can provide not only an opportunity for restoration of livelihoods and self-reliance but also an agency for self-efficacy and healing for IDPs. Most IDPs resort to agriculture and subsistence farming. In addition to providing a source of livelihood, most displaced persons have found the process of planting, tending for crops, and harvesting to be therapeutic. The farming cycle gives them something to look forward to while providing a daily means of activity and personal validation. More study is required to understand why a majority of employed IDPs are into agriculture in general, and farming in particular. Whatever the reasons are, there is need to provide more opportunities for farming for more IDPs. Some host communities in Adamawa state have loaned farmland to IDPs, while crops and 67. Udo-Udo et al. 2016. “Narratives of Displacement: Conversations with Boko Haram Displaced Persons in Northeast Nigeria.” Volume B: Country Case Studies  | 51 seedlings have been provided by aid agencies. This practice can be further studied, modelled, and integrated into a larger sustainable humanitarian response plan. 108.  National humanitarian responses can do more to provide support to host communities and IDPs residing in host communities. Nigeria’s NEMA and SEMA can do more to support communities that host IDPs as well as IDPs liv- ing in host communities. Presently, most relief materials are provided to IDP camps and settlements. Since there are no organized structures of patterned living in host communities, enough relief materials are not provided to IDPs in host communities. More attention is understandably given to IDP camps and settlements where there are more structured systems. The consequence is that most IDPs in host communities are left to fend for themselves and to depend on the continued benevolence of their hosts. To attend to the varied needs of IDPs, more subjective and less structured forms of intervention would have to be evolved and applied in host communities—including working with host communi- ties themselves to facilitate local integration. 109.  There are many possible reasons, but further research is needed to have a deeper awareness of why a majority of IDPs living in host communities would rather remain there than return home. An overwhelming majority of IDPs living in host communities prefer to remain there rather than return home. There are many possible reasons for this. Returning to their homes in most cases could be a traumatic experience for returnees. Home could resurrect memories and the pains of loss—of family members and livelihoods. In some cases, IDPs may not have anything to return to, as their homes and livelihoods may have been destroyed. Moreover, there are uncertainties about security conditions in villages and towns that have been liberated from the Boko Haram occupation. Basic services including health care, schools, electricity, and portable water, as well as law enforcement services may not have been restored in liberated villages. There is also the possibility of Boko Haram reattacking territories they previ- ously held. Moreover, several IDPs have found ways to make a living in their host communities and have developed strong interpersonal relationships with members of their host communities thus making return an unattractive proposition. 110.  The nature and patterns of relationships between host communities and IDPs deserve further study. Most host communities have a favorable opinion of IDPs in their communities and vice versa. Further study of the nature and patterns of relations between host communities and IDPs is needed to have a more nuanced understand- ing of the social dynamics that exist between both groups. Such study can potentially help provide models or keys to transforming indigene-foreigner/refugee relationships in other societies. 111.  A pathway for IDPs to access settler status in host communities deserves exploration. A significant number of IDPs have lived in host communities for more than four years. Possibly, the longer IDPs live in host communities, the less likely they are to return. To help IDPs achieve self-fulfillment, a pathway through which they can settle in their host communities should be explored. While a range of settler-indigene tensions may arise— including rights to land ownership, indigeneity, and intermarriage, among others—it is important to have early dis- cussions about long-term settlements with host communities to avoid possible future conflicts. Issues of identity, belonging, and imagined citizenships for IDPs in host communities should be explored more intentionally as part of a broad humanitarian response plan. One of the positive outcomes could be that children of IDPs growing up in host communities will feel less fragmented in their social and personal identities if their parents have a long-term settlement plan. 52  |  Informing Durable Solutions for Internal Displacement IDPs in Somalia Introduction and Country Context 112.  Somalia’s long history of conflict and drought have led the country to be one of the poorest in Sub- Saharan Africa. Since the Siad Barre government collapsed in 1991, the country has experienced successive cycles of conflict, mostly in the south. In 2017, Somalia experienced 1,537 organized violent events, making it the most con- flict-affected country in Africa ahead of South Sudan.68 In addition to civil war among different factions, many recruited along clan lines, the country also experienced violent jihadism, as well as conflicts over land, natural resources, pasture- land, and economic rents; and levels of criminality, interpersonal violence, and GBV are high.69 Somalia is also extremely vulnerable to climate shocks, and has long experienced cyclical droughts, as well as floods, desertification, and land degradation. Together, these cycles of conflict and drought have fragmented society, damaged people’s livelihoods, and caused deep-seated vulnerability: 77 percent of people live under the US$1.90 a day poverty line.70 They have also exacerbated Somalia’s two major famines, which took place in 1992 and 2011, and which together caused over half a million deaths. 113.  Recently, however, Somalia has begun to rebuild a federal state and has taken steps to emerge from conflict. In 2012, the Transitional Federal Government transferred power to a newly established Federal Government of Somalia (FGS), which was set up after a period of national dialogue and consensus building. The federal govern- ment aims to re-establish state authority through an institutional structure that incorporates the federal member states (FMS). To support this, the international community has begun to re-engage in the country. In 2013, international actors and regional countries endorsed a New Deal Compact for Somalia, which laid out key peace and state-building goals, including a special arrangement for Somaliland in the north. In 2017, this was succeeded by the New Partnership for Somalia, which outlines priority areas for development. This is aligned with a National Development Plan (2017–2019) developed by the FGS and the FMS. Substantial progress of building peace has thus been made, though violent events continue and much work remains to build peace further, notably to build trust among the different political entities, agree on a formula for sharing resources, and build local reconciliation. 114.  Significant challenges remain. The country had a gross domestic product (GDP) per capita in 2016 of only US$500, and its poverty rate makes it the third poorest country in Sub-Saharan Africa, after only Burundi and South Sudan. Almost half the population cannot meet the average consumption of food items. Because the economy is mostly pastoral and agro-pastoral, the country’s exposure to climate shocks, particularly drought, poses risks to people’s livelihoods and food security. These risks were visible during the recent 2016–2017 drought, the country’s most severe in years. This led to widespread crop failures, shortages of water and pasture, and livestock mortality, and caused more than 6 million people to face food insecurity. Unlike in previous severe droughts, however, this time the country man- aged to avoid famine. 68. Armed Conflict Location & Event Data Project, “Conflict Summaries of Hotspots of Political Violence.” 69. Federal Government of Somalia. 2018. “Somalia Drought Impact & Needs Assessment.” 70. World Bank, “Somali Poverty Profile 2018.” Volume B: Country Case Studies  | 53 115.  Forced displacement is a massive humanitarian and development challenge in Somalia. Between 1.8 to 2.1 million IDPs, out of a total population of 14.32 million, are currently estimated to live in Somali regions.71 Before the 2016–2017 drought, an estimated 1.1. million people were internally displaced, and over 877,000 Somali refugees lived in neighboring countries, making them one of the largest refugee populations in the world,72 with most living in Yemen, Kenya, and Ethiopia.73 The 2016–2017 drought led a further 926,000 people to be internally displaced.74 Somalia also has a population of refugee returnees. Their numbers have increased in recent years, in part due to the Govern- ment of Kenya’s decision to close the Dadaab Refugee Camp in 2016, but remain low: over 52,000 Somalia refugees have been supported to return to Somalia since 2014, of whom 29,000 returned between January and June 2017 (Fig- ure B.62). Forcibly displaced populations in Somali regions are thus a complex mix of IDPs, returnees, and the caseload of refugees seeking asylum within the country.   FIGURE B.62    Number of displacements occurring by month, January 2016–April 2018 350 300 250 Thousands 200 150 100 50 0 16 6 6 6 16 6 17 7 7 7 17 7 18 8 -1 -1 l-1 -1 -1 -1 l-1 -1 -1 n- p- n- p- n- ar ay ov ar ay ov ar Ju Ju Ja Se Ja Se Ja M M M M N M Conflict Drought N Source: UNHCR-PRMN, Jan. 2016–Apr. 2018. 116.  The drivers of displacement are multiple and overlapping, but are mainly related to armed conflict and climate-related events such as drought: these reinforce one another and exacerbate an already fragile human- itarian context. The recent Somalia Drought Impact Needs Assessment highlights how drought in Somalia has wors- ened conflict over natural resources and pastureland, and how armed conflict and insecurity, in turn, have undermined already precarious agricultural and pastoral livelihoods, thereby further exacerbating displacement. Forcible evictions and land acquisition have also contributed to displacement, particularly in urban areas such as Mogadishu, where land values have risen, and where IDPs often have insecure land tenure and lack access to affordable housing. Economic migration as a survival strategy in the face of precarious livelihoods is also a driver of displacement. 71. Estimates vary across humanitarian agencies working in Somali regions. UNCHR’s Protection and Return Monitoring Network estimates 1.88 million IDPs in the country (https://unhcr.github.io/dataviz-somalia-prmn/index.html), while IOM’s Displacement Tracking Matrix estimates roughly 2 million IDPs (http://www.globaldtm.info/somalia/). The PRMN is the UN’s latest data source on displacement, estimating 1.8 million, but the OCHA humanitarian needs overview reports that 2.1 million people are displaced. 72. “Refugees in the Horn of Africa: Somali Displacement Crisis.” http://data.unhcr.org/horn-of-africa/regional.php (May 5, 2018). 73. Ibid. 74. Federal Government of Somalia. 2018. “Somalia Drought Impact and Needs Assessment.” 54  |  Informing Durable Solutions for Internal Displacement 117.  Displaced populations face marginalization and vulnerability. ‘Gatekeepers’ from powerful clans commonly control access to humanitarian aid in IDP settlements, making it difficult for IDPs to get access to aid in an undistorted fashion. Clan affiliation remains important in much of Somalia and is a critical source of social, financial, and human protection and security. IDPs who have been dislocated from their clan networks or are in minority clans can find them- selves excluded from access to jobs, livelihoods and opportunities, and security. Security is also a challenge within IDP camps, where GBV is common and clan protection is limited. Women and children who lack decision-making power and access to resources within their households are also often particularly excluded and vulnerable. 118.  Finding durable solutions for displaced populations is a challenge, particularly in urban areas. Most recently, displaced people in Somalia have moved predominantly to urban and peri-urban areas seeking resources and humanitarian support. This has put a strain on these areas and contributed to overcrowding, and although the majority of IDPs and host community members currently report good relations, the strain contributes to social tension. 119.  Somalia’s National Development Plan recognizes the need to find durable solutions for forcibly displaced populations. The plan identifies the following five priorities for doing this: (a) Rule of law and governance. The plan recognizes that enhancing these at all levels of the state, including the municipality level, is necessary to ensure safety, security, freedom of movement, and access to basic services, labor markets, and documentation. (b) Access to land and tenure security and inclusive development. The plan recognizes that this is critical to allow IDPs to integrate into their urban environments, particularly in areas where forced eviction is common. (c) Individual documentation, social inclusion, and participation. The plan recognizes that documentation is essential to facilitate freedom of movement and to participate in public affairs. (d) Access to services and labor markets. The plan recognizes that IDPs struggle in getting access to schools, water and sanitation, health care, and jobs, which makes young IDPs particularly vulnerable to being forcibly recruited, and that improving this is essential to reducing that risk. (e) Rural reintegration capacity. The plan recognizes that rural areas lack infrastructure, security, services, and employ- ment and education conditions, and that these must be improved to allow IDPs who wish to return to do so. 120.  The World Bank has been supporting the FGS to find durable solutions for IDPs and returnees. This includes investing in technical assistance to the Somali authorities to enable them to define and formalize the institutional struc- tures that will enable implementation of the National Development Plan (NDP) commitments on forced displacement; a US$3 million grant to a new ‘Secretariat on Mixed Migration and Forced Displacement’ in Intergovernmental Authority on Development (IGAD); support to infrastructure investments in urban areas with high levels of displacement through the urban portfolio; analytical work on the dimensions, drivers, and impacts of forced displacement in the country; and policy dialogue, including through the Somalia Drought Impact Needs Assessment. Volume B: Country Case Studies  | 55 121. Addressing these challenges is complex and requires development as well as humanitarian policy responses. The long-standing development deficits and vulnerabilities of Somali regions, including in host communi- ties, render it challenging to address the needs of forcibly displaced populations effectively. The persistent and cyclical nature of the drivers of migration and conflict contribute to entrenched conditions, which requires a developmental, resilience-based approach to help affected populations cope with these shocks and stresses, combined with continu- ing humanitarian assistance to shore up basic needs. 122.  This chapter seeks to inform such approaches by analyzing displacement trends among IDPs in Somali regions using data collected by the SHFS in 2017. The data highlights the micro-effects of displacement across several dimensions, including poverty, health, food security, education, jobs, sex/gender, housing, and services. The analysis considers the heterogeneity of affected populations, comparing several subsets of IDPs (those living in and out of settlements, displaced by conflict and climate, in men- and women-headed households, recently displaced and in protracted displacement, displaced once and multiple times, and in rich and poor households), as well as host and non-host communities in urban areas, urban and rural residents, and the national population. This information provides a more comprehensive picture of displacement-related impacts and dynamics in Somali regions to better inform development-oriented, area-based solutions.   BOX B.2    Where are the IDPs? Timing of survey sampling and interpretation of spatial results The chapter examines IDPs across Somali regions and is nationally representative; however, the regional distribution of IDPs in the survey sample differs from that of other estimates. According to the Somali High Frequency Survey (SHFS) data, IDPs are clustered in Banadir, Bay, Lower Shabelle, Mudug, and Lower Juba. This differs, however, from UNHCR’s current PRMN (Protection and Return Monitoring Network) data, which has IDPs clustered in Banadir, Bay, Lower Shabelle, Hiraan, and Mudug. In the High Frequency Survey (HFS) sample, certain regions with substantial numbers of IDPs, including Hiraan and Sool (which have 7 percent and 5 percent of the total IDP population, respectively) are undersampled, while others such as Banadir, Mudug, and Lower Juba are oversampled (for instance, Banadir has 22 percent of the actual population but it is 28 percent in the HFS sample, Figure B.63). These differences are methodological. The SHFS sample of settlement IDPs was drawn using IDP location data from 2016, before the most recent drought event. The bulk of drought-related displacements (about 1 million IDPs) occurred from January to October 2017, influencing the spatial distribution of IDP households today.75 Further, the SHFS set of non-settlement IDPs were households in the rural and urban samples, who self-identified as having been displaced. Thus, it was not possible to stratify these households by region ex ante. Because of this, the chapter does not cut the sample of IDPs by region. The results on the regional distribution of IDPs are presented here, but are compared with that of the latest PRMN data and should be interpreted with caution. These differences do not affect how the broader survey results are interpreted. The survey itself was conducted from December 2017 to January 2018 after drought conditions improved, and its findings are nationally representative. The survey results further capture impacts of the drought. The timing of the sampling thus does not affect the accuracy or representative- ness of the survey results themselves, which captures the impact of the drought, but does mean that the results on the spatial distribution of IDPs presented should be interpreted with caution. 75. UN OCHA (United Nations Office for the Coordination of Humanitarian Affairs). 2018. “2018 Humanitarian Needs Overview: Somalia.” 56  |  Informing Durable Solutions for Internal Displacement   FIGURE B.63    Regional distribution of IDPs, HFS sample, and UNHCR-PRMN data Banadir Bay Lower Shabelle Hiraan Mudug Gedo Sool Bakool Middle Shabelle Sanaag Togdheer Galgaduud Woqooyi Galbeed Bari Middle Juba Lower Juba Awdal Nugaal 0 10 20 30 40 50 % of IDP population UNHCR—PRMN IDP population HFS sample—all IDPs HFS sample—settlement IDPs HFS sample—non-settlement IDPs Source: Authors’ calculation using Somali HFS 2017–18 and UNHCR-PRMN data, 2016–18. Displacement Profile 123.  IDPs and non-IDPs have an almost identical demographic structure: both are overwhelmingly young and slightly skew toward male. About one in two national residents76 and IDPs, both in and out of settlements, are under 15 years of age (national residents: 47 percent; IDPs: 51 percent; settlement IDPs: 50 percent). About two in three are under 25 (national residents: 62 percent; IDPs: 65 percent). The majority of IDPs are thus children and youth. IDP and non-IDP households alike have slightly fewer women than men: women make up 48 percent of national residents, non-settlement IDPs, and settlement IDPs (Figure B.64). Moreover, IDPs have a similar household composition as those living in host communities after controlling for other household characteristics (Table B.7).77 124.  As with the national population, every second IDP household is headed by a woman. About 48 percent of IDP households overall are headed by a woman, which is the same as the national population (Figure B.65). IDP 76. References to ‘national residents’, the ‘national population’, the ‘urban population’, ‘urban residents’, the ‘rural population’, ‘rural residents’, ‘host communities’, and ‘non-host communities’ in this chapter exclude IDPs and nomads. 77. The econometric analysis aims to provide some insights about the differences between IDPs and host communities. It only considers data from a cross- section and does not include some variables which might be associated with the status of households, such as time-variant elements, endowments, and social capital, among other determinants. Volume B: Country Case Studies  | 57 households living in settlements are more likely to be headed by a woman (54 percent) compared to IDP households outside settlements (37 percent, p < 0.01, Table B.4). This may be because women are seeking higher levels of security often present in more formal settlements, or because displaced women are separated or disconnected from family/ social networks and have fewer housing options outside formal settlements. The results are robust after considering other household characteristics, as there are no differences in terms of sex, age, and literacy of the household head between IDPs and households living in host communities (Table B.7). 125.  IDP and non-IDP households have similar characteristics. Households have similar numbers of dependents for every working-age adult: IDP households both in and out of settlements have an average of 1.4 each, compared to 1.2 nationally. The exception is women-headed IDP households outside settlements, which have only 1.1 depen- dents for every working-age adult compared to 1.5 in male-headed IDP households outside settlements (p < 0.05). Household sizes are also similar, except that IDP households outside settlements are slightly bigger, with 5.8 people on average compared to the national average of 5.1. Women-headed IDP households in settlements are also larger, with 5.7 people on average, compared to 5.1 in male-headed IDP households in settlements. Other differences between households are not statistically significant (Table B.4).   FIGURE B.64    Population structure for IDPs and non-IDPs by sex and age 1 1 1 0 50 1 1 1 1 19 18 16 19 40 % of population 17 16 17 16 30 7 7 9 6 8 7 7 7 20 26 27 24 27 23 27 24 10 21 0 Men Women Men Women Men Women Men Women National Overall IDP Non-settlement IDP Settlement IDP (excluding IDP) Under 15 years 15–24 years 25–64 years Above 64 years Source: Authors’ calculation using Somali HFS 2017–18.   TABLE B.4    Age dependency ratios and household size by sex of household head Non-settlement IDP Settlement IDP National Man Woman Man Woman Man Woman headed headed Overall headed headed Overall headed headed Overall Percentage of 62.7 37.3 100.0 45.6 54.4 100.0 51.7 48.3 100.0 households Dependency ratio  1.5  1.1   1.4  1.2  1.5   1.4  1.2  1.3   1.2 Household size  5.9  5.5   5.8  5.1  5.7   5.4  5.1  5.0   5.1 Source: Authors’ calculation using Somali HFS 2017–18. 58  |  Informing Durable Solutions for Internal Displacement 126.  Most IDP households are in urban areas and in formal settlements. About three in four IDP households overall (75 percent) are in urban areas (Figure B.66); and 6 in 10 IDPs (62 percent, Figure B.65) live in formal settlements. All such settlement IDPs in the SHFS sample are in urban areas.  FIGURE B.65    IDP profile   FIGURE B.66    Urban/rural composition of IDPs 100 100 80 80 % of overall sample % of households 60 60 40 40 20 20 0 0 C ett eme IDP n DP lim io P ev e an Oth nt a e r ot ea d Pr tra d is p ra d ed d d u e tto le p 0 60 -p or r C ett eme IDP n DP lim io P a ev e M n h ent ot e ed ot ed is p ra d ed d d u e tto e p 0 60 -p or r M h e oo oo e nc N n h ade pr de D is ot cte ac ce cte m onc To 4 om e nc D Dis rot acte ac ce cte m onc Bo ltipl Tom 4 C or v t ID C or v t ID Bo ltip e on o on o N n h ad pr ad fli me t I m N P fli me t I N P at le W at le l -s ttl all al on le n on le n a e r N S ver r o N S ve P pl la pl la O O -s tl om ct ct t on e on e W D Urban Rural Source: Authors’ calculation using Somali HFS 2017–18. 127.  Most IDPs have not gone far from home. About 7 in 10 IDP households live in the same districts as they did originally, and fewer than 1 in 10 are in a different region, federated member state, or country. Those who are displaced multiple times are more likely to travel out of their districts than those displaced only once (p < 0.01). Households headed by a woman are significantly more likely to stay in their districts than those headed by a man (women-headed households: 61 percent; men-headed households: 86 percent, p < 0.01). The limited distances traveled could be linked to limited freedom of movement for women, proximity of available humanitarian resources or secure settlements or possibly due to security risks linked to traveling long distances from home and outside environments with available clan protection (Figure B.67). 128.  Most IDPs have been displaced only once and have traveled to their current locations with their fami- lies, though this finding should be interpreted with caution. Approximately four in five IDPs (75 percent of non- settlement IDPs and 81 percent of settlement IDPs) report being displaced once, and only a tiny minority of IDPs report being displaced more than twice (Figure B.65). However, these findings should be interpreted with some caution, since they run counter to more common understandings of forced displacement in Somali regions, in which displaced populations often experience multiple displacements, due in part to forced evictions and/or new cycles of violence.78 Approximately 4 in 5 IDPs have traveled with their families to their current locations, about 1 in 10 alone, and about 1 in 10 as part of a larger group (Figure B.68). 78. UNHCR. 2016. “Internal Displacement Profiling in Mogadishu”; Federal Government of Somalia. 2017. “Somalia Drought Impact and Needs Assessment.” Volume B: Country Case Studies  | 59   FIGURE B.67    Original location relative to current location for   FIGURE B.68    Trends in traveling to current IDPs location 100 100 80 80 % of IDP households % of households 60 60 40 40 20 20 0 0 With my family Alone With a larger -s lem DP en P P e ce t ad d ed D isp rac d d ul e Bo ple To 40 60 -p r r en on oo oo he de ot cte te m nc on em t ID D at len ti I tI ev m p N P group ed d o an ea Pr ra ll l n tto ra lim io et e M nh ot ac e C v ve pl c pr or is la a O om ot on tt t ct N Se N fli W D C Same district Same region different district Different region Outside country Source: Authors’ calculation using Somali HFS 2017–18. 129.  Climate-related events (drought, famine, or flood) and conflict are the main causes of displacement cited by IDPs. About two in five IDP households (38 percent) are displaced from their original locations because of climate-related events (drought, famine, or flood). About another two in five (40 percent) are displaced because of armed conflict in their village or another village (Figure B.69).   FIGURE B.69    Reason for leaving original location   FIGURE B.70    Reason for arriving at current location 100 100 80 80 % of households % of households 60 60 40 40 20 20 0 0 -p or -p or P om nt P an IDP ot ead d ed D isp trac ed d ul e Bo iple To 40 60 r P nt DP P ev ce M n h nt ot eaded ed is p rac ed ac ce ted ul e Bo tiple p 0 60 r oo oo h de te m c m onc Tom 4 ID e ID -s ttle ll ID ID on o on o e n te en o ct h ad t t m p N P I c N P t ed d o an ea Pr tra Protra ll t on tlement ol tle en tto tto an e ra ra ed d M h e lim vi ac e o o ve ve m m et m pl c a pr pr C or is la pl la -s tle a O O om ct on et et D is on e N N fli N S N S W D D W C Armed conflict in village Better security Armed conflict in other village Water for livestock Increased violence but not conflict Home/land access Discrimination Education/health access Drought/famine/flood Employment opportunities Other Join family Humanitarian aid Source: Authors’ calculation using Somali HFS 2017–18. 60  |  Informing Durable Solutions for Internal Displacement 130.  The main reason IDPs live where they do is improved security. This is true whether their households are in or out of settlements, displaced by climate events or conflict, headed by men or women, or are rich or poor (Box B.3). Over three in five non-settlement and settlement IDP households, and almost four in five households displaced by conflict, report that they are in their current locations because of better security rather than for other reasons such as access to humanitarian assistance or better livelihoods. These patterns differ slightly for households displaced by climate, but even among these, approximately half (53 percent, p < 0.05) are in their current locations for better security, and the rest because they can get better access to livelihoods, employment, land and housing, or humanitarian assistance (Fig- ure B.70). There are some remaining differences in deciding factors across types of IDPs, with IDPs in settlements more likely to cite joining family as a reason for being where they are, and poor IDPs less likely than non-poor IDPs to cite security (poor: 58 percent, non-poor: 82 percent, p < 0.01), but overall security is the main driver for all groups.   BOX B.3    Drivers of displacement in Somali regions Although the household survey indicates that most people are displaced either by conflict or climate-related events, in practice, these categories are intertwined. The drivers of displacement in Somali regions are overlapping, multiple, and complex. Forced displacement in Somali regions is a consequence of decades of internal conflict, insecurity, political uncertainty, human rights violations, and governance failures, compounded by cyclical environmental challenges, including periods of acute drought and famine. While survey respondents were asked to indicate one primary driver motivating migration, it is more likely that individuals and households were influenced by several interrelated factors, including both climate and security-related events. Indeed, drought conditions in Somali regions have been known to exacerbate conflict, while the impacts of drought are worsened by conditions of violence and insecurity. The Somalia Drought Impact Needs Assessment reports that in Somali regions, drought conditions in 2017 have exacerbated conflicts over pasturelands and natural resources, with mediating impacts on food prices and livestock, and highlights the upsurge in communal and political violence in 2017 (particularly in the southern and central regions of the country) which compounded the devastating humanitarian and development impacts of drought and contributed further to displacement dynamics. Source: Federal Republic of Somalia, World Bank, UN and European Union. 2018. Somalia Drought Impact Needs Assessment. Vol II, page 147. 131.  Most IDPs have been displaced in the last five years, and those outside settlements more recently. IDPs outside settlements tend to have been displaced more recently than those in settlements: settlement and non- settlement IDPs alike arrived in their current locations about two years ago, but on average non-settlement IDPs have been displaced for about two and one-fourth years, whereas IDPs in settlements have been displaced for three years (p < 0.01). Non-settlement IDPs are also quicker to settle once originally displaced, taking on average four months to do so, compared to about a year for IDPs in settlements (Figure B.71, p < 0.01). 132.  Conflict and climate-driven IDPs have experienced continued and ongoing displacement. The pattern of displacement (Figure B.72; Figure B.73) shows clear peaks, which have increased since 2013. These peaks, however, are not as dramatic as that shown by similar data in other countries in the region with large-scale displacement.79 Although the Somali drought displacement pattern shows spikes between or at the edges of the Gu and Deyr rainy seasons, the displacement spikes also correlate less clearly to climate and conflict events than they do elsewhere. This suggests that displacement in Somali regions reflects underlying and continual uncertainties related to climate and conflict rather than one-off shocks. 79. For example, see South Sudan case study. Volume B: Country Case Studies  | 61   FIGURE B.71    Years since displacement and arrival in current location 8 7 6 5 Years 4 3 2 1 0 -p r P en DP P e ce t he ded ed ot ted ed m nce le To 4 0 60 r on o en oo ID D tip N Po at len ad ct em t I tI ev Pr rac m p ed o an a ul ra ll ttl n M he tto ra lim io ac ed e ot C v ve Se m Bo pl c an pr or is la tle O om ot D p ct et s N fli i -s W D on on C N Displaced from original location Arrived at current location Source: Authors’ calculation using Somali HFS 2017–18.   FIGURE B.72    Conflict events and dates of displacement of conflict-driven IDPs 10 400 % of conflict-displaced IDPs 9 350 8 300 Conflict events 7 6 250 5 200 4 150 3 100 2 1 50 0 0 Jan-2009 Apr-2009 Jul-2009 Oct-2009 Jan-2010 Apr-2010 Jul-2010 Oct-2010 Jan-2011 Apr-2011 Jul-2011 Oct-2011 Jan-2012 Apr-2012 Jul-2012 Oct-2012 Jan-2013 Apr-2013 Jul-2013 Oct-2013 Jan-2014 Apr-2014 Jul-2014 Oct-2014 Jan-2015 Apr-2015 Jul-2015 Oct-2015 Jan-2016 Apr-2016 Jul-2016 Oct-2016 Jan-2017 Apr-2017 Jul-2017 Oct-2017 Battle Remote violence Riots/protests Violence against civilians Conflict-displaced IDPs Source: Authors’ calculation using Somali HFS 2017–18, ACLED (conflict events 2006–2017). 133.  Most IDPs intend to stay in their current locations and only a few have revisited their original residence. About 7 in 10 IDPs (70 percent) wish to stay in their current locations, and only 2 in 10 (23 percent) intend to return to their original place of residence. A tiny minority intends to move elsewhere (Figure B.74). Few IDPs have revisited their original homes: over 9 in 10 IDPs have not gone back to their original residences. Those who have returned have done so mainly to visit family (Figure B.75). The return intention of IDPs in Somalia is equivalent for those that were displaced by the drought, compared to IDPs displaced by other reasons. Only receiving assistance is weakly associated with an intention to move from the current location (Table B.5).80 80. The econometric analysis aims to provide some insights about the different return intentions between IDPs. It only considers data from a cross-section and does not include some variables which might be associated with the decision to stay or move, such as time-variant elements, endowments, and social capital, among other determinants. 62  |  Informing Durable Solutions for Internal Displacement   FIGURE B.73    Rainfall anomalies, Gu-Deyr seasons, and displacement dates of climate-driven IDPs 10 450 % of conflict-displaced IDPs 9 400 8 350 Rainfall anomaly 7 300 6 250 5 200 4 3 150 2 100 1 50 0 0 Jan-2009 Apr-2009 Jul-2009 Oct-2009 Jan-2010 Apr-2010 Jul-2010 Oct-2010 Jan-2011 Apr-2011 Jul-2011 Oct-2011 Jan-2012 Apr-2012 Jul-2012 Oct-2012 Jan-2013 Apr-2013 Jul-2013 Oct-2013 Jan-2014 Apr-2014 Jul-2014 Oct-2014 Jan-2015 Apr-2015 Jul-2015 Oct-2015 Jan-2016 Apr-2016 Jul-2016 Oct-2016 Jan-2017 Apr-2017 Jul-2017 Oct-2017 Gu season Deyr season Climate-displaced IDPs Rainfall anomaly Rainfall average Source: Authors’ calculation using SHFS Wave 2, Vulnerability Analysis Mapping Unit (VAM) (Rainfall anomalies 2006–2017). Note: Rainfall anomaly is the monthly deviation of rainfall from the long-term average. The long-term rainfall average is scaled to 100, thus deviations are seen relative to this ‘100’ threshold.   FIGURE B.74    Return intentions of IDPs  FIGURE B.75    Trends in revisiting the original residence location for IDPs 100 100 80 % of households 60 80 40 % of IDP households 20 60 0 -s lem DP en P P e ce t ad d ed D isp rac d d ul e le To 40 60 -p r r en on o oo he e ot te te m nc 40 m ID D tip N Po at len d I tI ev Pr ac m p ed o an ea ll le nt tto ra lim io r d et e M h ot ac e C v ve Bo pl c an pr or is la O om ot on tt t ct N Se N 20 fli W D on C Don't want to move 0 Original place of residence Not gone Visit Check New area back family property status Source: Authors’ calculation using Somali HFS 2017–18. 134. The return intentions of IDPs are strongly motivated by security considerations. Almost 8 in 10 non- settlement IDPs and 9 in 10 settlement IDPS cite security as a motivation for wanting to stay where they are. Less than half cite other factors, which include homes, land, livestock, and employment; health, education, and humanitarian aid; or family. Settlement IDPs are more likely than non-settlement IDPs to cite security as a reason for wanting to stay where they are (89 percent of settlement IDPs vs. 75 percent of non-settlement IDPs, p < 0.01), which is likely because higher levels of security are available in formal settlements compared to outside formal settlements. Apart from that, IDPs cite similar motivations for wanting to stay where they are, whether their households are in or out of settlements, headed by men or women, displaced by conflict or climate, or are rich or poor (Figure B.76) Volume B: Country Case Studies  | 63   FIGURE B.76    Push factors for IDPs who don’t want to move 100 % of households who do 80 not want to move 60 40 20 0 P P P ce t ed ed ed ed ce 40 60 or le or en D D D tip Po o en on ad ad ct ct lI tI tI ev m p -p ul ra ra To l ol en en tto he he on ra ed m e ot ot vi ve at m m Bo N ac Pr an an pr ed or lim e tle O pl ttl om M ot ac ct et C is Se N fli pl -s W D on is on D C N Security Home, land, livestock, employment Health, education, humanitarian aid Family Source: Authors’ calculation using Somali HFS 2017–18. 135.  IDPs who do want to move have a broader range of motivations. These include getting better security, as well as family ties and improved housing or access to land, livestock, and employment. Over 7 in 10 IDPs who want to move cite security as a reason for wanting to do so, whether they are in or out of settlements, displaced by conflict or violence, live in households headed by men or women, or are rich or poor (Figure B.77). Yet at least 6 in 10 IDPs who want to move cite family as a motivation, and—apart from the poorest IDPs, who may get better services by being displaced—at least 5 in 10 are motivated by homes, land, livestock, and employment. Among the IDPs who want to move, richer IDPs plan to do so sooner than others (p < 0.01), which might reflect the lower capacity of poorer families to bear the costs of moving and to deal with uncertain livelihoods (Figure B.78).  FIGURE B.77    Pull factors for IDPs who want to move 100 % of households who do 80 not want to move 60 40 20 0 t P P P e ed ed ed ed e le 40 60 or r en oo nc c D ID D tip Po on ad ad ct ct I tI ev m p -p e ul tra ra ll t To ol en en he he tto on ra ed m e ot vi ro ve at em em Bo N ac Pr an an ed or p lim O pl ttl ttl om M ot ac ct C is Se se N fli pl W D - on is on D C N Security Home, land, livestock, employment Health, education, humanitarian aid Family Source: Authors’ calculation using Somali HFS 2017–18. 64  |  Informing Durable Solutions for Internal Displacement  FIGURE B.78    Return timeline for households that   FIGURE B.79    Legal identification and access to doc- intend to move umentation restitution mechanisms81 100 80 70 % of housholds that 80 60 intend to move Percentage 60 50 40 40 30 20 20 10 0 0 -p or on tle e P C or t IDP W at iole P v e M n h ent ot hea ded P otr d is p ra ed ed d d u e tto le p 0 60 r oo -p or om e e nc pr de ac ce cte m onc To 4 W imaviol P M n h vene N n head t pr ad d P ot ed ac ce cted ed d d Bo ulti ce p 0 N P 0 n P tto ple R an ho t C set lem ident fli m t t C or t IDP N S ral res t r C set lem ID ct en ID Bo ltip on o n- s s on tle e n oo D Dis rot act Tom 4 6 a e c ot e e is p tra te an ba ID ct en ID U no ho on o m N P s e om te en m on a - tt all fli m nt D Dis ro rac on e r id an e rb r ll n r U U era lim v e N S ve pl la a O v pl la a O on e rb u - tt N l Don't know yet More than 12 months Legal identification Access to restroration mechanisms 6–12 months Less than 6 months Source: Authors’ calculation using Somali HFS 2017–18. z Dependent variable: move vs. stay (reference) Independent variables (1) (2) (3) (4) (5) Displaced by the drought 0.636** 0.498 0.454 0.414 0.510 Overall satisfaction82 −0.100 −0.098 −0.100 −0.100 −0.086 Year of displacement −0.484 −0.469 −0.540 −0.606 −0.597 Poor household 0.349 0.230 0.241 0.139 Receive remittances 1.094 0.894 0.921 1.002 Receive assistance 0.551* 0.563* 0.545* 0.522* Household size 0.063 0.042 0.066 Age-dependency ratio −0.393 −0.440 −0.400 Share of women in the household −1.126 −1.022 −0.974 Share of children in the household 2.011 2.361 2.261 Share of literate members in household −0.399 −0.080 −0.129 Household head: men −0.074 −0.138 Household head: age 0.006 0.003 Household head: literate −0.346 −0.256 Improved water sources −0.231 Improved sanitation −0.378 Electricity 0.005 Region fixed effects Yes Yes Yes Yes Yes Observations 3,603 2,351 2,302 2,302 2,299 Source: Authors’ calculation using the Somali HFS 2017–18. Note: Significance level: 1 percent (***), 5 percent (**), and 10 percent (*). The coefficients were estimated from a logistic regression. The poverty status used in the regression was derived from total core consumption and a rescaled poverty line. 81. Access to legal identification is calculated at the individual level, whereas access to restoration mechanisms is calculated at the household level. 82. Refers to self-reported current satisfaction in terms of (a) livelihood/employment opportunities available, (b) quality of care at local health clinic, and (c) quality of primary education. Volume B: Country Case Studies  | 65 136.  Only a small proportion of Somalis, and an even smaller proportion of IDPs, have a legal identification or an access to mechanisms to restore documents. About 17 percent of IDPs have a legal identification, compared to 36 percent of urban residents (p < 0.01) and 50 percent of urban host community members (p < 0.01); similarly, few have access to mechanisms to restore documents. IDPs in households headed by a woman are more likely to have identification compared to those in households headed by a man (p < 0.01). The poorest 40 percent of IDPs are also less likely to have identification than the richest 60 percent (p < 0.05) and have less access to document restoration mech- anisms (p < 0.01; Figure B.79). Other than this, the rate of legal identification ownership does not differ much according to the displacement circumstances of IDPs. Standard of Living 137.  The incidence and depth of poverty are greater among IDPs than urban residents, but about the same as among rural residents. The poverty headcount ratio is the proportion of a population who live under the poverty line: it indicates how widespread poverty is. About three in four IDPs (74 percent) live under the US$1.90 PPP (2011) per day per capita international poverty line. Poverty is more widespread among IDPs than among urban residents (63 percent, p < 0.05), but there are no significant differences in the incidence of poverty when comparing IDPs and rural residents, 70 percent of whom are poor83 (Figure B.80). However, IDPs in Somalia are equally as poor as host communities after controlling for region fixed effects, receiving remittances, and demographic and housing characteristics (Table B.6).84 Poor and non-poor households differ in terms of house size and the share of literate members within the household, along with access to electricity and improved sanitation. Poverty is also deeper among IDPs than urban residents. The poverty gap measures how much less the average poor person consumes relative to the international poverty line: it measures not how widespread poverty is, but how deeply the average poor person feels it. In Somali regions, the pov- erty gap among IDPs relative to the international poverty line is 35 percent, meaning that IDPs below the poverty line typically consume only 65 percent of what is consumed by those who are at the US$1.90 per day per capita threshold. This gap is greater than that of urban residents (24 percent, p < 0.01) and the national population (27 percent, p < 0.01), but does not differ significantly compared to rural residents (32 percent) (Figure B.81).   FIGURE B.80    Poverty headcount ratio   FIGURE B.81    Poverty gap 100 60 80 50 % of poverty line % of population 40 60 30 40 20 20 10 0 0 Overall IDP Settlement IDP 2016 Host urban Non-host urban Urban resident Rural resident National resident Settlement IDP Non-settlement IDP Conflict or violence Climate event Not protracted Protracted Displaced once Displaced multiple Woman headed Man headed Overall IDP Settlement IDP 2016 Host urban Non-host urban Urban resident Rural resident National resident Non-settlement IDP Settlement IDP Conflict or violence Climate event Not protracted Protracted Displaced once Displaced multiple Woman headed Man headed Source: Authors’ calculations using Somali HFS 2016–18. 83. National populations reported in this chapter are of national residents, which include urban and rural residents, and exclude IDPs and nomads. 84. The econometric analysis aims to provide some insights about the differences between IDPs and host communities. It only considers data from a cross-section and does not include some variables which might be associated with the poverty status of households, such as time-variant elements, endowments, and social capital, among other determinants. 66  |  Informing Durable Solutions for Internal Displacement 138.  Poverty is more widespread and deeper among IDPs than non-host communities, but there is no signif- icant difference when comparing IDPs and host communities. The poverty headcount ratio among IDPs (74 per- cent) is higher than that of non-host communities in urban areas (64 percent, p < 0.05) (Figure B.80). The depth of poverty among IDPs is also greater. The poverty gap among IDPs (35 percent) is higher than that of non-hosts in urban areas (24 percent, p < 0.01), but there is no significant difference in the poverty gap when comparing IDPs and host communities (Figure B.81). 139.  IDPs in settlements are almost as poor as IDPs outside settlements. There is no significant difference in how widespread poverty is when comparing IDPs who live in settlements (76 percent) with those living outside settlements (73 percent) (Figure B.80), or in how deep the poverty gap is (settlement IDPs: 34 percent; non-settlement IDPs: 36 per- cent) (Figure B.81). 140.  Poverty is much more widespread and deeper among IDPs displaced by climate rather than conflict. Over four in five IDPs (88 percent) displaced by climate-related events (drought, famine, or flood) live under the US$1.90 PPP (2011) per day per capita international poverty line, compared to only three in five IDPs (65 percent, p < 0.01) displaced by conflict (Figure B.80). Poverty is also deeper among climate-displaced IDPs under the poverty line, who have a poverty gap of 40 percent, compared to 30 percent for poor IDPs displaced by conflict (p < 0.05). This means that IDPs displaced by climate events (drought, famine, or flood) are typically consuming only 60 percent of what is consumed at the US$1.90 per day international poverty line threshold (Figure B.81). 141.  Poverty is more widespread among recent IDPs than those in protracted displacement and among those who have been displaced only once. The poverty headcount ratio among IDPs who have been displaced for less than five years (76 percent) is significantly higher than that of IDPs who have been displaced for longer than five years (56 percent, p < 0.01)—though notably, all those in the SFHS sample who have been displaced for more than five years are in urban areas. The poverty headcount ratio among IDPs who have been displaced only once (73 percent) is signifi- cantly higher than that of IDPs who have been displaced multiple times (57 percent, p < 0.01). (Figure B.80). 142.  Poverty is somewhat more common among IDP households headed by men than women. About 75 per- cent of households headed by men live under the international poverty line, compared to 64 percent of households headed by women (Figure B.80, p < 0.1). 143.  Hunger is more common among IDPs than hosts, urban residents, and rural residents. About 55 percent of IDP households went at least once without having food of any kind in the last four weeks, compared to 17 percent of the host community (p < 0.01), 25 percent of urban residents (p < 0.01), and 43 percent rural residents (p = 0.107). While being inside or outside a settlement had no significant relation to hunger, IDPs displaced by conflict are more likely to face hunger than those displaced by climate events (p < 0.05), despite being less poor. This could indicate that conflict-driven IDPs are in areas that are more difficult for humanitarian actors to reach. IDPs that are in protracted displacement, or displaced more than once, are also more likely to face hunger than those who have been displaced for less time (p < 0.01) or displaced once (p < 0.01). Poor IDPs are more likely to be hungry than non-poor (p < 0.1, Fig- ure B.82). IDPs are also more likely to report hunger and less likely to be satisfied with their current conditions, compared to households in host communities after controlling for various household characteristics (Table B.7). Volume B: Country Case Studies  | 67   TABLE B.6    Demographic attributes of poor households Dependent variable: poor vs. non-poor status (reference) Independent variables (1) (2) (3) (4) (5) Host community Reference group Reference group Reference group Reference group Reference group IDP household 0.570* 0.180 0.078 0.055 −0.353 Receiving assistance 0.446 0.510 0.563 0.619 Receiving remittances −0.965*** −0.947** −0.923* −0.523 Household size 0.972*** 0.974*** 0.998*** Age-dependency ratio −0.114 −0.081 −0.292 Share of women in the household −0.360 −0.452 −0.187 Share of children in the household 1.451 1.409 2.163 Share of literate members in −1.560** −1.380** −1.474** household Household head: men 0.277 0.328 Household head: age −0.001 0.004 Household head: literate −0.210 −0.071 Improved water sources −0.094 Improved sanitation −0.710* Electricity −1.451*** Region fixed effects Yes Yes Yes Yes Yes Observations 949 949 946 944 944 Source: Authors’ calculation using the Somali HFS 2017–18. Note: Significance level: 1 percent (***), 5 percent (**), and 10 percent (*). The coefficients were estimated from a logistic regression. The poverty status used in the regression was derived from total core consumption and a rescaled poverty line.   FIGURE B.82    Hunger incidence in the last 4 weeks 100 % of households 80 60 40 20 0 -p r es P t n- st st si t nt en P P e ce t he ded ed ra d ed m nce le To 40 60 r on oo en re en en oo ot te l r ID m t ID D no ho ho tip de at len ad ct P id al id tI ev Pr rac m p ed o an a ul na ll an n ur s tle n M he tto io ra lim io ac d R n re rb ba et e ot pl ce C v at ve -s em Bo N an pr U r or is la U N O a l om ot D isp rb on tt ct N Se N U fli W D on C Source: Authors’ calculation using Somali HFS 2017–18. 68  |  Informing Durable Solutions for Internal Displacement   TABLE B.7    Demographic attributes of IDPs and host communities Dependent variable: IDP vs. host community (reference) Independent variables (1) (2) (3) (4) (5) Household size −0.034 −0.020 −0.048 −0.151 −0.144 Age-dependency ratio −0.042 −0.030 0.649 0.367 0.320 Share of women in the household −0.294 −0.310 0.231 0.288 0.440 Share of children in the household 1.634 1.677 −1.561 −0.690 −0.539 Share of literate members in household −1.269 −0.742 −0.174 −0.068 −0.158 Household head: men 0.419 0.313 0.283 0.331 Household head: age 0.001 −0.008 −0.004 0.000 Household head: literate −0.491 −0.226 −0.185 −0.165 Improved water sources 1.856*** 2.199*** 2.250*** Improved sanitation −1.246** −1.020* −0.981 Electricity −2.762*** −2.514*** −2.539*** Receiving remittances −1.525*** −1.618*** Household reported hunger in past 1.162** 1.158** month Overall satisfaction85 −1.094* Observations 946 944 944 944 944 Source: Authors’ calculation using the Somali HFS 2017–18. Note: Significance level: 1 percent (***), 5 percent (**), and 10 percent (*). The coefficients were estimated from a logistic regression. 144.  About one in four IDPs have access to improved housing, which is much worse than among the national population and host and non-host communities, but similar to the share among rural residents. Improved housing is defined as a structure made of wood, concrete, or block, and intended for habitation. About one in four IDPs (26 percent) currently have access to improved housing or had it before being displaced (27 percent). This is much lower than the share of the national population (59 percent), host communities (80 percent), and non-host communi- ties (75 percent) (p < 0.01) who have improved housing, but is not significantly different from the share of rural residents (18 percent) who have such access. The quality of housing is mostly homogenous for different types of IDPs: it is low for most groups, and does not differ significantly whether they are in settlements or not, displaced once or more, or are displaced by conflict or climate events. The only exception is the pre-housing quality of poor and non-poor IDPs: although they have similar rates of improved housing at present, non-poor IDPs had better housing than poor IDPs before displacement (p < 0.01) (Figure B.83). 85. Refers to self-reported current satisfaction in terms of (a) livelihood/employment opportunities available, (b) quality of care at local health clinic, and (c) quality of primary education. Volume B: Country Case Studies  | 69   FIGURE B.83    Percentage of population living in improved housing, now and before displacement 100 80 % of households 60 40 20 0 Poor Overall IDP Urban host Urban non-host Urban resident Rural resident National resident Non-settlement IDP Settlement IDP Conflict or violence Climate event Not protracted Protracted Displaced once Displaced multiple Woman headed Man headed Bottom 40 Top 60 Non-poor Overall IDP Origin Now Source: Authors’ calculations using Somali HFS 2017–18. 145.  About 8 in 10 IDPs have access to improved drinking water, but this does not account for likely over- crowding of drinking water access points in settlements, so should be interpreted with caution. The share of IDPs with access to improved drinking water (78 percent) is about the same as the share of the national population (77 percent), urban residents (85 percent), and non-hosts in urban areas (85 percent) who have such access. This is higher than among rural residents, only about 56 percent of whom have such access (Figure B.84, p < 0.01), and is similar across most types of IDPs, whether they are in or out of settlements, displaced by climate or conflict, were dis- placed recently or long ago, live in households headed by men or women, or are relatively rich or poor (Figure B.84). The differences in access to improved water sources for drinking water between IDPs and host communities are robust after considering other household characteristics (Table B.7). However, this finding should be interpreted with caution, as the survey question on which it is based does not account for (nor enable disaggregation for) possible overcrowding in access points for water, which other analysis has indicated is a serious problem. IDPs are reportedly 2.5 times more likely than others to experience problems with water points, including overcrowding.86 146.  IDPs appear on the surface to have better access to improved sanitation than rural residents, but this advantage disappears when discounting IDP households who share such facilities. Almost 8 in 10 IDPs have access to improved sanitation when including those whose households share such facilities as well as those who use them exclusively. This is about the same as among the national population and urban residents, and more than among rural residents (p < 0.01), only 6 in 10 of whom have access to such facilities. Yet such facilities are often overcrowded and are no longer classified as being ‘improved’ if they are shared. When discounting those who share, the higher rates of access among IDPs disappears. Only half of IDP households have their own exclusive access to improved sanitation facilities, which is about the same as among rural residents, and significantly less than among the national population, urban residents, host communities, and non-hosts in urban areas. After adjusting for sharing, there are no significant differences in improved sanitation access across different types of IDPs (Figure B.85). 86. Federal Government of Somalia. 2017. “Somalia Drought Impact and Needs Assessment.” 70  |  Informing Durable Solutions for Internal Displacement   FIGURE B.84    Access to improved drinking water for IDPs and residents 100 80 % of households 60 40 20 0 Overall IDP Urban host Urban non-host Urban resident Rural resident National resident Non-settlement IDP Settlement IDP Conflict or violence Climate event Not protracted Protracted Displaced once Displaced multiple Woman headed Man headed Bottom 40 Top 60 Poor Non-poor Overall IDP Source: Authors’ calculation using Somali HFS 2017–18.   FIGURE B.85    Access to improved sanitation for IDPs and residents 100 80 % of households 60 40 20 0 Overall IDP Urban host Urban non-host Urban resident Rural resident National resident Non-settlement IDP Settlement IDP Conflict or violence Climate event Not protracted Protracted Displaced once Displaced multiple Woman headed Man headed Bottom 40 Top 60 Poor Overall IDP Non-poor Unadjusted for sharing Adjusted for sharing Source: Authors’ calculation using Somali HFS 2017–18. 147.  Toilet crowding is more common among climate-displaced, non-settlement, and poorer IDPs. Having access to toilets is important in stopping disease and preventing GBV. IDPs have two households per toilet, meaning that toilet crowding is more common when compared to the national population, urban and rural residents, and hosts and non-hosts in urban areas, all of whom have fewer than one household per toilet. There are also large disparities in toilet access among different types of IDPs. IDP households in settlements, in protracted displacement, headed by women, and in the top 60 percent of households have between one and two households per toilet. Non-poor households experience less crowding than poor households (p < 0.05). Climate-displaced and non-settlement IDPs are much worse off, with three or more households per toilet (Figure B.86). This may reflect the rapid recent increase in drought-induced displacement: existing toilet facilities are likely insufficient to accommodate such a rapid expansion of migration flows. Volume B: Country Case Studies  | 71   FIGURE B.86    Number of households sharing a toilet   FIGURE B.87    Households more than 30 minutes from services 5 Number of households 50 4 sharing a toilet 3 40 % of households 2 30 1 20 0 Non-settlement IDP Settlement IDP Climate event Not protracted Protracted Bottom 40 Top 60 Poor Non-poor Overall IDP Host urban Non-host urban Urban resident Rural resident National resident Conflict or violence Displaced once Displaced multiple Woman headed Man headed 10 0 Urban host Settlement IDP Non-settlement IDP Overall IDP Water point Health School Market Source: Authors’ calculation using Somali HFS 2017–18. 148.  Host communities are closer to services than settlement IDPs. Host communities are more likely to be less than 30 minutes away from the closest health facility (p < 0.05), the nearest primary school (p < 0.1), and the closest market (p < 0.05), than settlement IDPs. However, there are no statistically significant differences between host com- munities and settlement IDP households in how far they are to the closest water point. Non-settlement IDPs are also at similar distances as settlement IDPs for all four services (Figure B.87). 149.  IDPs have lower access to charged mobile phones with network than urban resident communities. IDPs are less likely to have enough electricity to charge mobile phones than urban hosts and urban residents overall (p < 0.01 each). Conflict-motivated IDPs have more access than climate-driven IDPs (p < 0.05). The richest 60 percent and the non-poor are also more likely to have sufficient electricity to charge phones (p < 0.01 and p < 0.1 respec- tively; Figure B.88). IDPs are more likely than urban residents to be more than 15 minutes away from the closest point where they can get mobile phone reception (p < 0.01), but about as far as host communities. Non-settlement IDPs are closer to phone network reception than settlement IDPs (p < 0.05), and men-headed households are closer than women-headed ones (p < 0.01) (Figure B.89). 150.  IDPs have less access to health care than urban residents and more than rural residents—but the rates of access should be interpreted with caution. IDPs are twice as likely as urban residents, but about half as likely as rural residents, to give birth at home rather than in a maternity clinic, maternal and child health center, or hospital: almost 4 in 10 IDP women, less than 2 in 10 urban women, but 7 in 10 rural women (p < 0.01; Figure B.90) who gave birth in the last two years have done so at home. These figures, however, are higher than expected. IDPs are also much less likely than urban residents, but more likely than rural residents, to have their births attended by skilled health staff: only half of IDP women who gave birth in the last two years have done so assisted by a nurse, midwife, or doctor, compared to 8 in 10 urban women and 3 in 10 rural women (p < 0.01; Figure B.91). 72  |  Informing Durable Solutions for Internal Displacement   FIGURE B.88    Access to electricity to charge mobile   FIGURE B.89    Under 15 minutes to network recep- phone tion point 100 100 80 80 % of households % of households 60 60 40 40 20 20 0 0 U n n an P R an n-host N S ura res ost C settlem sid nt fli m nt t C or nt I P W limavio DP M n h ve e N an ea nt pr ea d D is rotrac d pl la ra d ed d ed Bo ulti ce tto ple p 0 N P 60 -p or r U n n an P R an n-host N S ura res ost C settlem sid nt fli m nt t C or nt I P W limavio DP M n h ve e N an ea nt pr ea d D is rotrac d pl la ra d ed d ed Bo ulti ce tto ple p 0 N P 60 -p or r on tle e en on tle e en oo oo a e c ot h de o de is p t te To 4 a e c ot h de o de is p t te To 4 a b ID ct e ID a b ID ct e ID - t e e - t e e on o on o om te len m on om te len m on ac ce ct ac ce ct rb o h rb o h on e l r id m on e l r id m U U rall U U rall ve ve rb r rb r P P O O D D Source: Authors’ calculation using Somali HFS 2017–18.   FIGURE B.90    Births in health facilities for IDPs, hosts,  FIGURE B.91    Births attended by skilled health staff and residents for IDPs, hosts, and residents 100 100 % of women who gave % of women who gave birth in last two years birth in last two years 80 80 60 60 40 40 20 20 0 0 C or nt DP W lim vio IDP N an ea nt ac c c d ed ed ted p 0 N P 60 -p or C or nt DP W lim vio IDP N an ea nt ac c c d ed ed ted p 0 N P 60 -p or U n n an P R an n-host N S ura res ost C -se tlem sident fli em nt nt M n h eve e pr ea ed D Dis Pr tra ed Bo ulti ce tto ple r U n n an P R an n-host N S ura res ost C -se tlem sident fli em nt nt M n h eve e pr ea ed D Dis Pr tra ed Bo ulti ce tto ple r oo oo pl la ra e Tom 4 pl la ra e Tom 4 a e nc a e nc a b D a b D on ttl e e on ttl e e on o on o m on m on is p ot ct is p ot ct ot h d o d ot h d o d rb o h rb o h rb r l I ct e I rb r l I ct e I on e l r id on e l r id om at le om at le U U ral U U ral t e t e ve ve O O At home Maternity clinic/MCH Hospital Relative/friend Traditional attendant Nurse/midwife/doctor Source: Authors’ calculation using Somali HFS 2017–18. 151.  Access to health care varies greatly across different types of IDPs. IDP women in settlements are half as likely (p < 0.01) to give birth at home compared to those outside settlements. Protracted IDPs, all of whom are in urban areas, also have better health care access than recent IDPs. Overall, the pattern of disparities across groups suggest that loca- tion is an important driver of disparities in access. 152.  IDPs have lower levels of literacy and schooling than urban residents and, like the rest of the population, there are stark sex differences between men and women in literacy. The literacy rate of IDP adults (52 percent) is lower than that of urban residents (73 percent, p < 0.01) (Figure B.92). School enrollment among ages 6–17 is also much lower among IDPs (35 percent) than urban residents (64 percent, p < 0.01) and hosts (62 percent, p < 0.01) (Figure B.93). Volume B: Country Case Studies  | 73 The sex gap in literacy is stark and consistent across groups: the share of adult men who can read and write, compared to women, is 22 percent higher among IDPs (p < 0.01), 18 percent higher among urban residents (18 percent, p < 0.01), and 20 percent higher among rural residents (p < 0.01) but there are no statistically significant sex gaps in school enroll- ment for primary (ages 6–13) or secondary school age (ages 14–17) children (Figure B.93).   FIGURE B.92    Adult literacy rate by sex, IDPs, and   FIGURE B.93    School enrollment among the school age residents 80 100 % of population 60 aged 6–17 80 aged 15 or more 40 % of population 60 20 40 0 U n n an P R an n-h st N S ral res ost C set lem sident fli m nt t C or nt I P W limavio DP M n h vene N n ea t pr ea d D is Protrac d pl la ra d ed d ed Bo ulti ce tto ple p 0 N P 0 -p or r 20 on tle e n oo a e c ot h de o de is p t te To 4 6 rb rb ID ct e ID rb o ho - tt e e on o om te len m on ac ce ct on e r id m U U rall ve 0 a O a u Overall Overall Overall Women Women Women Men Men Men D Overall IDP Urban resident Rural resident Source: Authors’ calculation using Somali HFS 2017–18. 153.  Adult literacy and schooling levels vary little when comparing IDPs and rural residents, and when com- paring different types of IDPs. There are no statistically significant differences across these groups, except that school enrollment among those ages 6–17 is somewhat lower among settlement IDPs (31 percent) than non-settlement IDPs (42 percent, p < 0.1), and among the bottom 40 percent of IDPs across the income distribution (28 percent) compared to the top 60 percent of IDPs (43 percent, p < 0.05). The overall similarities across different types of IDPs, however, sug- gest that the wider disparities in poverty across different types of IDPs are primarily because of their present circum- stances rather than educational endowments. Employment and Livelihoods 154.  IDPs participate in the labor force at similar rates to the urban and rural population. Almost 5 in 10 IDPs (48 percent) ages 15–64 are economically active, meaning that they have worked (45 percent) or are unemployed but have sought work (3 percent) in the last seven days. This is similar to the economically active share of the urban popu- lation (49 percent) and rural population (48 percent). Almost 2 in 5 inactive IDP working-age adults (1 in 10 of all IDPs, whether active or inactive) are enrolled in school. This is somewhat lower than the share of urban inactive adults who are in school (p < 0.05), but similar to rural enrollment. The remainder of IDPs in Somali regions, however (4 in 10 IDPs overall), are inactive and unenrolled: they are neither working, looking for work, nor in school. This is comparable to the share of the urban and rural residents in this category (Figure B.94). 74  |  Informing Durable Solutions for Internal Displacement   FIGURE B.94    Labor force participation for   FIGURE B.95    Changes in employment activity after dis- IDPs and residents placement 100 100 % of working-age population % of employed population 80 80 60 60 40 40 20 20 0 0 Men Women Overall Men Women Overall Men Women Overall P tI P C or v DP ev e t ot ead d ed D isp tra ted pl c cted ul e Bo ple To 4 0 60 -p or r en oo W ate enc h e m nc -s le l ID en D on o d ti I o c m p N P ed d o an ea Pr tra on lem nt l l tto ra lim io t e M nh ac e o ve et m pr is la a O om t ct t on e N IDP Urban Rural fli N S D C Active, employed Active, unemployed Inactive, enrolled Inactive, not enrolled Did not work before No Yes Source: Authors’ calculation using Somali HFS 2017–18. 155.  There are significant sex gaps between IDP men and women in labor force participation and employ- ment. IDP women are less likely to be employed than men (p < 0.01), and over half of IDP women are neither active nor enrolled in school, compared to less than a third (31 percent) of IDP men (p < 0.01). This sex gap in being neither active nor enrolled is smaller among IDPs (a 23 percent gap between men and women) than it is among urban and rural residents (a 33 percent gap for each). This may be because IDP women have greater access to schools than rural women (9 percent of IDP women ages 15–64 are enrolled, compared to 5 percent of rural women), or because male IDPs are much more likely to be neither active nor enrolled than urban men. It may also be because rural women lack alternate sources of income and are required to find work to support the home. About 3 in 10 male IDPs (31 percent) are neither working, looking for work, nor enrolled in school, compared to only 2 in 10 urban men (20 percent; p < 0.05) (Figure B.94). Male IDPs are almost twice as likely as IDP women to work as salaried labor (51 percent of male IDPs vs. 31 percent of IDP women, p < 0.01). IDP women are more likely than men to work on their own business (27 percent for women IDPs vs. 18 percent for male IDPs, p < 0.1), and to be working as unpaid helpers in family businesses (31 percent for women IDPs vs. 18 percent for male IDPs, p < 0.01) (Figure B.98) 156.  Women are much more likely than men to be economically inactive because they are caring for their families or households. Unpaid care work is not counted in labor force participation statistics as being economically ‘active’. Most Somalis (about 7 in 10 of IDPs and non-IDPs alike) believe that most or all women in their communities are allowed to work outside the home, despite a significant minority reporting that only some or almost none are. (Fig- ure B.96). Yet even if social norms permit, women are much more likely than men to be unable to work or be enrolled in school because of family and household care responsibilities: among IDPs, 59 percent of women and only 24 percent of men are economically inactive because of family and household care responsibilities (p < 0.01). Rural and urban women are also more likely than men to be economically inactive because of family and household care (rural women: 64 percent, rural men: 24 percent, p < 0.01; urban women: 69 percent, urban men: 38 percent, p < 0.01). IDP men, in contrast, are much more likely than IDP women to not be working because of illness or disability (p < 0.01), the reason cited by 30 percent of IDP men for economic inactivity (Figure B.97). Volume B: Country Case Studies  | 75   FIGURE B.96    Proportion of women perceived to be al-  FIGURE B.97    Reasons for economic inactivity lowed to work outside the home 100 not-enrolled population 100 80 % of inactive and 80 60 % of households 60 40 40 20 20 0 Men Women Overall Men Women Overall Men Women Overall 0 rb rb d e P U no n h t R an -host al s t C on e de t fli ett em t C r v ent W lim io le t ce a t pr ad d P ot ed ac ce cted ed d d Bo ltip e tto le p 0 N P 0 -p o r r an a n u r re s N S esi en on -s ttl n o mn an he n oo ot he de is p tra te u c Tom 4 6 U U esi ID n o ct le e Man eve on o om ate n m on r id D Dis ro rac l r a ll na er IDP Urban Rural io v p l la at O rb N N Family and household care In school Almost none Some Majority Almost all Too young/old Waiting for busy season/on leave Ill/disabled Source: Authors’ calculation using Somali HFS 2017–18. 157.  The employment patterns of IDPs and host communities, and of IDPs in and out of settlements, differ. Most employed IDPs work as salaried labor or labor paid in kind, including in agriculture (43 percent), or in non-farm businesses that they (22 percent) or their households (23 percent) own. These patterns differ from those of host com- munities, who are almost twice as likely to work in their own businesses (40 percent for hosts vs. 22 percent for IDPs, p < 0.01), but are less likely to be helping in their families’ businesses (14 percent vs. 23 percent for IDPs, p < 0.1) (Fig- ure B.98). Settlement and non-settlement IDPs also have different employment patterns, which may be because settle- ment IDPs are only in urban areas. Settlement IDPs are more likely than non-settlement IDPs to work as salaried labor (p < 0.05) and are less likely to farm, hunt, or fish for themselves or help on family farms (p < 0.01). This difference may partly be because all settlement IDPs (100 percent) are in urban areas, compared to only 35 percent of non-settlement IDPs (Figure B.98).  FIGURE B.98    Main employment activity for IDPs, hosts, and rural residents 100 % of emplpoyed population 80 60 40 20 0 Overall Overall Overall Overall Overall Overall Overall Women Women Women Women Women Women Women Men Men Men Men Men Men Men Overall IDP Urban host Rural resident Camp IDP Non-camp IDP Conflict IDP Climate IDP Salaried labor Own business Help in business Own account agriculture Apprenticeship Source: Authors’ calculation using Somali HFS 2017–18. 76  |  Informing Durable Solutions for Internal Displacement 158.  Most IDPs do the same work they did before being displaced, but about half of the poorest and those outside settlements have had to change their main employment. Almost 7 in 10 employed IDPs (67 percent) report the same main employment activities as before being displaced. These figures are even higher for IDPs in pro- tracted displacement (82 percent), who may have had more time than others to re-establish their livelihoods, and set- tlement IDPs (84 percent). However, non-settlement IDPs and the poorest 40 percent of IDPs (who, because 69 percent of all IDPs are under the poverty line, are a subset of the poor), are more likely to have changed their employment. Every second IDP (49 percent) living outside a settlement has had to change his or her main employment activity since being displaced, as have over 4 in 10 (44 percent) of the poorest 40 percent of IDPs (Figure B.95). 159.  Most IDPs rely on salaried labor, small family businesses, or aid/zakat to provide their main source of income, and are more likely than most others to rely on small family businesses. When examining household income (what households live on) rather than employment (what they do), a more nuanced picture of IDP livelihoods emerges, which captures how aid, zakat, remittances, trade, property, and other income sources contribute to house- hold livelihoods. In Somali regions, about two in five IDPs have salaried labor as their main income source, and one in five rely on small family businesses, but almost none rely on trade or property income. IDPs are more likely to rely on small family businesses than urban residents or host communities (IDPs: 19 percent; urban residents: 12 percent, p < 0.05; host communities: 12 percent, p < 0.05). IDPs are also less likely than before to make a living from agriculture (15 percent before being displaced vs. 7 percent after being displaced, p < 0.01) (Figure B.99).   FIGURE B.99    Main source of household income for IDPs, hosts, and residents 100 80 % of households 60 40 20 0 Urban host Urban non-host Urban resident Rural resident Before Current Before Current Before Current Before Current Before Current Non displaced (current) Overall IDP Camp IDP Non-camp IDP Conflict IDP Climate IDP Salaried labor Remittances Small family business Agriculture Trade, property income Aid or zakat Other Source: Authors’ calculation using Somali HFS 2017–18. 160.  Few IDPs rely on remittances, aid, or zakat, and although climate IDPs are poorer, they are less likely than conflict IDPs to rely on aid or zakat. Less than 1 in 13 IDPs (7 percent) rely on remittances as their main source of income, and only 1 in 10 IDPs overall (12 percent) rely on aid or zakat. IDP aid dependency is much lower than in other countries in the region, such as South Sudan, where over 7 in 10 IDPs, and 9 in 10 refugees, rely on humanitarian Volume B: Country Case Studies  | 77 assistance.87 The difference in receiving remittances between IDPs and host communities is robust after considering other household characteristics (Table B.7). Climate IDPs in Somali regions are also much less likely to rely on aid or zakat than conflict IDPs (Figure B.99). This is even though they are much poorer (Figure B.80; Figure B.81), and even though drought, famine, or flood have disrupted their agricultural livelihoods: as might be expected, far fewer climate IDPs now rely on agriculture, fishing, hunting, and animal husbandry (28 percent before vs. 13 percent now, p < 0.01) as before being displaced, shifting to salaried labor (25 percent before vs. 41 percent currently, p < 0.01) to earn an income. Only a tiny minority of climate IDPs (3 percent) rely on aid or zakat, which is much lower than the share of conflict-displaced IDPs who do so (23 percent, p < 0.01) (Figure B.99). This suggests not that climate IDPs need less assistance, but that they may get less. This may again reflect the recent rapid increase in rates of displacement due to the most recent drought, the growth of which has outpaced the ability of humanitarian actors to expand their assistance to fully meet the scale of demand. 161.  Average remittance amounts vary considerably across different types of IDPs, with settlement, pro- tracted, women-headed, and the bottom 40 percent of IDP households receiving low amounts. The average annual value of remittances for all IDP households, whether they receive remittances or not, is US$27 per capita, which is about half of what urban residents get on average (US$56, p < 0.1). There are considerable disparities in how much different types of IDPs get on average. Settlement IDPs receive about an eighth of what non-settlement IDPs get, receiving only US$7 on average per capita per year in remittances, compared to US$59 for IDPs outside settlements (p < 0.1). Protracted IDPs get less than a dollar on average per year, compared to US$37 for recent IDPs (p < 0.05). The bottom 40 percent of IDPs get US$6 on average, compared to US$39 for the top 60, and women-headed households get far less than men-headed households, getting US$7 on average, compared to US$45 for male-headed households (p < 0.1). These findings are consistent with earlier surveys and likely reflect the extent to which such households are marginalized and disconnected from social networks that would otherwise provide such support. This may be particu- larly true of minority clans that are disconnected from social networks and may have no mechanisms of support other than formal settlements (Figure B.100).   FIGURE B.100    Average remittances for IDPs, hosts, and residents 120 amount (USD) 90 Remitance 60 30 0 Overall IDP Urban host Urban non-host Urban resident Rural resident Non-settlement IDP Settlement IDP Woman headed Man headed Not protracted Protracted Bottom 40 Top 60 Overall IDP Source: Authors’ calculation using Somali HFS 2017–18. 87. See South Sudan and Ethiopia case studies. The comparison, however, should be interpreted with some caution as the South Sudan survey that was conducted in Protection of Civilian (PoC) camps only. 78  |  Informing Durable Solutions for Internal Displacement Social Cohesion and Security Perceptions 162.  Most IDPs feel safe where they are and report good relations with the communities around them. Almost 8 in 10 IDPs (78 percent) feel safe (moderately or very) where they are, which is similar among the national population, but somewhat less than among host community members (92 percent, p < 0.05). IDPs displaced by conflict are less likely to feel safe, as are IDPs displaced multiple times: 3 in 10 conflict-displaced IDPs (31 percent) and almost 4 in 10 IDPs displaced multiple times (37 percent) feel very unsafe, moderately unsafe, or neither safe nor unsafe, compared to 15 percent of climate-displaced IDPs (p < 0.1) and 19 percent of IDPs displaced only once (p < 0.05) (Figure B.101). This overall perception of safety among the IDP population at large is in line with perceptions of host community rela- tions. Almost 9 in 10 IDPs (87 percent) think that their relations with the communities around them are good or very good. This likelihood is consistent across different types of IDPs, whether in or out of settlements, in households headed by men or women, in protracted or non-protracted displacement, displaced once or multiple times, or are rich or poor (Figure B.102).   FIGURE B.101    Perceptions of safety   FIGURE B.102    Perceived relations with surrounding community 100 80 % of households 60 100 % of households 40 80 20 60 0 40 oor rb rb id P U an an ent R an n-host on e l r id t C -se tlem esident fli lem nt nt C or nt P W lim vio IDP M n h ve e N an head nt pr ad d D Dis Protra d pl la ra ed ed d ed Bo ult ce tto iple p 0 60 r 20 N S ura res os -poo a ee c ot e e o e To 4 U U res ID D on tt e e om at en m on is p t ct ac ce ct rb no h ct e I m on P l ll l na ra 0 io ve N at O C ettl me P ct ent DP C v DP an eve e M he nt ot ea d r d D isp tr ed ac e ted u e tto le To 40 60 -p or r t oo om e c N n h de pr de m nc -s tle ll ID Bo ltip on o W at olen D Pro act I I is la ac m p N P ed d o N a a on em nt a ot N S ver lim i pl c o r O on et Very unsafe Moderately unsafe fli Neither safe nor unsafe Moderately safe Very safe Very bad Bad Neither good nor bad Good Very good Source: Authors’ calculation using Somali HFS 2017–18. Targeting Analysis 163.  Most IDPs are productive but poor, and they concentrate in Banadir, middle Shabelle, Gedo, Woqooyi Galbeed, and Bay. In Somalia, only a small share of households are classified as support-dependent (less than 1 per- cent). Over 73 percent of IDPs are productive but poor, and 26 percent are self-reliant. Host communities have a larger share of self-reliant households compared to IDPs. The vulnerability status of households varies markedly by (pre-war) region. Almost all IDPs in lower Juba are self-reliant, whereas most households in Banadir, middle Shabelle, Gedo, Woqooyi Galbeed, and Bay are productive but poor. This indicates a need for region-wise targeting to improve IDPs’ gaps to a durable solution (Figure B.103; Figure B.104). Volume B: Country Case Studies  | 79   FIGURE B.103    Vulnerable population by status of   FIGURE B.104    Vulnerable IDP population by pre-war the household region 100 100 80 80 % of households % of households 60 60 40 40 20 20 0 0 Host community IDPs a ug ri er r le o ed y di Ba Ba b ed el he Ju ud na be ab G gd M Ba al er Sh G Self-reliant To w Lo yi e dl oo Productive but poor id oq M W Support-dependent Source: Authors’ calculation using the Somali HFS 2017–18 Self-reliant Productive but poor Support-dependent Source: Authors’ calculation using the Somali HFS 2017–18 Typology of IDPs 164.  Clustering analysis for Somalia yields two groups. The groups differ in their displacement trajectories, partic- ularly from the cause-based and needs-based lens (Figure B.105). Group 1, which accounts for about 40 percent of the sample (Table E10 in the appendix) had more agricultural livelihoods at the origin, was more likely to be drought displaced, and had poorer living conditions before displacement and currently. Group 2, while less agricultural and with better hous- ing at the origin, was more likely to be displaced by armed conflict. The households of Group 2 are less poor, less hungry, and in better housing than those in Group 1. The differences among the groups, while still present, are lesser in the solu- tion-based lens. Most members of both groups prefer to stay in the current location rather than return or relocate, though Group 2 is more likely to be guided by security and Group 1 by basic amenities and livelihoods in addition to security.   FIGURE B.105    Visualization of the two clusters from the clustering analysis Source: Authors’ calculation using Somali HFS 2017–18. Note: Group 1 is represented by the blue circles and Group 2 by the black triangles. 80  |  Informing Durable Solutions for Internal Displacement Cause Profile 165.  Conflict and climate triggered the displacement of both groups, but in differing degrees. More than half the households in Group 1 were displaced by climate events, including drought, famine, or flood (Figure B.106). The drought of 2016–2017 triggered massive displacements, bringing the number of IDPs to nearly 2 million in Somalia.88 Households that were not displaced from climate events were motivated to move due to conflict. About 2 in 10 house- holds of Group 1 had to leave due to armed conflict in their own village, while another 1 in 10 households left due to armed conflict in nearby villages. Group 2, in contrast, was twice as likely to be displaced by armed conflict in the village (4 in 10 households). Armed conflict in the household’s own village indicates a more direct exposure to the conflict. Another 3 in 10 households were displaced by climate events. While Group 1 was more likely to be displaced by climate and Group 2 by armed conflict, the overlapping of climate and conflict events makes it difficult to separate their effects, especially as Somali regions have a long-standing history of both conflict and cyclical droughts and climate shocks.   FIGURE B.106    Reasons for displacement from origin 80 60 % of households 40 20 0 Armed conflict Armed conflict Increased Discrimination Climate event Other in village other village violence Group 1 Group 2 Source: Authors’ calculation using Somali HFS 2017–18. 166.  Though motivations for displacement varied, security was the key reason that both groups chose the current location. In both groups, security was the most commonly cited reason for arriving at the current location. In Group 1, more than 4 in 10 households arrived at the current location for better security (Figure B.107). Another 2 in 10 did so due to home, land, or livelihood opportunities, and another 2 in 10 settled in the current area due to health, education, and aid. In Group 2, 8 in 10 households settled at the current location for better security. The near-unani- mous choice of security is likely motivated by the armed conflict, often in their own village, that many households in Group 2 faced. 88. Estimates vary across humanitarian agencies working in Somali regions. UNCHR’s Protection and Return Monitoring Network estimates 1.88 million IDPs in the country (https://unhcr.github.io/dataviz-somalia-prmn/index.html) while IOM’s Displacement Tracking Matrix estimates roughly 2 million IDPs (http://www.globaldtm.info/somalia/). The PRMN is the UN’s latest data source on displacement, estimating 1.8 million, but the OCHA humanitarian needs overview reports that 2.1 million people are displaced. Volume B: Country Case Studies  | 81   FIGURE B.107    Reasons for coming to current location 100 80 % of households 60 40 20 0 Security Water Home, land, Health, Family livestock livelhood education, aid Group 1 Group 2 Source: Authors’ calculation using Somali HFS 2017–18. 167.  The group that was more exposed to conflict had closer access to basic amenities and better housing at the origin. Nearly 9 in 10 households in Group 2 were on average within a 30 minute-walk of amenities such as pri- mary school, health facility, water point, and market (Figure B.108), as opposed to about 6 in 10 households in Group 1. Group 2 was also more likely to live in improved housing, defined as living in a dwelling that is made of block, wood, or concrete, and intended for habitation. About 35 percent of the households in Group 2 and 14 percent of Group 1 had improved housing at the origin. However, Group 1 was more likely to have access to agricultural land (which was not necessarily owned). More than 3 in 10 households of Group 1 had such access compared to less than 1 in 10 in Group 2. The access to agricultural land indicates an agricultural livelihood rather than wealth. Group 2 was better off in terms of living conditions and less agricultural.  FIGURE B.108    Closeness to basic facilities,   FIGURE B.109    Primary source of household income at origin agricultural land access and housing at origin Other 100 Aid or zakat 80 % of households Trade, property 60 Own account agriculture 40 Family business 20 Remittances 0 Near Land Improved Salaried labor facilities access house at at origin at origin origin 0 20 40 60 80 % of households Group 1 Group 2 Group 1 Group 2 Source: Authors’ calculation using Somali HFS 2017–18. 82  |  Informing Durable Solutions for Internal Displacement 168.  Group 1, which was more likely to be drought displaced and relied more heavily on agriculture for income. Most of Group 2 relied on salaried labor (54 percent) and nonagricultural family businesses (30 percent) as the primary source of household income (Figure B.109). The households in Group 1 were more heterogeneous. More than 3 in 10 households relied on own-account agriculture, 2 in 10 on salaried labor, 1 in 10 on remittances, and another 3 in 10 on miscellaneous other livelihoods such as savings, pensions, or asset sales. The households of Group 1 were 15 times more likely to be involved in agriculture, and 6 times less likely to be involved in a nonagricultural business. This establishes the more agricultural and possibly rural leanings of Group 1, which had more access to agricultural land and lower living standards. 169.  The differing socioeconomic patterns of the groups can be linked to their drivers of displacement. Group 1 had more access to agricultural land and was more likely to have an agricultural income source. This can explain why it was more exposed to the adverse shocks of drought, triggering the climate-related displacements that are more prev- alent in Group 1. The households of Group 2 relied very little on agricultural livelihoods, had better housing, and were located closer to basic amenities. This can signal more wealth relative to Group 2, or more urban locations. Both higher wealth and lower dependence on agriculture could lead to a lower impact of the drought on these households. It is also possible that their relative wealth made these households likelier targets in plundering or lootings in the conflict, which could explain the higher conflict-related displacement and concern with security in Group 2. Needs Profile 170.  Differences in the cause-based conditions of the groups have translated into different needs in their current situations. Group 1, which was more likely to be driven by climate events, has higher household sizes and dependency ratios, driven by children (Table B.8). Households in Group 1 have six members on average, compared to a little more than five in Group 2. The age dependency ratio is also higher in Group 2, at 1.7, meaning that for each working-age member there are nearly two non-working–age dependents. Larger household sizes and more fertility can be associated with rural or agricultural lifestyles. Roughly half the households are headed by a woman in both groups (no significant difference).   TABLE B.8    Household characteristics   Group 1 Group 2 % of women headed households 51 46 Household size 5.9 5.3 Age dependency ratio 1.7 1.2 Source: Authors’ calculation using Somali HFS 2017–18. 171.  In keeping with worse living conditions at the origin, Group 1 faces greater poverty today. While poverty is high for the IDPs as a whole, it is more widespread in Group 1. More than 8 in 10 individuals in Group 1 are living in poverty, as opposed to 1 in 7 in Group 2 (Figure B.110). Households in Group 1 also do worse on a range of living con- ditions. About 67 percent of Group 1 faced hunger in the last four weeks, as opposed to 47 percent of Group 2. Only 13 percent live in improved dwellings today, while 34 percent of Group 2 have improved housing. About 36 percent have improved sanitation facilities while 58 percent of Group 1 use such facilities. However, 16 percent of Group 1 have access to agricultural land, as opposed to nearly none of the households in Group 2. As at the origin, the current access to agricultural land appears to signal an agricultural livelihood rather than wealth or a better standard of living. Volume B: Country Case Studies  | 83   FIGURE B.110    Poverty and current living conditions   FIGURE B.111    Primary source of household income, current 100 Other 80 % of households Aid or zakat 60 Trade, property 40 Own account agriculture Family business 20 Remittances 0 Poverty Hungry Improved Improved Land Salaried labor headcount house sanitation access ratio 0 20 40 60 80 % of households Group 1 Group 2 Group 1 Group 2 Source: Authors’ calculation using Somali HFS 2017–18. 172.  The two groups have modified their income sources somewhat, with Group 1 shifting away from agri- culture. Most households in Group 2 rely on salaried labor (49 percent), followed by a nonagricultural family business (25 percent) and aid or zakat (17 percent; Figure B.111). The livelihood structure of Group 1 is, as before displacement, more heterogeneous. Households have shifted away from own-account agriculture, with only about 15 percent of Group 1 relying on this source for income (compared to more than 30 percent before). About 25 percent of the house- holds rely on salaried labor, while 16 percent rely on remittances and 9 percent on a non-farm family business. More than 27 percent of Group 1 also rely on other miscellaneous sources of income such as pensions, savings, and asset sales, among others. 173.  Group 1 has shifted more into remittances, while Group 2 has shifted into aid or zakat. In both groups, about 2 in 10 households have sources of household income that are from external sources: remittances and aid or zakat (Figure B.111). Interestingly, Group 2 has shifted more into aid or zakat while Group 1 has shifted more into remittances. Signs of this selection were present in the livelihood structures at the origin, where about 5 percent of the households in Group 2 relied on aid or zakat, and 11 percent in Group 1 relied on remittances (Figure B.111). Over the course of displacement, the households possibly switched into external income sources which were more accessible to them. Group 2 might be better connected to networks for external aid, which can explain why its households are more likely to rely on this source despite being less poor than Group 1. Solutions Profile 174.  Most IDPs across both groups prefer to stay in their current locations. About 63 percent of households in Group 1 and 73 percent in Group 2 want to stay in their current locations rather than return to their origin or relocate (Figure B.112). Households of Group 2, which were more likely to be displaced by armed conflict, are more likely to pre- fer staying. Of the households that plan to return or relocate, those in Group 1 envision a longer timeline (12 months or more) or are uncertain when to relocate. 84  |  Informing Durable Solutions for Internal Displacement   FIGURE B.112    Intention and timeline to move 100 80 % of households 60 40 20 0 Stay 6 6–12 ≥12 Don't months months months know when Source: Authors’ calculation using Somali HFS 2017–18. 175.  Group 2 is more likely to seek security to settle anew, while Group 1 seeks security along with services and livelihoods considerations. Group 2 is more likely to cite security as the key factor in a decision to move (Fig- ure B.113). This can explain why most households (7 in 10, Figure B.112) plan to stay. Group 2 is also less likely to seek more information in deciding whether to move than Group 1. About 50 percent of the households in Group 1 seek more information to inform a moving decision, with 25 percent seeking information on the security and political situation, and another 25 percent seeking information on a host of other non-security–related factors such as basic amenities, livelihood opportunities, and travel options. Group 1 is also more likely to cite non-security–related inputs (compared to Group 2) in helping them to settle anew, while 85 percent of the households in Group 2 cite better secu- rity as the key factor they require to settle.   FIGURE B.113    Factors guiding a decision to settle in the future 100 80 % of households 60 40 20 0 Security Non-security Have Security and Other Non-security Security all info political info info inputs inputs Factors in moving decision Information required to decide Help needed to settle Group 1 Group 2 Source: Authors’ calculation using Somali HFS 2017–18. Volume B: Country Case Studies  | 85 Policy Implications of IDP Typology 176.  Uncertainty around security dominates return intentions for both groups, outweighing their potentially different tendencies. Both Group 1 and Group 2 cite security as a major reason for their return intention: whether to stay, return, or resettle. Security is also the key support that both groups seek to settle anew. This indicates the nature of the displacement situation, which is still entrenched in uncertainty regarding security, and which could possibly moti- vate large numbers of both groups to stay in the current location rather than move. It is possible that in a post-conflict stage, differences among the two groups might become more pronounced. Group 1 had more agricultural livelihoods and continues to have more access to agricultural land today, which could create scope for a move (or a return) to agri- cultural locations. Group 2, in contrast, depended more on salaries and businesses, which might allow for integration in an urban setting, whether at the origin or in a new location. 177.  Home and livelihood restoration efforts would look different for the two groups. Group 1 had more agri- cultural livelihoods at the origin, and was more likely to have access to agricultural land both before displacement and currently. Agricultural land access can be key in restoring the own-account agricultural livelihoods of Group 1. Group 2 had better housing and was wealthier than Group 1 before displacement. Also affected more by armed conflict, Group 2 may have lost more housing and assets, the restoration of which could be a possible channel to restore their livelihoods. 178.  Resilience to drought would be key in a durable solution, especially for Group 1. Group 1’s dependence on agriculture, displacement due to climate events, lower living conditions, and possibly lower income, indicate a vul- nerability to climate shocks. Given Somalia’s long-standing history of climate shocks, including droughts and flooding, approaches of drought-resilient rehabilitation of livelihoods will be key in achieving a stable durable solution. Conclusions Informing Durable Solutions 179.  The estimated 1.8 to 2 million IDPs in Somalia regions are mostly young, poor, and out of work; often go hungry; have poor housing, water, health, and schooling; and are increasingly concentrated in urban areas. Half of IDPs are under the age of 15, half experience hunger, and three in four live on less than the international poverty line of US$1.90 PPP (2011) per day per capita, consuming on average about 35 percent less than US$1.90 per day. About three in four are in already strained urban areas. About a third have had to change their livelihoods, many shifting out of agriculture; and four in ten are neither working, looking for work, nor in school. IDPs have poor housing and access to sanitation, and are farther away than others from schools, health centers, and markets. They receive low levels of remittances and have few safety nets. They also have less access to health care and schooling, which, combined with hunger, can translate into persistent, lifelong gaps in well-being. 180.  Advancing durable solutions for displacement-affected populations in Somali regions is thus a central challenge for longer term stability and development. Displacement is widespread, its deprivations many and deep. Development and poverty alleviation strategies for Somali regions will not be achieved without addressing displacement-related vulnerability and ensuring that displaced populations are integrated into society, the econ- omy, and development policy and planning.  86  |  Informing Durable Solutions for Internal Displacement 181.  IDPs should be able to choose freely whether to return, stay, or settle elsewhere. International standards highlight that durable solutions for displaced populations may entail returning sustainably to places of origin, locally integrating in current communities, or settling in another part of the country; particularly important is the right of displaced populations to choose freely between these options.89 More specifically, in the context of Somali regions, advancing durable solutions for displacement-affected populations—including IDPs, returning Somali refugees, and host communities—should further reflect the following: • Support for return to communities of origin in areas where conflict- and climate-related events have abated and where voluntary, safe, and dignified return is feasible; • Support for local integration for those unwilling to return to areas affected by continuing conflict- or climate-related events, or other factors; and • Support as feasible for those currently displaced in areas of continuing conflict and/or humanitarian emergency or for those interested in return even in the context of ongoing instability. 182.  Providing durable solutions in Somali regions requires a broad-based approach led by the govern- ment. This entails a combination of area-based, cross-sectoral, multi-stakeholder needs and rights-based policies and investments in which humanitarian and development partners engage collaboratively under government leadership. Enabling government ownership and leadership across any policies and investments is a priority. Interventions should align with the development priorities for durable solutions outlined in the National Development Plan, as well as other government-led efforts, including the Recovery and Resilience Framework in development to respond to the most recent drought. Efforts should further build on other ongoing initiatives, including the Durable Solutions Initiative and regional initiatives such as the CRRF, the Nairobi Declaration on Durable Solutions for Somali Refugees and Reintegra- tion of Returnees in Somalia, and ongoing engagement by IGAD’s Regional Secretariat on Mixed Migration and Forced Displacement. 183.  Policy and program recommendations include the following: (a) Continue to provide humanitarian assistance to address basic needs and support resilience. With more than half of IDPs reporting hunger, continuing life-saving activities to support basic needs remains critical. Expand- ing access to basic services, including health and education, is also important in enabling communities to become more resilient. (b) Strengthen the viability of urban and peri-urban areas, and enable IDPs to better integrate into them. About 70 percent of IDPs express a desire to stay in their current locations, which are mostly in urban areas. This is consistent with other studies, which indicate that even when climate-related conditions in communities of origin improve, IDPs may feel too unsafe to return.90 Given that IDPs are concentrated in urban centers and secondary towns and that rapid urbanization is having an impact on existing development deficits, vulnerability, and mar- ginalization in Somali cities, strengthening the viability and resilience of Somalia’s urban and peri-urban areas to enable IDPs to integrate into the local economy and become more self-reliant is critical. This will entail investing in 89. Council, U.E.a.S. 1998. “Guiding Principles on Internal Displacement”; IASC. 2010. “IASC Framework on Durable Solutions for Internally Displaced Persons.” 90. UNHCR. 2016. “Internal Displacement Profiling in Mogadishu.” Volume B: Country Case Studies  | 87 services and infrastructure (including housing, shelter, water and sanitation, health, and education) to help cities better absorb massive population growth and provide services for displacement-affected populations and host communities alike. There is also a need to empower municipal authorities to plan, monitor, and budget for city growth. At the same time, the cities are already sites of innovation, with extensive private sector delivery mecha- nisms for services, financial investment, and job creation, which can be further harnessed. (c) Support rural resilience and recovery to enable safe and voluntary return and reintegration. Although IDPs have mainly moved from rural to urban areas, investing in rural solutions to support return and recovery, and to provide opportunities in rural areas, should also be pursued. The survey findings highlight that socioeconomic and human development indicators of IDPs are often comparable or even better than those of rural residents, high- lighting the vulnerabilities and development deficits confronting rural populations in Somali regions. Improving access to basic services and investing in socioeconomic infrastructure will be critical in supporting IDPs who wish to return. This will likely require start-up assistance and support to restore livelihoods. Consideration for invest- ments may include cash transfers for basic consumption, skills development, and other forms of livelihood support, including inputs for agricultural production or restocking of livestock for pastoralist activities. Interventions may also include consideration for developing systems to enable recovery of lost assets and land, or repair/restoration of housing. (d) Promote livelihood and employment opportunities. Employment and labor force participation among IDPs are low. Enabling access to livelihoods, employment, and opportunities to earn an income is critical both for house- hold stability and resilience, as well as for local economic development and growth. In urban settings, this may include expanding salaried labor opportunities, for example, through public work schemes or other infrastruc- ture investment activities. Development investments targeting male or youth employment should investigate integrated approaches that combine business skills development, vocational training, or cash transfers with cog- nitive and non-cognitive skills building. These have demonstrated effectiveness in other high-risk contexts and may be appropriate for addressing psychosocial challenges—for example, trauma, depression, dislocation—that may impede participation in employment opportunities.91 Employment and livelihood initiatives should include gender-responsive approaches to address barriers to employment opportunities, such as disadvantageous social norms and domestic labor burdens on women. Policies and interventions to enable women to engage in eco- nomic opportunities should consider key protection provisions to minimize potential exposure to harm, harass- ment, or forms of GBV. (e) Support policy and planning solutions for improved access to land, housing, and shelter. Insecurity of land tenure constitutes a significant challenge in Somalia, which has had a major influence on the success of housing and resettlement provisions in Somalia to date. Other studies and humanitarian reports indicate that forced evic- tions due to land tenure insecurity are a common feature of urban life and perpetuate cycles of displacement. As this survey further highlights, lack of access to improved housing for three-quarters of IDPs also constitutes a major barrier to development and resilience across multiple dimensions. These findings underscore the need for laws, frameworks, and policies to assure both secure property rights and to identify housing planning and policy solutions for IDPs and host communities alike. Addressing land tenure and housing disputes may further require establishment and mediation through local-level dispute resolution mechanisms. 91. Rift Valley Institute. 2013. “The Impact of War on Somali Men”; World Bank. 2015. “The Sustainable Transformation of Youth in Liberia (STYL) Program.” 88  |  Informing Durable Solutions for Internal Displacement (f ) Promote protection and social cohesion. While survey findings indicate general positive feelings of safety and cohesion by displaced populations, humanitarian and development programming should consider interventions to strengthen social cohesion and protection considerations to minimize potential grievance and monitor or address tensions between displaced and host communities in both urban and rural environments. IDPs in South Sudan Introduction and Country Context 184.  The Republic of South Sudan is the youngest nation in Africa. The country gained independence from (Northern) Sudan in 2011 after a years-long secessionist war. The war was divided in two phases. The first began in 1955 and lasted until 1972. The second confrontation between the North and the South began in 1983 and ended with a peace agreement in 2005. This peace agreement set the stage for a referendum that ended with the independence of South Sudan in 2011. The war for independence was largely rooted in Sudan’s colonial past and patterns of marginalization of the Southern societies. Power concentration among Northern Sudanese elites and the social and cultural hegemony of Muslim groups had alienated the Southern Sudanese elites. This lead to dis- content and eventually to war.92 After the Addis Ababa agreement of 1972 which ended the first war, the violation of Southern semi-autonomy and a struggle over the control of the newly found oil reserves in the late 1970s and early 1980s led to the renewal of the guerilla movements and the emergence of the Southern Peoples Liberation Army/Movement (SPLA/M) that led the armed struggle up to 2005.93 The region that is now South Sudan suffered from displacement during the many years of conflict with Sudan. Millions of people returned after the South- ern Peoples Liberation Movement (SPLM) and the Government of Sudan (GoS) signed the Comprehensive Peace Agreement (CPA) in 2005. 185.  In late 2013, a power struggle within the ruling party, SPLM, resulted in the outbreak of conflict. President Salva Kiir, an ethnic Dinka, fired Vice-President Riek Machar, an ethnic Nuer, suspecting a plot to overthrow the govern- ment. Machar denied the claim and formed a group called the Southern Peoples Liberation Movement—in opposition (SPLM-IO). Kiir and Machar had also been rivals during the long war with Sudan, which led to the independence of South Sudan in 2011. Fighting broke out in Juba between forces loyal to the two men, igniting a civil war that quickly spread across the country. Areas initially affected by violence were Greater Upper Nile (Unity, Upper Nile, and Jonglei) as well as some areas of Greater Bahr el Ghazal. Eventually the war also reached Greater Equatoria, spreading local violence by allies of the SPLM-IO.94 186.  The conflict map in South Sudan is complex and multilayered, and the pace of nation building has left the government overwhelmed and created precarious institutions. The long exposure to violent conflict of South Sudan’s societies, as opposed to its fairly short history within their own state, make Southern Sudan prone to complex national conflicts. The root causes for conflict are diverse. Competing political visions about the future of Southern 92. Poggo. 2008. “First Sudanese Civil War”; Sharkey. 2008. “Arab Identity and Ideology in Sudan”; Ylönen. 2009. “On Sources of Political Violence in Africa.” 93. Johnson. 2003. “The Root Causes of Sudan’s Civil Wars.” 94. Justin and De Vries. 2017. “Governing Unclear Lines.” Volume B: Country Case Studies  | 89 Sudan after the death of John Garang, the iconic leader of the SPLA/M, who died after the 2005 CPA, have created new political rivals within the ruling party. The government’s challenges in enforcing inclusive policies and tackling social and economic problems immediately after independence have led to discontent and unrest. On the local level, long-lasting rivalries that have been shaped by decades of intra-Southern competition have not been overcome.95 Further complications have been created in the wake of the 2015 administrative reorganization of the national territory by presidential decree. The 10 states that existed since 1992 were reorganized into 28 new administrative regions. This resulted in several unclear boundary agreements that led to conflict on various governmental and local levels. In addi- tion, these new boundary agreements further exacerbated resource competition between pastoralists and agricultur- alists, and between autochthonous and migrant groups in some areas.96 187.  A peace deal between the government, opposition, and other parties was signed in September 2018, but continued outbreaks of violence render the peace precarious. The Revitalized Agreement on the Resolution of Conflict in South Sudan (R-ARCSS) was signed on September 12, 2018, by President Salva Kiir, First Vice President Riek Machar, and other parties. While the R-ARCSS officially ended the conflict, outbreaks of violence have continued. Previous peace deals were also followed by conflict—for instance, a peace agreement between the government and opposition had been signed in August 2015, but opposition splintered, and fresh violence broke out in July 2016. Thus, while positive steps have been taken in addressing tension between the opposing parties, a long way remains before a stable, consolidated peace is achieved. 188.  The recent conflict displaced 4 million people—one in three South Sudanese, though returnee move- ment has accelerated since the signing of the R-ARCSS in 2018. During the conflict, 2.1 million became refugees in neighboring countries, including Uganda (1 million), Sudan (453,000), Ethiopia (419,000), Kenya (111,000), and DRC (87,000). By 2017, the number of IDPs swelled to 1.9 million, with most of them concentrated in the Greater Upper Nile Region—former Unity (539,000), Jonglei (365,000), and Upper Nile (220,000) states. As the war moved southward, the Equatoria Region (413,000) that includes the former states of Western, Central, and Eastern Equatoria received many IDPs.97 As of March 2019, since the signing of the peace deal, there are 1.2 million IDPs in South Sudan and more than 800,000 returnees.98 189.  The majority of IDPs in South Sudan are not in camps. As of 2018, only 15 percent of IDPs lived in camps or camp-like settings where they could easily access humanitarian assistance. This included 210,000 IDPs in PoC sites, 58,000 in collective centers, and 28,000 in informal settlements.99 PoC sites are IDP camps inside United Nations Mission in South Sudan (UNMISS) peacekeeping bases. There are, at the time of writing, six PoC sites in Bentiu, Bor, Juba, Malakal, and Wau.100 95. De Vries and Schomerus. 2017. “South Sudan’s Civil War Will Not End with a Peace Deal”; Johnson. 2014. “Briefing”; Rolandsen. 2015. “Another Civil War in South Sudan”; Sefa-Nyarko. 2016. “Civil War in South Sudan.” 96. Justin and De Vries. 2017. “Governing Unclear Lines.” 97. UN OCHA. 2019. “South Sudan: Humanitarian Snapshot.” 98. IOM. 2019. “Displacement Tracking Matrix, South Sudan Mobility Tracking Round 4.” 99. UN OCHA. 2018. “South Sudan: Humanitarian Needs Overview 2018.” 100. UNMISS. 2018. “UNMISS PoC Update.” There are two PoCs in Wau, one on the UNMISS base, and one adjacent to it. 90  |  Informing Durable Solutions for Internal Displacement 190.  The legal framework for the protection of IDPs has been promising. In 2012, South Sudan was signatory to the Kampala convention expressing her determination to protect IDPs. The 2015 peace agreement between the Gov- ernment of South Sudan (GoSS) and the opposition included a mandate of the Transitional Government for National Unity, “expediting the relief, protection, voluntary and dignified repatriation, rehabilitation and resettlement of IDPs.”101 The stability of the latest peace agreement, the R-ARCSS, will be instrumental in determining the return of IDPs and disbarments of armed groups. 191.  The conflict and displacement as well as the fall in oil prices have contributed to widespread food inse- curity. In September 2017, before the harvest, about half of the South Sudanese population (6 million people) were severely food insecure. South Sudan is largely an agricultural society where most people rely on their own production to sustain themselves. Violence, insecurity, and displacement disrupted agricultural activities, including several planting seasons. In 2017, this created a cereal deficit of 500,000 tons, which is enough to feed 2.3 million people for a year.102 In addition, the loss of foreign exchange due to falling oil prices deprived the country of food imports. As of 2019, food insecurity continues to plague half of South Sudan’s population, with about 6.45 million people facing acute food inse- curity in February–April 2019.103 192.  At the same time, the South Sudanese economy has until recently been contracting with rapid devalu- ation and soaring inflation. The economy contracted by 11.2 percent in 2016, 6.9 percent in 2017, and 3.5 percent in the FY2018 (2017–2018).104 These contractions were mostly due to falling oil prices, which triggered a strong deval- uation of the South Sudanese pound and contributed to large price rises. Between September 2015 and September 2016, the consumer price index increased by 549 percent.105 After the signing of the peace deal, prospects look slightly better. The economy is projected to grow at 1.8 percent in FY2019 (2018–2019), and inflation, while high, fell to around 40 percent in December 2018. Oil production increased following the peace deal and is expected to be the major driver of growth in the medium term.106 193.  The conflict and macroeconomic crisis as well as drought have contributed to considerable increases in poverty rates. In 2016, about 80 percent of people lived under the international poverty line of US$1.90 PPP per capita per day. This is a significant increase from 2009 and 2015 when the poverty rate was 51 percent and 66 percent, respec- tively. The depth of poverty in 2016 was as alarming as its breadth, with the average household consuming about half of the international poverty line. These incredible levels of depravation have made South Sudan one of the poorest countries in the world.107 101. ICRC. 2017. “Translating the Kampala Convention into Practice: A Stocktaking Exercise,” 409. 102. FAO (Food and Agricultural Organization of the United Nations). 2017. “South Sudan: Situation Report.” 103. WFP (World Food Programme). 2019. “WFP South Sudan Situation Report #242.” 104. World Bank. 2019. “Macro Poverty Outlook South Sudan.” 105. World Bank. 2017. “South Sudan Poverty Assessment.” 106. World Bank. 2019. “Macro Poverty Outlook South Sudan.” 107. World Bank. 2017. “South Sudan Poverty Assessment.”  BOX B.4    The CRS collects rich micro-data on IDPs to complement the HFS 2017 The CRS represents four of the largest IDP camps in South Sudan as of 2017. The CRS was conducted in 2017 in the four largest PoC camps with defined boundaries. The camps, all in urban areas, are Bentiu PoC, Bor PoC, Juba PoC, and Wau PoC, located in the pre-war states of Upper Nile, Jonglei, Central Equatoria, and Western Bahr-el-Ghazal, respectively (Figure B.114). The CRS collects rich micro-data about consumption, poverty, education, and labor outcomes of IDPs in these camps. It also collects details on displacement-specific outcomes, including motivations for displacement, return intentions, social capital, and pre-displacement outcomes in the standard of living, education, and labor. Though only an estimated 15 percent of South Sudan’s IDPs live in camp or camp-like settings, the detailed micro-data on the PoCs fills important information and knowledge gaps for IDP-focused programming.   FIGURE B.114    HFS 2017 and CRS 2017 coverage (pre-war states) Source: HFS 2017 and CRS 2017. The fourth wave of the High Frequency Survey South Sudan (HFS 2017) allows for comparisons of IDPs to urban resi- dents. The HFS 2017 represents urban areas in 7 of the 10 pre-war states of South Sudan. As the PoC camps covered in the CRS are in urban areas, HFS 2017 allows for comparisons in the outcomes of IDPs and the residents of the areas where they are now located. However, HFS 2017 does not cover two of the pre-war states in which CRS camps are located (Jonglei and Unity). Secu- rity concerns prevented survey activity from being feasible in these states. Thus, comparisons are drawn at the overall urban and IDP level rather than for specific camps or pre-war states. The focus of the analysis is primarily on comparisons between IDPs and urban residents today, and IDPs’ living conditions today compared to their living conditions before displacement. Comparisons to resident populations from IDPs’ place of origin are less clear given data limitations, especially identifying which IDPs in the PoCs were from rural or urban areas. The CRS and HFS 2017 inform how IDPs are different from urban non-IDPs, as well as how different types of IDPs have heterogeneous outcomes. The urban residents present a relevant comparison group since the IDP households surveyed in the CRS are located in urban areas. The urban population provides a benchmark for access to services—such as housing, sanitation, and health—for IDPs in urban camps. Urban education and labor outcomes establish the human capital and labor market condi- tions of the areas that IDPs now find themselves in. Finally, the relationship with surrounding communities affects IDPs’ socioeco- nomic integration. This is especially pertinent since many IDPs do not plan to move from their location in the foreseeable future, and a majority of those who plan to move do not know when the opportunity will arise. Among IDPs, specific characteristics of the household reflect different trajectories and needs, creating the potential for more customized program response. House- holds headed by women can have missing male spouses and larger dependency ratios; poorer households can have lower social and economic capital which affects integration or moving. Thus, different groups of IDPs are also compared to each other.108 108. See Appendix C for details on the surveys and comparison groups. 91 92  |  Informing Durable Solutions for Internal Displacement Demographic Profile 194.  IDPs are significantly younger than urban residents, driving higher dependency ratios. About 45 percent of IDPs are under 15 years old (compared with 32 percent of urban residents, p < 0.01). Consequently, fewer IDPs are of working age, which is between 15 and 64 years old. About 54 percent of IDPs are of working age (compared to 65 percent of urban residents; Figure B.115). This translates to higher dependency ratios, defined as the number of dependents under 15 and over 64 years old, compared to the working-age population. The dependency ratio of IDP households is 1.2, almost twice the ratio of urban residents (0.7) (p < 0.01) (Table B.9). 195.  IDPs and urban residents have fewer adult men than women following years of civil war. South Sudan has fewer men than women.109 In the adult age group, this disparity is most pronounced among urban residents (18 per- cent are men and 23 percent are women, p < 0.01).110 For IDPs, this disparity is less pronounced among adults (15 per- cent men and 18 percent women) but is also present among the youth (9 percent are men and 12 percent are women) (Figure B.115). Most woman-headed households have missing male spouses (83 percent for IDPs and 87 percent for urban). The lack of a male spouse further drives up dependency ratios in woman-headed IDP households (1.4 com- pared to 0.96 for man-headed households for IDPs) (Table B.9). Differences in the sex of the household head between IDPs and urban residents are significant after controlling for various household characteristics (Table B.9).   FIGURE B.115    Population structure for IDPs and   FIGURE B.116    Ethnic composition for IDPs urban residents, by sex and age 100 1.5 0.6 1.4 0.6 80 50 % of population 17.7 40 23.4 15.3 60 18.0 % of population 30 40 9.0 11.5 20 11.9 12.1 20 10 23.3 22.2 15.8 15.9 0 IDP Urban Bentiu Bor Juba Wau 0 PoC PoC PoC PoC Men Women Men Women Urban IDP Overall IDP Under 15 years 15–24 years Nuer Zande Balanda Viri 25–64 years Above 64 years Dinka Bari Shilluk Other Source: Authors’ calculations using HFS 2017 and CRS 2017.  TABLE B.9    Dependency ratio and household size, by sex of household head Urban IDP Male head Female head Overall Male head Female head Overall % of all households 54.5 45.5 100.0 54.4 45.6 100.0 Dependency ratio  0.57  0.79   0.66  0.96  1.41   1.16 Household size  5.3  4.4   4.9  5.9  5.2   5.6 Source: Authors’ calculations using HFS 2017 and CRS 2017. 109. World Bank. 2017. “South Sudan Poverty Assessment.” 110. Age groups are defined as follows: children are under 15 years old, youth are between 15 to 24 years old, adults are between 25 to 64 years old, and elderly are above 64 years old. Volume B: Country Case Studies  | 93 196.  IDPs are mostly from the Nuer tribe, which is associated with the opposition group. The two largest ethnic groups in South Sudan are the Dinka and the Nuer. Other large groups include the Zande, Bari, and Shilluk. The conflict was largely between the government, who is linked with the Dinka, and the opposition, who is affiliated with the Nuer but also includes other groups. About three in four IDPs are Nuer (compared to less than 1 percent of the urban popula- tion), while one in three urban residents are Dinka (compared to less than 1 percent of the IDP population, p < 0.01).111 PoCs often shelter IDPs associated with the opposition but are located in areas controlled by the government.112 The PoCs in Bentiu, Bor, and Juba are almost entirely composed of the Nuer tribe. Other studies have also found IDPs in PoCs belong to the same ethnic group.113 Wau PoC is the only multiethnic camp, with Balanda Viri and Zande as well as other groups, but few Nuer (Figure B.116). Displacement Profile 197.  The displaced overwhelmingly link their displacement to security. IDPs predominantly fled their original residences due to armed conflict (79 percent of IDP households). Discrimination and violence in the absence of conflict are also key reasons for IDPs in the Juba PoC (21 percent and 15 percent, respectively) (Figure B.117). As a result, secu- rity sharply outweighs other factors like humanitarian assistance when choosing a camp location, with more than 9 in 10 households coming to the current camp for security (Figure B.118).  FIGURE B.117    Reasons for leaving original location  FIGURE B.118    Reasons for arriving at current loca- for IDPs tion for IDPs 100 100 80 80 % of households % of households 60 60 40 40 20 20 0 0 Man head Woman head Bentiu PoC Bor PoC Juba PoC Wau PoC Overall Overall Man head Woman head Bentiu PoC Bor PoC Juba PoC Wau PoC Armed conflict Violence but not conflict Security Humanitarian aid Discrimination Other Join family or known people Other Source: Authors’ calculations using HFS 2017 and CRS 2017. 111. Importantly, the urban sample does not include Bentiu and Bor towns. Including these towns would alter the ethnic makeup of the urban population. 112. Norwegian Refugee Council. 2017. “Protection of Civilian Sites: Lessons from South Sudan for Future Operations.” 113. REACH. 2016. “South Sudan Intentions Study.” 94  |  Informing Durable Solutions for Internal Displacement 198.  The geographic trajectory of the conflict can explain displacement dates for IDPs in specific pre-war states. Displacement dates of the IDPs map closely with conflict intensity.114 Most IDP households were displaced when violence escalated. About 28 percent of IDPs were displaced around December 2013 when the conflict broke out. Another 17 percent were displaced in July 2016 when the conflict reignited. Trends in conflict and displacement events appear even more clearly when looking at PoCs. For example, more than 66 percent of IDPs in Juba PoC were displaced in December 2013 when the clashes broke out in Juba. More than 90 percent of Bor PoC IDPs were also displaced at this time too, as the opposition seized control of Bor town a week after the clashes in Juba. For Wau PoC, 78 percent of the households were displaced in June 2016 when battle broke out between the government and the opposition following tensions building in the areas since 2015 (Figure B.119).   FIGURE B.119    Conflict events and displacement dates for IDPs, January 2013–July 2017 200 100 % of households displaced 80 150 Number of events 60 100 40 50 20 0 0 Jan-2013 Mar-2013 May-2013 Jul-2013 Sep-2013 Nov-2013 Jan-2014 Mar-2014 May-2014 Jul-2014 Sep-2014 Nov-2014 Jan-2015 Mar-2015 May-2015 Jul-2015 Sep-2015 Nov-2015 Jan-2016 Mar-2016 May-2016 Jul-2016 Sep-2016 Nov-2016 Jan-2017 Mar-2017 May-2017 Battle Strategic development Remote violence Riots/protests Violence against civilians Bentiu PoC Bor PoC Juba PoC Wau PoC Source: Authors’ calculations using ACLED 2013–2017 and CRS 2017. 199.  Most IDPs are displaced within their state of origin and have not travelled far. Most IDPs are from the former states of Unity (30 percent), Western Bahr el Ghazal (25 percent), Jonglei (21 percent), and Central Equatoria (17 percent), where Bentiu, Wau, Bor, and Juba PoCs, respectively, are located (Figure B.120). About 7 in 10 IDPs are now displaced in their state of origin, with half of them even in their county of origin (Figure B.121). This indicates that they did not travel far to reach the safety of the camps. This also indicates that they would not have to cover large distances, security permitting, to check on their dwellings and livelihoods or, eventually, return home. 200.  IDPs and urban residents have comparable rates of family separation. About 37 percent of IDP households compared with 30 percent of urban households have separated members. IDP households have a slightly larger num- ber of separated members (3.5 compared to 3.1 for urban households). This trend suggests that separation may be driven by conflict but exacerbated by displacement. Among IDPs, woman-headed households are more likely to have separated members, but they have a somewhat smaller number of separated members. Among the camps, house- holds in Bor are the most likely to have separated members (about three in four) and have the largest number of sepa- rated members (almost five members) (Table B.10). 114. ACLED were used for conflict events. Volume B: Country Case Studies  | 95   FIGURE B.120    IDPs’ place of origin, by state   FIGURE B.121    IDPs’ place of origin vs. current location Outside South Sudan 1% Same boma Different state 11% 16% Same payam 7% Same state Same county 35% 30% Source: Authors’ calculations using HFS 2017 and CRS 2017.   TABLE B.10    Trends in separation for IDPs and urban residents Overall IDP Man Woman Bentiu Bor Juba Wau Urban IDP head head PoC PoC PoC PoC % households with separated members 29.9 36.5 34.1 39.3 29.8 76.3 34.9 48.1 Average number of separated members (in  3.1  3.5  3.6  3.4  3.6  4.9  2.9  3.7 households with separated members) % separated members who were women or girls 49.1 45.3 46.9 43.4 46.3 40.7 47.0 43.3 Average age of separated members 28.1 28.7 28.8 28.5 27.0 25.1 30.0 30.1 % households that can contact separated members 88.4 62.0 69.3 52.9 74.6 26.1 57.2 58.5 % households with access to reunification systems N/A 32.4 28.4 35.7 25.0 41.5 29.1 41.2 Source: Authors’ calculations using HFS 2017 and CRS 2017. 201.  IDP households have less contact with separated members, and most do not have access to family reuni- fication mechanisms. Urban households are more able to contact their separated members (about 9 in 10 compared with 6 in 10 IDP households). Among the camps, only one in four households in Bor PoC can contact their separated members compared with three in four households in Bentiu PoC. However, about one in three IDP households have access to family reunification mechanisms. Perhaps because of the considerable number of separated members in Bor PoC (about five) and the household’s inability to contact them, access to reunification mechanisms rises to 42 percent in this camp (Table B.10). Most IDP and urban separated household members are displaced to another location. Very few have stayed behind (10 percent for IDPs and 5 percent for urban residents) and even fewer have been recruited into armed groups (Figure B.122). However, some studies assume separated members have stayed behind or been recruited into armed groups.115 115. REACH. 2015. “South Sudan Displacement Trends Analysis.” 96  |  Informing Durable Solutions for Internal Displacement   FIGURE B.122    Reasons for separation of household  FIGURE B.123    Return intentions of IDPs members for IDPs and urban residents 100 100 80 % of households 80 % of households 60 60 40 40 20 20 0 0 l oC l C C C ra ad ad Po Po Po ve rP IDP Urban Man head Woman head Bentiu PoC Bor PoC Juba PoC Wau PoC he he O au a iu Bo an b an nt W Ju M Be om W Stay here/another camp Return to origin Overall IDP New area/country Deceased Recruited by armed forces Stayed behind Displaced to other location Other Source: Authors’ calculations using HFS 2017 and CRS 2017. 202.  Only one in three IDPs wish to return to their place of origin. About 58 percent of IDPs do not want to leave their current location compared with 34 percent who want to return to their place of origin and 7 percent who want to resettle in a new location. IDP households headed by women are less likely to want to return (29 percent compared to 39 percent for households headed by men). Among the camps, IDPs in Bor PoC are the most likely (75 percent) and in Wau PoC the least likely (25 percent) to want to return to their place of origin (Figure B.123). The reluctance of IDPs in Wau PoC may be due to their more recent experience of conflict and displacement. Households displaced by armed conflict and insecurity are more likely to report they want to move from their current location after controlling for household characteristics and other factors (Table B.11).116 Satisfaction of households and poverty is also associated with a desire to stay in the current location. 203.  IDPs who want to stay in their current location are motivated by better security, services, and assistance in the camps. The main reasons IDPs do not want to leave the PoCs are the provision of security (99 percent), health and education services (74 percent), and humanitarian assistance (66 percent) by the international community. Across the PoCs, humanitarian assistance is more important for IDPs in Bor, and health and education services in Benitu. A 2015 assessment found ongoing violence and insecurity to be barriers to return. In addition, it also found the destruction of assets discourages returning.117 In contrast, family reasons and other economic indicators (that is, improved livelihood, access to land, and other assets) do not feature prominently (Figure B.124). 116. The econometric analysis aims to provide some insights about the different return intentions between IDPs. It only considers data from a cross-section and does not include some variables which might be associated with the decision to stay or move, such as time-variant elements, endowments, and social capital, among other determinants. 117. REACH. 2015. “South Sudan Displacement Trends Analysis.” Volume B: Country Case Studies  | 97   FIGURE B.124    Reasons for staying in current location for IDPs who do not intend to relocate 100 % of households who do not want to move from camp 80 60 40 20 0 Overall Man Woman Bentiu Bor Juba Wau head head PoC PoC PoC PoC Security Better access to home/land/livestock Better access to education and health services Better access to livelihood Family reasons Access to humanitarian aid Source: Authors’ calculations using HFS 2017 and CRS 2017. 204.  Better security and services are also the most important concerns for IDPs who want to leave their cur- rent location. The primary reasons IDPs want to return to their place of origin or resettle in a new location are better security (83 percent), and health and education services (67 percent). A study found that IDPs expect the humanitarian community to continue providing assistance and services when they return home.118 Surprisingly, better access to home/land/livestock and employment are cited by fewer IDPs (43 percent and 40 percent, respectively) (Figure B.125). Other studies have found the reasons IDPs want to return include security, livelihoods, and assets, especially land.119 The main reasons IDPs want to leave their current location include low access to assets, employment, services, and humanitarian assistance, as well as improper management of the camps.   FIGURE B.125    Reasons for moving to new location for IDPs who intend to relocate 100 % of households who do not 80 want to move from camp 60 40 20 0 Overall Man Woman Bentiu Bor Juba Wau head head PoC PoC PoC PoC Better security Better access to home/land/livestock Better access to education and health services Better access to livelihood/employment opportunities Family reasons Access to humanitarian aid Source: Authors’ calculations using HFS 2017 and CRS 2017. 118. REACH. 2016. “South Sudan Intentions Study.” 119. REACH. 2015. “South Sudan Displacement Trends Analysis.” 98  |  Informing Durable Solutions for Internal Displacement   TABLE B.11    Return intention of IDPs Dependent variable: move vs. stay (reference) Independent variables (1) (2) (3) (4) (5) Displaced by armed conflict/insecurity 0.668*** 0.673*** 0.713*** 0.689*** 0.694*** Year of displacement 0.013 0.005 0.032 0.031 0.024 Overall satisfaction120 −0.855*** −0.837*** −0.921*** −0.926*** −0.913*** Poor household −0.345 −0.577** −0.551** −0.540** Receive assistance 1.381* 1.346* 1.376* 1.769** Age-dependency ratio 0.070** 0.065** 0.063** Share of women in the household −0.035 −0.011 −0.000 Share of children in the household −0.322 0.118 0.141 Share of literate members in household 0.277 0.385 0.312 Household head: woman −0.041 −0.150 −0.208 Household head: age −0.376** −0.362** Household head: literate 0.005 0.006 Improved water sources 0.102 0.118 Improved sanitation −0.293 Region fixed effects 0.228 Observations 2,396 2,396 2,382 2,354 2,354 Source: Authors’ calculation using the using HFS 2017 and CRS 2017. Note: Significance level: 1 percent (***), 5 percent (**), and 10 percent (*). The coefficients were estimated from a logistic regression. The poverty status used in the regression was derived from total core consumption and a rescaled poverty line. Standard of Living 205.  About 9 in 10 IDPs are living in poverty. In a country with a staggering prevalence of poverty, IDPs are a partic- ularly marginalized group. About 91 percent of IDPs fall under the international poverty line of US$1.90 PPP (2011) per day per capita compared with 86 percent of rural residents and 75 percent of urban residents (p < 0.01). Among the IDPs, poverty rates vary considerably across the camps. Bentiu PoC has the highest poverty rate, while Bor PoC has the lowest poverty rate (96 percent and 76 percent, respectively, p < 0.01) (Figure B.126). Differences in poverty between IDPs and urban residents seem to be associated with regional effects and household characteristics. The poverty rates become similar after controlling for regional differences, household characteristics, and living conditions (Table B.12).121 Overall, poor households tend to have more members, a smaller share of women in the household, and are less likely to have a literate household head. 206.  Along with higher prevalence, IDPs also have deeper poverty gaps than urban residents. The poverty gap, defined as the average consumption shortfall relative to the poverty line, is 54 percent for IDPs compared with 51 per- cent for rural and 40 percent for urban residents (p < 0.01). IDPs and rural residents who are poor live on less than half the income threshold of US$1.90 PPP (2011) per day per capita. Where poverty is more prevalent, it is also more severe. Bentiu PoC, which has the highest poverty incidence, has the deepest poverty gap (60 percent), while Bor PoC, which has the lowest poverty incidence, has the smallest poverty gap (34 percent, p < 0.01) (Figure B.127). 120. Refers to self-reported overall current life satisfaction. 121. The econometric analysis aims to provide some insights about the differences between IDPs and urban residents. It only considers data from a cross-section and does not include some variables which might be associated with the poverty status of households, such as time-variant elements, endowments, and social capital, among other determinants. Volume B: Country Case Studies  | 99   FIGURE B.126    Poverty headcount   FIGURE B.127    Poverty gap relative to US$1.90 PPP (2011) per day ratio for IDPs and residents per capita poverty line for IDPs and residents 100 140 120 Poverty gap in SSP (mean consumption 80 shortfall relative to % of population 100 40% 34% 45% PPP P line) 51% 54% 53% 56% 54% 60 80 60% 40 60 40 20 20 0 0 Urban Rural 2016 IDP Man head Woman head Bentiu PoC Bor PoC Juba PoC Wau PoC Urban Rural 2016 IDP Man head Woman head Bentiu PoC Bor PoC Juba PoC Wau PoC Overall IDP Overall IDP Mean income Poverty line Source: Authors’ calculations using HFS 2017 and CRS 2017.   TABLE B.12    Demographic attributes of poor households Dependent variable: poor vs. non-poor status (reference) Independent variables (1) (2) (3) (4) (5) Urban residents Reference group Reference group Reference group Reference group Reference group IDP household 0.319 −0.095 −0.584 −0.435 −0.538 Receiving assistance 0.526* 0.635* 0.509 0.290 Household size 0.190*** 0.213*** 0.218*** Age-dependency ratio 0.176 0.247 0.226 Share of women in the −2.617*** −2.983*** −2.733*** household Share of children in the 1.423 1.251 1.409 household Share of literate members in −1.110** −0.359 −0.338 household Household head: woman 0.313 0.119 Household head: age −0.001 0.000 Household head: literate −0.927** −0.995*** Improved water sources 1.100** Improved sanitation −0.386 Region fixed effects Yes Yes Yes Yes Yes Obs. 3,324 3,324 3,281 3,281 3,270 Source: Authors’ calculation using the using HFS 2017 and CRS 2017. Note: Significance level: 1 percent (***), 5 percent (**), and 10 percent (*). The coefficients were estimated from a logistic regression. The poverty status used in the regression was derived from total core consumption and a rescaled poverty line. 100  |  Informing Durable Solutions for Internal Displacement 207.  Despite being poorer, IDPs are less hungry than urban residents. About 24 percent of IDPs have experienced hunger three or more times during the last four weeks compared with 32 percent of urban residents. The lower hun- ger rates among IDPs may be due to a more predictable and stable access to food through aid. For IDPs, hunger rates are similar across the consumption quintiles and across sexes. Among the camps, Bor PoC has the lowest poverty and hunger rates (5 percent), while Bentiu PoC, despite being the poorest camp, experiences less hunger (15 percent) than Juba PoC (31 percent) and Wau PoC (29 percent, p < 0.01) (Figure B.128).   FIGURE B.128    Frequency of facing hunger in the   FIGURE B.129    Food aid and core food consumption, past four weeks for IDPs and urban residents per capita per day for IDPs and urban residents 100 80 (SSP) per capita per day 80 Food consumption % of households 60 60 40 40 20 20 0 0 IDP Urban Man head Woman head Bentiu PoC Bor PoC Juba PoC Wau PoC Poorest quintile Q2 Q3 Q4 Richest quintile IDP Man head Woman head Bentiu PoC Bor PoC Juba PoC Wau PoC Poorest quintile Q2 Q3 Q4 Urban Richest quintile Overall IDP Overall IDP Often (more than 10 times) Sometimes (3–10 times) Aid food Non-aid food Rarely (1–2 times) Never Source: Authors’ calculations using HFS 2017 and CRS 2017. 208.  IDPs receive considerably more food aid than urban residents but still consume less food overall. IDPs receive five times more food aid than urban residents (SSP 15 per capita per day compared with SSP 3 per capita per day for urban residents (p < 0.01). However, they still consume SSP 12 per capita per day less food than urban residents (p < 0.01).122 Across the quintiles, IDPs receive similar levels of food aid. Yet the overall food consumption differs drastically, with the richest quintile consuming more than four times the poorest quintile (p < 0.01) (Figure B.129). This indicates that food aid is not targeted based on income or consumption. 209.  IDPs had better housing before the conflict but now occupy unimproved and temporary camp dwell- ings.123 Before the December 2013 conflict, about 43 percent of IDPs had access to improved housing, and 86 percent owned these dwellings. A recent study confirms that most IDPs living in PoCs owned their homes before displace- ment.124 The pre-conflict housing conditions of IDPs were better than those of urban residents today; 21 percent of urban residents occupy improved housing and 78 percent own the dwelling. The housing standards of IDPs have severely fallen to levels that are well below the urban residents; less than 1 percent of IDPs live in improved housing today, and 94 percent of the dwellings are temporary shelters provided by NGOs or the UN. This indicates that before displacement, IDPs may have been somewhat better off than urban residents but are now significantly worse off (Fig- ure B.130 and Figure B.131). 122. Food aid and overall consumption values are calculated using ‘core’ food items, which reflect about 75 percent of the total food consumption of the sample comprising urban residents and IDPs. However, the total imputed food consumption of IDPs is also less than that of urban households. 123. Improved housing is defined as a structure that is made of wood, concrete, or block and is intended for habitation. 124. REACH. 2015. “South Sudan Intentions Study.” Volume B: Country Case Studies  | 101   FIGURE B.130    Access to improved housing, now and   FIGURE B.131    Trends in tenure of housing, now pre-displacement and pre-displacement 100 100 80 80 % of households % of households 60 60 40 40 20 20 0 0 Urban IDP Overall Man head Woman head Bentiu PoC Bor PoC Juba PoC Wau PoC Poorest quintile Q2 Q3 Q4 Richest quintile Current Origin Current Urban IDP Temporary shelter Squatting Relatives/friends Work Current Origin IDP Rented Owned overall Source: Authors’ calculations using HFS 2017 and CRS 2017. 210.  Camps offer improved water, sanitation, and hygiene (WASH) facilities and closely situated services to IDPs, while urban residents are farther from services. IDPs have nearly universal access to improved drinking water sources. IDPs also have higher rates of improved sanitation facilities, defined as toilets with certain types of disposal and drainage systems, than urban residents (78 percent and 56 percent, respectively) (p < 0.01) (Figure B.132).125 These differences in access to WASH facilities between IDPs and urban residents are robust after controlling for household characteristics. In line with this, IDPs are more likely to be satisfied with their current conditions (Table B.13). Further, IDPs are typically much closer to a health facility, food market, and water point, than urban residents. Before displacement, IDPs were much farther from these services, indicating an increased, current ease of access to these basic amenities (Figure B.133).  FIGURE B.132    Trends in access to improved   FIGURE B.133    Time (one way) to amenities for IDPs and water and sanitation for IDPs and urban resi- urban residents dents 56 100 48 80 40 % of households 32 Minutes 60 24 40 16 8 20 0 IDP origin IDP now Urban now IDP origin IDP now Urban now IDP origin IDP now Urban now IDP origin IDP now Urban now 0 Urban IDP Improved water source Improved sanitation facility Water Health facility School Market Source: Authors’ calculations using HFS 2017 and CRS 2017. 125. World Health Organization and UNICEF. 2006. “Core Questions on Drinking Water and Sanitation for Household Survey.” 102  |  Informing Durable Solutions for Internal Displacement   TABLE B.13    Demographic attributes of IDPs and urban residents Dependent variable: IDP vs. urban residents (reference) Independent variables (1) (2) (3) (4) (5) Household size −0.074** −0.059* −0.066** −0.064* −0.069* Age-dependency ratio −0.129 −0.088 −0.073 −0.062 −0.029 Share of women in the household −1.286** −0.938 −0.839 −0.942 −0.936 Share of children in the household 2.703*** 2.556*** 2.561*** 2.480*** 2.290*** Share of literate members in household −0.469* −0.060 −0.327 −0.441 −0.556 Household head: woman −0.430** −0.626*** −0.629*** −0.506** Household head: age −0.023*** −0.030*** −0.030*** −0.032*** Household head: literate −0.412** −0.468** −0.439** −0.339 Improved water sources 3.433*** 3.352*** 3.167*** Improved sanitation 1.397*** 1.381*** 1.392*** Household reported hunger in past month −0.328* −0.280* Overall satisfaction126 1.501*** Observations 3,281 3,281 3,270 3,264 3,264 Source: Authors’ calculation using the using HFS 2017 and CRS 2017. Note: Significance level: 1 percent (***), 5 percent (**), and 10 percent (*). The coefficients were estimated from a logistic regression. 211.  However, severe overcrowding in camps effectively decreases access to services. Along with quality of flush and waste disposal in toilets, WASH guidelines on improved sanitation also indicate that toilets should not be shared with other households. However, in IDP camps, multiple households share a toilet; a 2015 assessment of PoCs found 57 IDPs per latrine.127 Thus, even though 78 percent of IDP households have access to a toilet with an ‘improved’ waste disposal system, virtually none of them have access to improved sanitation after accounting for toilet sharing (Figure B.134). Overcrowding also affects other living conditions of the displaced, such as housing. The UN defines insufficient living space as having four or more persons per room.128 IDP homes are at least seven times more likely to be overcrowded than urban resident dwellings (58 percent and 9 percent, respectively). Households that are poorer or headed by men are more likely to experience overcrowding (p < 0.01) (Figure B.135). A 2015 assessment also found five in seven PoCs to be overcrowded.129 Overcrowding can adversely affect welfare, especially for women and girls. Having insufficient living space contributes to spreading of communicable diseases such as cholera, diarrhea, and malaria. It can also increase psychological distress. Focus groups in Bor PoC and Juba PoC found that overcrowded shelters and bathing facilities deprived privacy for women, increasing their exposure to certain forms of GBV, such as harassment.130 126. Refers to self-reported overall current life satisfaction. 127. IOM/Camp Coordination and Camp Management. 2015. “South Sudan’s Crisis Response Displacement Tracking Matrix.” 128. UN-Habitat. 2016. “Sustainable Cities and Communities. SDG Goal 11. Monitoring Framework. A Guide to Assist National and Local Governments to Monitor and Report on SDG Goal 11 Indicators.” 129. IOM/Camp Coordination and Camp Management. 2015. “South Sudan’s Crisis Response Displacement Tracking Matrix.” 130. Oxfam. 2017. “South Sudan Gender Analysis.” Volume B: Country Case Studies  | 103   FIGURE B.134    Access to improved sanitation ac-  FIGURE B.135    Crowding in dwellings for IDPs and counting for sharing urban residents 100 100 80 80 % of households 60 % of households 60 40 40 20 0 20 Urban IDP Man head Woman head Bentiu PoC Bor PoC Juba PoC Wau PoC Poorest quintile Q2 Q3 Q4 Richest quintile 0 Urban IDP Unadjusted Adjusted for sharing Overall Source: Authors’ calculations using HFS 2017 and CRS 2017. 212.  IDPs have better educational outcomes than rural residents, but worse than urban residents. About 53 percent of IDPs above 14 years old are literate, compared with 33 percent of rural and 62 percent of urban residents. Women are much less likely than men to be literate in all the three groups, with a disparity of about 35 percent for IDPs and urban residents, and 28 percent for rural residents. Among IDPs, members of households headed by men or in the richest quintile are more literate (p < 0.01 each) (Figure B.136). While more than half of IDPs are literate, few have studied beyond primary school. About one in four IDPs have a secondary school or university education. This is driven by stark sex differences, with 43 percent men but only 10 percent women having studied beyond primary school. Women are also more than twice as likely to have had no education (63 percent and 25 percent, respectively) (Figure B.137).   FIGURE B.136    Literacy rates, 15 years and above for   FIGURE B.137    Adult educational attainment for IDPs IDPs, urban, and rural residents and urban residents, by sex 100 100 % of 15 years and above 80 80 18 years or above % of population 60 60 40 40 20 20 0 Men Women Total Men Women Total Men Women Total Man head Woman head Poorest quintile Q2 Q3 Q4 Richest quintile 0 Men Women Overall Men Women Overall Men Women Overall Urban Rural 2016 IDP IDP Urban Rural 2016 No Education Primary Secondary University Overall Source: Authors’ calculations using HFS 2017 and CRS 2017. 104  |  Informing Durable Solutions for Internal Displacement 213.  IDPs have lower secondary school attendance than urban residents. About 72 percent of primary school age IDP children attend primary school compared with 76 percent of urban children (p < 0.01) (Figure B.138). While secondary school attendance is low in South Sudan, it is particularly low for IDPs. Only 8 percent of secondary school age IDP children attend secondary school (compared to 22 percent of urban children; p < 0.01). Households headed by women, while having lower literacy rates, have higher primary school attendance rates (79 percent compared with 66 percent for households headed by men; p < 0.01). Across the quintiles, primary school attendance is similar but secondary school attendance falls as households become poorer. Most secondary school age students who are not in secondary school are instead in primary school (81 percent of IDPs) (Figure B.139).  FIGURE B.138    Enrollment rates for school-age  FIGURE B.139    Reasons for not attending secondary children school 100 100 80 not in secondary school 80 % of secondary aged Percentage 60 40 60 20 40 0 20 IDP Urban Man head Woman head Bentiu PoC Bor PoC Juba PoC Wau PoC Poorest quintile Q2 Q3 Q4 Richest quintile 0 Boys Girls Overall Boys Girls Overall Urban IDP Overall IDP Too young Look after home Primary school attendance among the primary school aged Lack of finances In primary school Secondary school attendance among the secondary school aged Source: Authors’ calculations using HFS 2017 and CRS 2017. Employment and Livelihood 214.  IDP youth are more likely to be idle than urban youth. In general, youth have lower labor force participation because most of them are enrolled in education.131 However, IDP youth have lower labor force participation than urban youth (32 percent and 63 percent, respectively). Further, the IDP youth are more likely than the urban residents to be inactive (not enrolled or in the labor force; 25 percent and 9 percent, respectively). Thus, one in four IDP youth are idle— neither working, nor looking for work, nor studying. While this is more pronounced for women, a significant number of men are also idle (32 percent and 16 percent, respectively) (Figure B.140).132 131. Labor force participation is the sum of working-age individuals who are looking for work (employed or unemployed). 132. UNDP. 2005. “Youth and Violent Conflict: Society and Development in Crisis.” Volume B: Country Case Studies  | 105   FIGURE B.140    Labor activity status among the working age (15–64 years) 100 % of working age population 80 15–65 years old) 60 40 20 0 Men Women Overall Men Women Overall Men Women Overall Men Women Overall Youth Adults Youth Adults Urban IDP Not enrolled Enrolled Unemployed Employed Source: Authors’ calculations using HFS 2017 and CRS 2017. 215.  Sex-based disparities in the working-age population are much starker for IDPs. Young women have higher labor force participation and lower educational enrollment than young men, suggesting that young men prioritize education over working. This disparity is not very pronounced for urban residents, but more pronounced for IDPs, where most young men are studying, while young women are either working, looking for work, or not studying (51 per- cent of youth IDP men are in education, compared to 28 percent women). Among adults, the labor force participation trends are reversed. Men are more likely to be active in the labor force while women are idle. This could be explained by the number of women who work in their youth but are not educated further, thus staying idle in adulthood. This sex-based disparity is more serious among IDPs than among the urban residents (45 percent of adult IDP women are idle compared to 28 percent adult IDP men and 14 percent adult urban women; Figure B.140). 216.  Current labor trends among adults have implications for the large youth force. If the existing labor trends for adults continue, IDP youth may not find work as adults. The proportion of IDP adults who are neither looking for work nor enrolled is greater than the corresponding proportion for youth. If this trend continues, it is possible that the education enrollment of IDP youth will not translate into labor force participation or employment in adulthood. Rather, a number of the youth who are currently enrolled in education will be inactive in the labor force as adults. For urban residents, youth currently enrolled in education would join the labor force as adults. The proportion of adults who are in the labor force is similar to the proportion of youth who are either in the labor force or enrolled in education. 217.  The employment structure of IDPs and urban residents was similar before the conflict, and it is compa- rable now. About half of IDPs and non-displaced urban populations were receiving salaries, while one in five for all groups ran their own business before the conflict. However, urban residents were more likely to be involved in agricul- ture than IDPs (28 percent compared to 9 percent of IDPs). After the conflict, both groups shifted from salaries to help- ing non-farm businesses—a trend that is more pronounced for IDPs than urban residents. Across IDP camps, IDPs were more likely—before the conflict—to receive salaries in Bor and less likely in Wau (82 percent compared with 42 percent, respectively). In Bor PoC, interestingly, one in three IDPs are now training or apprenticing (Figure B.141). 106  |  Informing Durable Solutions for Internal Displacement   FIGURE B.141    Primary employment activity for IDPs and residents, now and before December 2013 100 80 % of population 60 40 20 0 Before Now Before Now Before Now Before Now Before Now Before Now Before Now Urban Rural 2016 IDP Bentiu PoC Bor PoC Juba PoC Wau PoC Overall IDP Salaried labor Run non-farm business Help in non-farm business Apprenticeship/training Own account agriculture Source: Authors’ calculations using HFS 2017 and CRS 2017.   FIGURE B.142    Main source of livelihood for IDPs and residents, currently and before December 2013 100 80 % of households 60 40 20 0 Before Now Before Now Before Now Before Now Before Now Before Now Before Now IDP Urban Rural 2016 Bentiu PoC Bor PoC Juba PoC Wau PoC Overall Agriculture Wages and salaries Own business Remittances Aid Other Source: Authors’ calculations using HFS 2017 and CRS 2017. 218.  The primary sources of income for IDPs and urban residents were also similar before the conflict, but currently IDPs overwhelmingly rely on humanitarian assistance. IDPs and urban residents relied mostly on agri- culture and salaries (42 percent and 28 percent of IDPs, respectively, and 50 percent and 29 percent of urban residents, respectively) before the conflict. Interestingly, IDPs derived a large part (42 percent) of their livelihoods from agriculture, although few of them were employed in the sector. After the conflict, urban residents rely slightly less on agriculture and salaries and slightly more on businesses (compared to before the conflict) while IDPs largely rely on humanitarian assistance (Figure B.142). 219.  Currently, IDPs have very little agricultural land and livestock and productive assets. Before the conflict, urban residents had access to more land than IDPs (2 acres and 0.8 acres, respectively). While both groups suffered land losses since the conflict, it has resulted in IDP households holding about 0.2 acres of land on average while urban households still hold 1.4 acres. Households in Bentiu PoC had the largest land loss, from 1 acre before the conflict to Volume B: Country Case Studies  | 107 virtually no land after the conflict (Figure B.143). IDP households also suffered from a nearly complete loss of livestock holdings, from 42 livestock units before the conflict to currently 2 units, a fact that suggests IDPs’ involvement in pas- toralism rather than agriculturalism.133 Cattle ownership is an important indicator of social and economic status, and cattle raiding is the source of communal violence in the country. Bentiu PoC had access to the most livestock before the conflict (70 livestock units) and has access to virtually no livestock now (Figure B.144). The sharpest losses in land and livestock can explain why Bentiu PoC is the poorest and most aid-dependent camp. The loss of productive assets was also much starker for IDPs than urban residents.134 While 65 percent of IDP households had access to at least one productive asset pre-conflict, only 13 percent have access currently. Households in Wau PoC are most likely to have access to assets (41 percent), which can explain why Wau PoC is the least aid-dependent camp (Figure B.145).   FIGURE B.143    Holdings of agricultural land, current-   FIGURE B.144    Holdings of livestock units, currently ly and pre-December 2013 and pre-December 2013 3 100 2.5 80 Livestock units 2 60 Acres 1.5 40 1 20 0.5 0 0 IDP Urban Bentiu Bor Juba Wau IDP Urban Bentiu Bor Juba Wau PoC PoC PoC PoC PoC PoC PoC PoC Overall IDP Overall IDP Before December 2013 Current Before December 2013 Current  FIGURE B.145    Ownership of at least one productive asset, currently and pre-December 2013 100 80 % of households 60 40 20 0 IDP Urban Bentiu Bor Juba Wau Richest PoC PoC PoC PoC quintile Overall IDP Before December 2013 Current Source: Authors’ calculations using HFS 2017 and CRS 2017. 133. Livestock include cattle, horses, donkeys/mules, pigs, sheep, goats, poultry, and camels. Livestock units are used to aggregate different types of livestock and allow for regional and global comparisons. They are obtained by converting body weight into metabolic weight. The livestock unit coefficients used here are for the Near East and North Africa region. Cattle—0.70; sheep—0.10; goats—0.10; pigs—0.20; asses—0.50; horses—0.40; camels—0.75; chickens—0.01. Chilonda and Otte. 2006. “Indicators to Monitor Trends in Livestock Production at National, Regional and International Levels.” 134. Productive assets include car, truck, motorcycle, rickshaw, bicycle, boat, plough, computer, refrigerator, and hoe, spade, or axe. 108  |  Informing Durable Solutions for Internal Displacement Social Capital and Cohesion 220.  Many IDPs do not feel safe in the camps, and perceptions of safety are quite low. Almost half of IDPs feel unsafe or very unsafe in PoCs despite the presence of UN peacekeepers (Figure B.146). This is particularly true at night when 68 percent do not feel safe compared to 22 percent during the day (Figure B.147). For IDPs, households headed by women are more likely to feel unsafe (51 percent compared with 42 percent for households headed by men). According to qualitative reports, widows or separated women are more likely to face GBV in the absence of husbands.135 IDPs have directly or indirectly experienced considerable violence. More than three in four households have members who have been threatened with a weapon. About half have been robbed, kidnapped, or extorted. And two in five households have members who have been forced to join armed groups (Figure B.148). These findings show that displacement in South Sudan is usually accompanied by the threat and/or use of violence.   FIGURE B.146    Trends in perceived safety for IDPs   FIGURE B.147    Trends in perceived safety for IDPs and urban residents136 100 100 80 % of households 80 % of households 60 60 40 40 20 0 Urban IDP Man head Woman head Bentiu PoC Bor PoC Juba PoC Wau PoC Poorest quintile Q2 Q3 Q4 Richest quintile 20 0 Violence Walk in the day Walk in the night Very unsafe Unsafe Overall IDP Neither safe nor unsafe Safe Very safe Very unsafe Unsafe Neither safe nor unsafe Safe Very safe Source: Authors’ calculations using HFS 2017 and CRS 2017. 221.  The different nature of risk that men and women face can explain why women-headed households feel less safe inside camps. Men face a higher threat of being killed or recruited into armed groups, which is elevated outside the camp setting. Women face the threat of GBV, which is prevalent not only outside the camp but also inside camps by the police or civilian strangers.137 In a 2016 study, men reported feeling constricted from visiting the forest for collecting firewood or cutting poles for construction, for fear of being killed. In the same study, women and girls were identified as having significantly higher exposure to GBV, even inside camps and settlements. The key perpetrators of sexual violence or rape in PoCs were identified as police, soldiers, and civilian strangers. In addition, overcrowding of dwellings and sanitation facilities translates to a lack of privacy, creating the potential for certain forms of GBV, especially for women and girls. 135. Oxfam. 2017. “South Sudan Gender Analysis.” 136. The safety indicator here is a combined scale of three measures: safety from violence, safety walking in the day, and safety walking at night. Cronbach’s alpha for the scale is 0.65. 137. Oxfam. 2017. “South Sudan Gender Analysis.” Volume B: Country Case Studies  | 109  FIGURE B.148    Trends in exposure to violence after  FIGURE B.149    Relations with neighbors within the December 2013 for IDPs camp for IDPs 100 100 80 80 % of households % of households 60 60 40 40 20 20 0 t t n t n ry g s ul ea en 0 ce po tio in be ns pp hr m r r ea fo to ob li ss lt na ll Bo oC Ju oC W oC c 2 3 4 e w Ex ba ed e ba R ra ra Po Q Q Q til d til ith r P rP P in ve ar Ki m r Ve in Ve au qu w ar tiu ba lh qu O at n ua in st st Be re jo he x rre Th Se to ic Po R ed rc Fo Very bad Bad Neither good nor bad Good Very good Source: Authors’ calculations using HFS 2017 and CRS 2017. 222.  Social capital of IDPs can be analyzed using a bonding, bridging, and linking lens.138 The social relations and networks that IDPs form within and across communities have a direct impact on durable solutions (return, local integration, and resettlement). IDPs experience strong bonding social capital: in fact, most IDPs have positive relations with their current neighbors in the camps (59 percent very good and 24 percent good). This is even true in Wau PoC, which unlike the other camps is multiethnic. Intra-camp relations are better in Bor PoC and worse in Bentiu PoC. Richer households have better relations with their neighbors than poorer households, indicating that they have more social as well as financial capital (Figure B.150). 223.  On the other hand, IDPs have significantly less bridging social capital. Many IDPs do not have good rela- tions with the communities outside the camps. Only 37 percent describe their relationships as good or very good. This could be because most IDPs are from the Nuer tribe and associated with the opposition, while the host communities are largely from the Dinka tribe and supporters of the government.139 A 2015 study found that more than one in three South Sudanese did not trust someone from another ethnic group.140 The IDPs in Bentiu PoC suffer from the worst rela- tions (only 22 percent have good or very good relations), while those in Wau PoC enjoy the best relations (over 72 per- cent have good or very good relations,). Poverty and conflict dynamics help explain these different trajectories. On the one hand, Bentiu is one of the most conflict-affected areas and has changed hands several times. On the other hand, as it was seen above, IDPs in Wau PoC experienced the least change in employment activity from before the conflict, are most likely to have access to assets, and are the least aid dependent. 138. “Bonding social capital refers to relationships among members of a network who are similar in some form (Putnam, 2000). Bridging social capital refers to relationships among people who are dissimilar in a demonstrable fashion, such as age, socio-economic status, race/ethnicity and education (Szreter and Woolcock, 2004). Linking social capital is the extent to which individuals build relationships with institutions and individuals who have relative power over them (e.g. to provide access to services, jobs or resources) (Woolcock, 2001; Szreter and Woolcock, 2004).” Hawkins and Maurer. 2010. “Bonding, Bridging and Linking.” 139. Norwegian Refugee Council. 2017. “Protection of Civilian Sites: Lessons from South Sudan for Future Operations.” 140. South Sudan Law Society and United Nations Development Program. 2016. “Search for a New Beginning: Perceptions of Truth, Justice, Reconciliation and Healing in South Sudan.” 110  |  Informing Durable Solutions for Internal Displacement   FIGURE B.150    Relations with host communities   FIGURE B.151    Frequency of attending public meet- outside the camps for IDPs ings for IDPs and urban residents 100 100 % of households 80 80 % of households 60 60 40 40 20 20 0 Urban IDP Poorest Quintile Q2 Q3 Q4 Richest Quintile 0 l Bo oC Ju oC W oC c e 2 3 st Q4 e l ra Po til Q Q til P rP P in in ve au qu qu iu ba O nt st Be rre he ic Po R Overall IDP Very bad Bad Neither good nor bad Good Never 1–4 times 5 or more times Very good Source: Authors’ calculations using HFS 2017 and CRS 2017. 224.  IDPs have a moderate amount of linking social capital as compared to urban residents. IDPs are more likely to attend public meetings than urban residents (52 percent and 39 percent, respectively). This could be because there are more opportunities in camps run by the international community. Among IDPs, richer households are more likely to attend meetings than poorer households (Figure B.151). Targeting Analysis 225.  About 1 in 10 IDP households are support-dependent, most of them located in Bor PoC, Juba PoC, and Wau PoC. Almost 13 percent of IDP households are support-dependent, 80 percent productive but poor, and 7 percent self-reliant. Urban resident and IDP households are equally likely to be support-dependent. However, urban households are four times more likely to be self-reliant than IDP households. The support-dependent IDP population is concen- trated in the PoCs of Bor, Juba, and Wau. This exposes the heightened vulnerability of IDP households compared to urban residents. Camp-level policies will help target households with specific needs (Figure B.152; Figure B.153).   FIGURE B.152    Vulnerable population by status of   FIGURE B.153    Vulnerable IDP population by (pre- the household war) state 100 100 % of IDP households 80 80 % of households 60 60 40 40 20 20 0 0 Urban resident IDP Bor PoC Juba PoC Wau PoC Bentiu PoC Self-reliant Self-reliant Productive but poor Productive but poor Support-dependent Support-dependent Source: Authors’ calculation using the using HFS 2017 and CRS 2017. Volume B: Country Case Studies  | 111 Typology of IDPs 226.  The analysis for South Sudan shows that IDPs have two distinct profiles. The two groups, Group 1 and Group 2, are of a roughly equal size and represent 40 percent and 60 percent of the IDPs respectively (Figure B.154) (for details about the methodology see Volume C). Before displacement, Group 2 had more agricultural livelihoods, worse housing, and was more likely to be displaced by armed conflict. Contrary to this, Group 1 had wage- and busi- ness-based livelihoods. A majority of Group 1 was also driven by armed conflict, but to a lesser degree than Group 2. Differences in the current conditions of the two groups are possibly derived from their different pre-displacement sit- uations. Group 2 has larger households, higher dependency ratios, higher poverty and aid dependence, and feels less safe. Group 2 is more likely to be confident of a moving timeline, but also more likely to seek information to decide on a move, while Group 1 is more optimistic about the future. Group 1 is largely in Juba PoC and Bor PoC, and Group 2 is concentrated in Bentiu and Wau PoCs.   FIGURE B.154    Visualization of groups from the clustering analysis Source: Authors’ calculation using CRS 2017. Note: Group 1 is represented by the black triangles and Group 2 by the blue circles. Cause Profile 227.  The two groups hail from different (pre-war) states. About 60 percent of households from Group 1 reported that their original residential location was in the (pre-war) state of Central Equatoria, which also holds the national cap- ital, Juba (Figure B.155). In contrast, only 3 percent of Group 2 reported being from Central Equatoria, with a majority— 65 percent—hailing from the oil-rich Unity state, and another 32 percent from Western Bahr-el-Ghazal. All these states have seen waves of conflict, which broke out in Juba in December 2013 and spread to other parts in the country. 228.  Group 2, which is largely from Unity state, was more likely to be displaced by armed conflict, and traveled less far over the course of displacement. Over 8 in 10 households from Group 2 cited armed conflict in their own village as the reason for displacement, as compared to 5 in 10 households from Group 1. Armed conflict in a house- hold’s village indicates a close if not direct exposure to the violence. Group 1 also reported armed conflict as the most common driver of displacement, though not as unanimously as Group 2. Households in Group 1 were also displaced by 112  |  Informing Durable Solutions for Internal Displacement discrimination (19 percent) and increased insecurity without violence (12 percent). In addition to being more driven by armed conflict, households in Group 2 also seem to have travelled less far over the course of displacement. Most house- holds in Group 2 are now in camps that are in the same district (a subdivision of the state), or even the same payam or boma (lower subdivisions from the district), while Group 1 was more likely to travel outside of the district (Figure B.156). Despite these differences in distance, both groups largely stayed in the same state, which could indicate an urgency to find the closest camp given the danger of prolonged travel.   FIGURE B.155    (Pre-war) state of origin   FIGURE B.156    Current location relative to origin 70 50 60 40 % of households % of households 50 40 30 30 20 20 10 10 0 0 Central WBeG Jonglei Unity Upper Same Same Same Same Same Different Eq Nile boma payam county district state state Group 1 Group 2 Group 1 Group 2 Source: Authors’ calculation using CRS 2017. Source: Authors’ calculation using CRS 2017.   FIGURE B.157    Main source of livelihood pre-displacement 60 % of households 40 20 0 re es ss s d er ce Ai tu th ne ag an ul O si W ric itt Bu em Ag R Group 1 Group 2 Source: Authors’ calculation using CRS 2017. 229.  The two groups had different livelihoods before displacement, with Group 2 being more agricultural. More than 6 out 10 households in Group 2 relied on agriculture before displacement, while a majority of Group 1 relied on wages (4 in 10 households) or businesses (2 in 10 households). IDPs in Group 1 have more heterogeneous income sources, while Group 2 is largely agricultural, which can indicate a rural origin (Figure B.157). Though aid and remittance were generally uncommon sources of livelihood, Group 1 was six times more likely to rely on both aid and remittances. This could be linked to a greater access to safety nets and networks for Group 1, possibly due to a presence in urban areas. The differences in sources of income between both groups are significant after controlling for regional effects and other household characteristics, such as the household size and the literacy of the household head. Volume B: Country Case Studies  | 113 230.  Along with having an agricultural livelihood, households in Group 2 were more likely to have access to livestock, land, and assets. Households in Group 2 are four times more likely to have access to any agricultural land than Group 1.141 Further, while households in Group 2 were more likely to have access to at least some land, house- holds in Group 1 had a higher average acreage of land available to them, indicating that Group 2 had broader access to largely smaller landholdings. Group 2 is also one and four-tenths times more likely to own productive assets and live- stock before displacement. Group 2 also had more holdings of livestock on average than Group 1 (48 and 32 livestock units, respectively). This ownership of assets is possibly associated with the agricultural livelihood rather than reflecting wealth. Indeed, Group 2 was less likely to live in improved housing, defined as a structure that is made of concrete or wood, and intended for habitation. Needs Profile 231.  Group 2 has larger households and greater dependency ratios than Group 1. A household in Group 2 on average has almost seven members, compared to four in Group 1 (Table B.14). With members of both groups primar- ily occupying relief tents, this results in overcrowded dwellings for most of Group 2 (82 percent of households are in overcrowded dwellings).142 A larger size for households in Group 2 is driven by children—Group 2 have an average of around four children, compared to two from Group 1. This also results in higher dependency ratios; 0.8 for Group 1 and 1.4 for Group 2.143 In addition, Group 1 is more likely to have woman-headed households. Differences between both groups are significant after controlling for other household characteristics.   TABLE B.14    Current household characteristics and poverty status   Group 1 Group 2 Household size  4.0  6.6 Number of children  2.3  3.8 Dependency ratio  0.8  1.4 Share of households headed by women 53.3 40.4 Poverty incidence (% of population) 87.5 94.6 Poverty gap (% of the poverty line) 51.0 55.8 Source: Authors’ calculation using CRS 2017. 232.  Group 2 is poorer and more likely to rely on aid compared to Group 1, but both groups have similar ser- vice access in the camps. More than 94 percent of the individuals in Group 1 live below the standard international monetary poverty line, compared to 87 percent in Group 1 (Table B.14).144 Regardless of the livelihood source before displacement, both groups now rely primarily on aid. However, households from Group 2 are more likely to depend on aid as the main source of household income (80 percent vs. 69 percent, respectively). Households in both groups largely feel unsafe moving in and out of the camp, and have access to improved but overcrowded WASH facilities.145 Households from Group 1 have seen an improvement in water and sanitation conditions since displacement, though 141. Having access to land does not necessarily imply ownership. 142. Overcrowded dwellings are defined as those that have more four or more people per room. 143. The age dependency ratio is defined as the proportion of children and old age dependents to working-age population (15–64). 144. The poverty line corresponds to a daily value of US$1.90 PPP per day. 145. Access to improved water and sanitation sources is high in the camps, but after accounting for sharing (WASH guidelines state that toilets, apart from having a quality of structure and drainage, also should be shared within the household only, to be ‘improved’), the access to improved sanitation virtually drops to 0 as all multiple households share toilets. 114  |  Informing Durable Solutions for Internal Displacement this improvement is due to the stay in camps, assumed to be temporary. The IDP camps offer services at close hand, with both groups being within 30 minutes of key amenities such as health clinics, schools, and markets.   FIGURE B.158    Perceptions of current situation Feel safe from crime & violence Good relations with neighbors in camp Bad living conditions today 0 20 40 60 80 100 % of households Group 1 Group 2 Source: Authors’ calculation using CRS 2017. 233.  Both groups have similar perceptions about their current living conditions and relations within the camp, but Group 1 perceives more safety from violence. Households in both groups have a similar perception of their rela- tions with neighbors within the camp—more than 8 in 10 households believe that they enjoy positive relations with others. Both groups also hold a positive perspective of the living conditions today, possibly due to the close access to amenities such as schools and health centers and WASH facilities, even though the camps are severely overcrowded. However, households from Group 1 feel safer from crime and violence (58 percent) compared to those from Group 2 (39 percent, Figure B.158). Group 2 was more likely to be displaced by armed conflict in the village. This exposure could have a bearing from the trauma they experienced and the concern for safety. Solutions Profile 234.  In both groups, a decision to move is guided by security, and a majority of households want to stay in the camp. The widespread conflict is likely the overarching factor that unites the considerations of households from these different backgrounds. For 9 out of 10 households in each group, security is the primary factor guiding a decision to stay or move, and also the primary factor required for them to settle anew. Households from both groups have very similar intentions of returning: about 6 out of 10 households in each group want to stay in the current location while 4 out of 10 want to move, primarily to the original residence (Figure B.159). 235.  Households in Group 2 are more confident of a moving timeline and seek more information to inform their decision to move. In Group 2, the households that would like to move are more likely to have a sense of when they would want to relocate. Around half of the households from Group 2 that intend to move are planning on doing it within the next 12 months. Contrary to this, half of the households from Group 1 that intend to move do not have a clear timeline (Figure B.159). While being surer of when they would like to leave, Group 2 is also more likely to perceive having an incomplete amount of information, and of wanting to know about the political and security situation to inform a decision (Figure B.160). The higher exposure to armed conflict in the village might explain the need for security information. Volume B: Country Case Studies  | 115  FIGURE B.159    Return intention and timing of moving 70 60 50 % of households 40 30 20 10 0 Stay 6 months 6–12 months ≥12 months Don't know when Group 1 Group 2 Source: Authors’ calculation using CRS 2017.  FIGURE B.160    Information required to decide whether to stay or move 80 70 60 % of households 50 40 30 20 10 0 Have all info Security and political info Other info Group 1 Group 2 Source: Authors’ calculation using CRS 2017. 236.  Group 1 is more optimistic about the future than Group 2. About one in three households in Group 1 expect their personal future to be better or much better, compared to one in five households from Group 2. A better percep- tion about the future in Group 1 is probably associated with a lower poverty incidence and less dependency on aid, as well as a better perception of safety. Policy Implications of IDP Typology 237.  The two groups can be differentiated based on current and pre-displacement parameters. Group 2 has larger households, more children, and higher dependency ratios. They were also more likely to be in agriculture before displacement, while Group 1 was either involved with wages or business. Further, Group 2 was less likely to have lived in improved housing before displacement (concrete or wood-based structures intended for habitation). Most of Group 2 are found in Wau PoC and Bentiu PoC, while Group 1 is largely in Juba PoC. 116  |  Informing Durable Solutions for Internal Displacement 238.  The two types of IDPs have different policy needs. Group 2 has more children and, as a result, larger house- holds. This makes the need to ensure regular attendance and schooling quality a higher priority for these households. Further, child-specific vulnerabilities, including child abuse—mostly to girls—need to be addressed. Ensuring basic and timely care for young children, particularly infants and those under five years of age, is critical to maintain and enhance long-term human capital, which will have a bearing on the future of the nation’s productivity, poverty, and workforce. To address the high dependency ratios, better vocational training targeted toward the working age, and other oppor- tunities to build technical skills and prevent a protracted economic inactivity which could lead to skill erosion, are key avenues for a developmental response. 239.  Group 2 also has greater exposure to armed conflict and a lower perception of safety today. Interventions that focus on armed conflict–related trauma are especially relevant. Cognitive and non-cognitive approaches that can help overcome conflict-induced and displacement-induced trauma can work well with vocational development and efforts to prevent human capital depletion, and further enhance it. This is especially pertinent because physical capital in the form of housing and assets have largely been lost, and replacing them in a post-conflict stage will need to be accompanied by strong human capital, which can start being nurtured and built in the current stage through health, education, and skill-based programming. 240.  Both groups reported information needs. Regular and reliable information about the security and political situation in the origin as well as in potentially new areas is needed, as nearly 4 in 10 households in Group 1, and 5 in 10 households in Group 2 believe that they do not have enough knowledge to inform a decision to return, move, or stay. Reliable information flows to the displaced communities about their origins and areas will also continue to be relevant in a post-conflict stage concerning any news about security, infrastructure, and the reconstruction efforts for towns and cities affected by the conflict. Conclusion Informing Durable Solutions 241.  During the war, IDPs experienced devastating violence, separation, and losses. They suffered considerable violence, including exposure to armed conflict as well as robbery, kidnapping, and extortion. The adverse psychological, social, and economic effects of experiencing these types of violence can take years to overcome. Their dwellings were destroyed or abandoned. This means that few have a home that they can return to with certainty. They lost substantially more livestock and productive assets than urban residents. This makes it difficult for them to sustain themselves during their displacement or re-establish their livelihoods if they return to their place of origin. They also suffered slightly more family separation than urban residents. Urban residents were also affected by the war and experienced violence, sepa- ration, and losses but not on the same scale as IDPs. 242.  They are now worse off in most areas because the adverse impacts of the conflict are exacerbated by displacement. IDPs are significantly poorer and more likely to be unemployed than urban residents. They mainly rely on assistance and live in tents provided by the humanitarian community, which are not sustainable. They suffer con- siderably more from overcrowding than urban residents, which can lead to psychological distress, tensions with other camp dwellers, and disease outbreaks. They tend not to feel safe in the camps or enjoy freedom of movement. Finally, they have tense relations with the host community, which is a critical barrier to local integration. Volume B: Country Case Studies  | 117 243.  But IDPs are also better off in some areas thanks to the humanitarian community. They have good rela- tions with their neighbors in the camps. They are somewhat less hungry because they receive considerably more food assistance. They have better access to most services. Many also have access to dispute resolution mechanisms, while some even have access to family reunification mechanisms. Finally, they participate more in public affairs. However, all of this is unsustainable as it is largely due to IDPs living in camps where the humanitarian community provides them assistance and services. 244.  Before the conflict, IDPs were comparable to urban residents in many areas. They had similar access to services as well as comparable employment and livelihood structures. However, IDPs were slightly less likely to be employed in and derive their livelihoods from agriculture. This indicates that they were somewhat more urbanized, as involvement in agriculture is usually associated with rural residents. IDPs also had slightly lower literacy rates than urban residents but significantly higher literacy rates than rural residents, suggesting again that they were city dwellers. These similarities can indicate that IDPs who decide to stay will not need to make major adjustments to their way of life. They could also indicate that the habits and skills they develop while they are displaced are transferable if and when they decide to return. 245.  They were even better off in some areas. IDPs had access to better housing and tenure arrangements. They also owned substantially more livestock. Livestock ownership is an important indicator of economic and social well- being in South Sudan. It can also lead to conflict as cattle raiding is a common source of communal violence in the country. Urban residents, on the other hand, had access to significantly more land. This is not surprising given that they seem to have been more involved with agriculture. These differences in asset ownership suggest that some IDPs who reared livestock before displacement may now need to learn how to grow crops if they decide to stay. 246.  Security, services, and humanitarian assistance influence durable solutions. Most IDPs want to stay in their current location (58 percent). The main reasons they want to stay include access to security, services, and humanitar- ian assistance. While security is the main reason that IDPs left their places of origin and settled in the camps, access to health, education, and humanitarian assistance are also important. Though a peace agreement was signed recently, outbreaks of violence continue. Once security conditions stabilize, the provision of health, education, and humanitarian assistance in places of origin could encourage IDPs to return. About one in three IDPs want to return home now. The main reasons they want to move include access to security, services, assets, and employment/livelihood opportunities. Where security conditions have already improved, the provision of health, education, and employment opportunities, combined with property and asset restoration mechanisms, can help IDPs make sustainable returns (Table B.15).   TABLE B.15    Summary of factors motivating mobility intentions Positive factors Negative factors Stay in camp • Security • Few assets (home/land/livestock) • Health and education services • Little employment • Humanitarian assistance • Low access to health and education services • Low access to humanitarian assistance • Improper management of the camp Return home or resettle elsewhere • Better security • Conflict/insecurity • Health and education services • Discrimination/persecution • Assets (home/land/livestock) • Low access to health and education services • Livelihoods/employment opportunities • Low access to humanitarian assistance 118  |  Informing Durable Solutions for Internal Displacement 247.  Any durable solutions will depend on the stability of the peace agreement and the provision of services. The PoCs were never seen as durable. Since PoCs were mostly built in UNMISS camps, the political implications were enormous, and on several occasions the GoSS accused the UNMISS of hosting SPLM-IO and armed groups. Possible solutions hence include transferal of IDPs’ alternative locations, encouraging voluntary movement, or return inclusive of ‘continued services and shifting the security support so the IDPs could be located outside UNMISS perimeters’.146 248.  Return will largely depend on the improvement of the security situation as well as pro-active manage- ment of the sensitive land conflicts. In several cases, land has been grabbed from groups after they have abandoned their plots due to fighting. These are major issues that require immediate attention by the GoSS and should be a priority in search for sustainable peace.147 249.  Looking ahead, policies and programs need to shift from encampment toward sustainable development for IDPs, returnees, and their hosts. For IDPs who want to stay, in the short term the focus needs to be on (a) main- taining and building their human capital (food security, health, and education) and (b) improving their living conditions (housing and sanitation). In the medium term, the focus needs to gradually shift to (a) providing opportunities for the socioeconomic integration of IDPs outside camps (social cohesion, skills development, jobs, and access to assets), (b) continuing to build their human capital (nutrition, health, and education), (c) reducing humanitarian assistance where possible, and (d) supporting host communities to absorb IDPs and improve their own living conditions (services and economic opportunities). For IDPs who want to return, the focus needs to be on (a) providing opportunities for the socioeconomic reintegration of returnees (social cohesion, skills development, jobs, and property and asset restoration; the latter can include providing some guarantee to returning IDPs that they can return to their pre-displacement land and, if possible, housing), (b) continuing to build their human capital (nutrition, health, and education), (c) providing humanitarian assistance where necessary, and (d) supporting host communities to absorb the returnees and improve their living conditions (services and economic opportunities).148 As many IDPs do not currently wish to return, a possi- ble mechanism can be for willing groups to first return under a pilot plan, with a feedback mechanism to the remaining IDPs, about the viability of return. IDPs in Sudan Introduction and Country Context 250.  Sudan has one of the largest stocks of IDPs in the world. Decades of protracted conflicts and human rights violations have been the main drivers of forced displacement in Sudan. Although the number has been decreasing since 2016, the most recent available estimates indicate that as many as 2 million individuals, 5 percent of Sudan’s pop- ulation, are internally displaced as a result of conflict. This makes Sudan the country with the seventh largest IDP pop- ulation in the world.149 In addition to the burden of internal displacement, Sudan is a transit and destination country for asylum seekers and a destination country for refugees, most notably from South Sudan but also from Syria, Eritrea, Chad, and the DRC, among others. 146. Arensen. 2016. “If We Leave We Are Killed—Lessons Learned from South Sudan Protection of Civilian Sites (2013–2016),” 61. 147. Arensen. 2015. “Historical Grievances and Fragile Agreements: An Analysis of Local Conflict Dynamics in Akobo”; Arensen. 2016. “If We Leave We Are Killed—Lessons Learned from South Sudan Protection of Civilian Sites (2013–2016).” 148. Bohnet. 2016. “Back to Turmoil: Refugee and IDP Return to and within South Sudan,” 26. 149. UNHCR. 2018b. “UNHCR Global Trends: Forced Displacement in 2017.” Volume B: Country Case Studies  | 119 251.  Forced displacement constitutes a severe development challenge. Extreme poverty is increasingly concen- trated among vulnerable groups, including individuals who had to flee in the wake of conflict, violence, human rights abuses, and natural disasters. Because of violence and displacement, these groups have typically endured physical and psychological harm as well as loss of their economic livelihoods. Displacement is also often associated with a disruption of long-term investment in human capital as access to health and education is interrupted. A sudden increase in popu- lation density in the area of refuge, often akin to an accelerated urbanization, puts pressure on scarce resources such as land, housing, food, and public services affecting host communities, as well as forcibly displaced people.150   FIGURE B.161    Nominal wholesale prices in Khartoum and Al Fashir (SDG per kg), January 2016–July 2018 25 25 25 20 20 20 SDG per kg 15 15 15 10 10 10 5 5 5 0 0 0 Jan-16 Jul-16 Jan-17 Jul-17 Jan-18 Jul-18 Jan-16 Jul-16 Jan-17 Jul-17 Jan-18 Jul-18 Jan-16 Jul-16 Jan-17 Jul-17 Jan-18 Jul-18 Khartoum El Fasher Khartoum El Fasher Khartoum El Fasher (a) Wheat (b) Sorghum (c) Millet Source: Authors’ calculations based on monthly price data from FAO’s FPMA. The prices are nominal wholesale prices per kilogram. Sorghum is feterita. 252.  Adding to years of slow growth and macroeconomic imbalances, recent food price hikes likely had adverse economic consequences for urban populations, including IDPs and host communities in Al Fashir. The secession of South Sudan in 2011 resulted in an abrupt decline in exports of oil and associated government reve- nue. Macroeconomic imbalances subsequently increased as evidenced by the emergence of multiple exchange rates, deficit monetization and subsequent hikes in the inflation rate, and low rates of economic growth. The lifting of U.S. economic sanctions against Sudan in October 2017 ended an embargo in place for more than two decades. While this step initially fostered optimism, Sudan remained on the U.S. list of State Sponsors of Terrorism. A US$8.9 billion dollar fine against BNP Paribas in 2015 for sanction violations resulted in a loss of correspondent banking relations that did not improve after the end of economic sanctions. Deficit monetization and the removal of wheat subsidies in early 2018 further accelerated inflation: the consumer price index increased by 112 percent between January 2016 and March 2018 and by 31 percent since December 2017 alone. Wholesale prices for wheat, sorghum (feterita), and millet in Al Fashir in May 2018, when most of the survey work was conducted, were 73, 116, and 178 percent higher than one year earlier, respectively (Figure B.161). 253.  Conflict and internal displacement have been concentrated in Darfur. Approximately two-thirds of all con- flict events in Sudan since 2003 took place in the five Darfuri states with one in four in the state of North Darfur (of which Al Fashir is the capital) alone (Figure B.162). In 2016, the most recent year for which data are available, 83 percent of 150. Alix-Garcia and Saah. 2010. “The Effect of Refugee Inflows on Host Communities: Evidence from Tanzania.” 120  |  Informing Durable Solutions for Internal Displacement Sudan’s IDPs were concentrated in Darfur.151 Between 2003 and 2013, the war in Darfur had already internally displaced an estimated 1.7 million people and driven another 280,000 into Chad. Yet the entire population of Darfur—around 8 million as of 2013—has been affected by the conflict. Around 40 percent of Darfur’s population has received monthly food aid for the past decade.152 New displacement related to inter-communal conflicts in the region peaked again at around 430,000 individuals in 2014 after a low of 80,000 in 2011.153 The majority of the displaced are likely to be pro- tracted cases.   FIGURE B.162    Conflict events in Sudan since 2003, percentage by state Source: Authors’ calculations using ACLED (conflict events 2003–2017) data. 254.  Most IDPs in Darfur live in camps that increasingly resemble permanent, informal settlements but rely heavily on humanitarian support. While internal displacement in Darfur has become protracted and camps increas- ingly resemble permanent yet informal settlements, most IDPs in Darfur either rely heavily on humanitarian assistance or fall short of meeting their basic needs. For instance, in 2016 two-thirds of IDPs reportedly struggled to meet their basic food needs by themselves. Long-term development perspectives are equally lacking. UN OCHA estimates that out of the 1.7 million Sudanese children who lack access to education, more than half are IDPs.154 255.  The present note uses original survey data from two IDP camps at the outskirts of Al Fashir in North Darfur to shed light on the living situations of IDPs and to inform actions toward durable solutions. Household-level data were collected between May and July of 2018 among IDPs from the two camps and hosts in the neighboring Al 151. UNHCR. 2016. “SUDAN: Refugees, Asylum Seekers, IDPs and Others of Concern to UNHCR by State.” 152. GoS. 2013. “Developing Darfur: A Recovery and Reconstruction Strategy.” 153. UN OCHA. 2017a. “Sudan Humanitarian Needs Overview 2017”; UN OCHA. 2016. “Sudan: Humanitarian Bulletin.” 154. UN OCHA. 2017a. “Sudan Humanitarian Needs Overview 2017.” Volume B: Country Case Studies  | 121 Fashir (Box B.5). Both Abu Shouk and Al Salam camps were established at the height of the conflict. Founded in 2004, Abu Shouk IDP camp is located 2.5 km northwest of Al Fashir town and was originally established by the Spanish Red Cross and the GoS for approximately 30,000 IDPs from Jebel Si, Korma, and Tawilla.155 Due to a water shortage in the camp and additional inflows in 2005, about 23,000 IDPs were moved to the newly created El Salam, about 1 km east (Box B.6).156   BOX B.5    The IDP Profiling Survey Sudan The IDP Profiling Exercise in Sudan covered 2,000 IDP households from the two camps of Abu Shouk and El Salam as well as 1,000 households from the immediately neighboring city of Al Fashir. IDPs that live in the city but outside of the camps were excluded, as were refugees from other countries. The interviews were conducted between April and July 2018, with a five-week pause to avoid the collection of unrepresentative consumption data during the month of Ramadan. Since both camps are in the immediate neighborhood of the city, this setup allows for good comparability between IDPs and the host population. The urban population provides a benchmark for access to services such as housing, sanitation, and health for IDPs in urban camps. The urban education and labor outcomes indicate the human capital and labor market conditions of the areas that IDPs now find themselves in. Finally, the relationship with surrounding communities affects IDPs’ socioeconomic integration. This is especially pertinent since many IDPs do not plan to move from their location in the foresee- able future, and a majority of those who plan to move do not know when the opportunity will arise. Among IDPs, specific charac- teristics of the household reflect different trajectories and needs, creating the potential for more customized program response. At the same time, this setup limits the representativeness of the data to IDPs in camps around Al Fashir. Based on a comparison between 25 camps in Darfur, Abu Shouk and El Salam appear to be two of the more permanent and established camps in that they have high rates of people living in permanent structures with their own source of income (Figure B.163 and Figure B.164). They are also more accessible to researchers. While most IDPs in Darfur live in camps close to urban centers and some of the results may well hold for IDPs in Darfur more broadly, these limitations need to be kept in mind in interpreting ana- lytical results.   FIGURE B.163    Camp comparison: dwelling made   FIGURE B.164    Camp comparison: primary income of mud is nonagricultural labor 60 90 80 50 70 % of population % of population 40 60 50 30 40 20 30 20 10 10 0 0 Al Shabab Sisi Shangli Tobaya Abu Shouk El Salam ZamZam New ZamZam Nertiti Deleij Elsalam Tayba Shadad Krenik Sisi Shangli Tobaya Abu Shouk Kabkabia Taweela ZamZam New Bindisi Deleij Nertiti Abasi Elsalam Shadad Krenik Source: Authors’ calculations using WFP data. Source: Authors’ calculations using WFP data. 155. Soliman et al. 2017. “Water Associated Diseases amongst Children in IDPs Camps and Their Relation to Family Economic Status: Case Study of Abuschock IDPs Camp, North Darfur State, Sudan.” 156. Tearfund. 2007. “Darfur: Water Supply in a Vulnerable Environment: Phase Two of Tearfund’s Darfur Environment Study.” 122  |  Informing Durable Solutions for Internal Displacement Demographic Profile 256.  IDPs are young and in equal numbers of men and women; their demographic profile resembles that of non-IDP populations more than that of newly registered IDPs in Sudan. Both IDP and host populations are young: 43 percent of the IDP population is less than 15 years of age, compared to 40 percent of the host population (p < 0.05, Figure B.165). This is close to one 2014 estimate for the entire population of Sudan (45 percent).157 Setting the threshold at 18 years, 50 percent of the people in Abu Shouk and El Salam are children, a smaller proportion than what is typically observed among newly registered IDPs.158 The IDP population studied in this note appears to be closer therefore to the host population than to other IDPs in Sudan. The share of women is one in two and is not statistically different between IDPs and hosts in Al Fashir. This again is a slightly lower proportion of women among IDPs than for the newly registered in 2016, where it was 54 percent.159 These differences point toward Abu Shouk and El Salam as being different from the average camp in these basic demographics.  FIGURE B.165    Population structure for IDPs and urban populations, by sex and age 60 2.17 1.83 2.37 1.52 50 % of population 40 15.3 15.59 16.55 18.22 30 11.21 11.24 10.82 11.49 20 15.4 14.5 13.59 13.81 10 6.41 6.39 6.08 5.57 0 Male Female Male Female IDP Host Under 5 years 5–14 years 15–24 years 25–64 years Above 64 years Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 257.  A large share of IDP households are headed by women, especially in Abu Shouk. A woman is responsi- ble in 56 percent of the households in this camp, which is in contrast to only 41 percent in the neighboring El Salam (Table B.16). However, in the host population two in three households are headed by a man, which is still significantly more than for both IDP camps (p < 0.01). Still, the relatively moderate shares of women-headed households contrast with some earlier reports which state figures of up to 80 percent.160 These trends can only be partly attributed to miss- ing spouses of women household heads: most are married (74 percent and 60 percent for IDPs and hosts), albeit less so than man household heads (94 percent and 96 percent for IDPs and hosts). 258.  IDP households have fewer members and more dependents, although the differences seem small. Hosts have on average one dependent household member per working-age adult, which constitutes a significant burden for the households. IDPs even have slightly more dependents (p < 0.01). On average, households in the camps have 6.1 157. Central Bureau of Statistics (CBS) and UNICEF Sudan. 2014. “Multiple Indicator Cluster Survey 2014 of Sudan, Final Report.” 158. UN OCHA. 2016b. “Sudan Humanitarian Needs Overview 2017.” 159. IOM. 2016. “Displacement Tracking Matrix Sudan: Dashboard January–December 2016.” 160. See for example, World Bank. 2017. “Sudan IDP Profiling Preparation,” on the basis of WFP data. Volume B: Country Case Studies  | 123 members while host households consist of 6.6 people (p < 0.01, Table B.16). This is at least partly driven by the small IDP woman-headed households, which only have 5.7 members on average. Again, the difference to the host community is not very pronounced.   BOX B.6    The Abu Shouk and El Salam IDP camps The two IDP camps are located just on the northern fringes of Al-Fashir town and are the size of a town on their own, yet their exact population numbers are difficult to estimate. Abu Shouk was opened in 2004, 2.5 km from the northwestern corner of Al Fashir. A 2013 biometric registration exercise carried out by WFP and IOM counted 44,531 individuals, far less than the 80,000 that qualitative reports estimated in 2012.161 The El Salam camp was added in 2005 just next to Abu Shouk toward the east. The latest population numbers are from 2011, when the IOM estimated its population at 35,552.162 The camps provide numerous services, educational facilities, and institutions to improve livelihood opportunities. Abu Shouk camp has 3 health centers, 45 water sources, 18 kindergartens, 18 primary and 4 secondary schools, 52 Quranic khalwa schools, 1 market, 83 mosques and prayer sites, 1 police station, and 2 public transport routes. It also hosts a microfinance firm funded by the Ministry of Finance and a women’s center to improve women’s livelihood opportunities and self-reliance.163 Sim- ilarly, El Salam camp has 8 health facilities, 56 water sources, 13 kindergartens, 13 primary and 4 secondary schools, 22 khalwa schools, 2 markets with over 150 shops, and 14 community centers, including 2 women’s centers and 1 training center.164 IOM offers vocational training and income-generating activities for IDPs such as agricultural training and courses for welding, auto mechanics, construction, food processing, handicraft, and business planning in the camps, although it is unclear how regularly.165 The facilities often are in bad working condition or lack materials. Health facilities in El Salam lack electricity, medical staff, and equipment; schools are located in unsuitable structures and lack books and other supplies; and water sources are often not functional and overused.166 For Abu Shouk, there are reports of similar shortcomings.167 The camps and Al Fashir town are economically interrelated. Different types of water sources are used in Abu Shouk, includ- ing pipelines from Al Fashir town. By 2007, the increase in demand reduced water tables in reservoirs, resulting in dried up bore- holes. Water services in Al Fashir town were estimated to serve only 40 percent of the town population in 2009. At the same time, water distributed by aid agencies to IDPs in camps was partly sold to residents from Al Fashir town, undercutting urban water sellers.168 There are also reports of the IDP camps serving as tax havens, where products such as coal and animals are shipped to the camps to avoid taxes that are payable upon entering the town. The products are then either sold directly in the camp markets, which are frequented by all Al Fashir residents, or smuggled into town.169 161. IOM. 2015b. “North Darfur State—Abu Shouk IDP Camp”; SUDO UK. 2012. “The Humanitarian Situation in Abu Shouk Camp, Al-Fasher, North Darfur.” 162. IOM. 2015a. “North Darfur—Al Salam IDP Camp Profile 2015.” 163. Ali and Mahmoud. 2016. “From a Temporary Emergency Shelter to an Urbanized Neighborhood: The Abu Shoak IDP Camp in North Dārfūr”; IOM. 2016a. “Abu Shouk IDP Camp Celebrates the Official Opening of Its First Women’s Centre.” 164. IOM. 2015. “North Darfur—Al Salam IDP Camp Profile 2015.” 165. IOM. 2014. “IOM Sudan: Annual Report 2014”; IOM. 2016b. “Sudan Livelihoods Factsheet.” 166. IOM. 2015a. “North Darfur—Al Salam IDP Camp Profile 2015.” 167. IOM. 2015b. “North Darfur State—Abu Shouk IDP Camp”; SUDO UK. 2012. “The Humanitarian Situation in Abu Shouk Camp, Al-Fasher, North Darfur.” 168. DFID and UN-Habitat. 2009. “Darfur: Profile of El Fasher Town and Abu Shouck IPD Camp”; Tearfund. 2007. “Darfur: Water Supply in a Vulnerable Environment: Phase Two of Tearfund’s Darfur Environment Study.” 169. Buchanan-Smith and Fadul. 2008. “Adaptation and Devastation: The Impact of the Conflict on Trade And Markets in Darfur”; IOM. 2015b. “North Darfur State—Abu Shouk IDP Camp.” 124  |  Informing Durable Solutions for Internal Displacement   TABLE B.16    Household composition for IDP and host populations, by sex and household head IDP Host Man head Woman head Overall Man head Woman head Overall % of all households 49.19 50.81 100 68.02 31.98 100 Dependency ratio  1.11  1.16 1.14  1.05  0.89 1.00 Household size  6.46  5.69 6.06  6.68  6.36 6.58 Abu Shouk (%) 44.37 55.63 100 El Salam (%) 59.31 40.69 100 Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Displacement Profile 259.  Conflict is the principal reason for displacement and settling in the camps. For more than 9 in 10 IDP house- holds, armed conflict in their village of origin was the most important reason for leaving (Figure B.166). By including conflict in neighboring villages and general violence increases, this figure rises to 98 percent of IDPs. Of those who stated conflict in their village as the most important driver, four in five say that conflict in other villages and general violence were the second most important reason. Similarly, for 92 percent the good security situation in the camps was the main reason for settling there (Figure B.167). About two in three of these say that the second most important reason was the good access to assets and services. Overall, however, other factors than immediate security play a negligible role in the displacement motivations of the people of Abu Shouk and El Salam (Box B.7).  FIGURE B.166    First and second most important reasons for leaving original location for IDPs 100 80 % of households 60 40 20 0 IDP Overall IDP Overall Man head Woman head First reasons Reasons apart from conflict in own village Family reason Lack of access to employment opportunities Lack of access to assets/services Drought/famine/flood Increased crime/violence Armed conflict in neighboring villages Armed conflict in my village Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 260.  The IDPs in these camps were largely displaced in the initial phase of the conflict and have been living in the camps for almost 15 years. About four in five of the people living in the camps today left their original place of residence in the first two years of the war in Darfur, 2003 and 2004 (Figure B.168). In the chronology of arrivals, it is still evident that El Salam served as a replacement when the capacities of Abu Shouk were exhausted closely after the out- break. Only 10 percent of the IDPs were displaced during the more violent recent outbreaks since 2012. This matches the overall numbers of recorded newly displaced in Darfur, which peaked in 2003 and 2004 with a much smaller increase in 2014.170 Considering all the years of possible return migration, this supports the view that these camps are very established, permanent settlements and that most IDPs are protracted cases. 170. UN OCHA. 2017a. “Sudan Humanitarian Needs Overview 2017.” Volume B: Country Case Studies  | 125  FIGURE B.167    First and second most important reasons for settling in camp for IDPs 100 80 % of households 60 40 20 0 IDP overall Abu Shouk El Salam Man head Woman head First reasons Reasons apart from better security Better security Access to humanitarian aid Better access to assets/services Better access to livelihood opportunities To join family/friends Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018.  FIGURE B.168    Conflict events and displacement dates 500 60 400 50 % of IDPs by camp Number of events 40 300 30 200 20 100 10 0 0 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Violence against civilians Other conflict events Abu Shouk 20 El Salam Source: Authors’ calculations using ACLED (conflict events 2001–2017) and Sudan IDP Profiling Survey, 2018. 261.  IDPs largely lived in the same district before displacement and traveled with groups, although not nec- essarily with their families. Almost all IDPs arrived in the camps in a group, although only about half of them were accompanied by their families (Figure B.169). Only 3 percent traveled alone, which might have been the least safe option. Also, most did not travel very far: 97 percent of IDPs lived in the same state before displacement as they do now, North Darfur (Figure B.170). About three in four even lived within the Al Fashir district. This confirms a general trend of regional displacement: in the 2008 census, more than 95 percent of recorded IDPs identified their current state to coincide with their usual state of residence.171 262.  Most IDPs were displaced twice and arrived in the camp less than a year after displacement. This suggests that the camps offered quick and available shelter. For one in three the camp is the first place of residence after they left their place of origin, and one in two settled once somewhere else before coming to the camp. The time between displacement and arrival in the camp is relatively short (Figure B.171), on average 9 months. At 21 months, it is signifi- cantly longer for those that arrived since 2008 (p < 0.01). It is unclear why it took these recently displaced a longer time to find somewhere to settle. 171. GoS. 2013. “Developing Darfur: A Recovery and Reconstruction Strategy.” 126  |  Informing Durable Solutions for Internal Displacement   FIGURE B.169    Trends in travelling to current location   FIGURE B.170    Places of origin of IDPs, North Darfur state 100 % of population 80 60 40 20 0 k m ad ad 08 8 ou 00 la he he 20 Sh Sa r2 an en re te u El fo M om Ab af be ed W ed riv riv Ar Ar With a larger group but not my family With my family Alone Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018.  FIGURE B.171    Years since displacement and arrival at current location for IDPs 16 14 12 10 Years 8 6 4 2 0 k m ad ad 08 8 ou 00 la he he 20 Sh Sa r2 an en re te u El fo M om Ab af be ed W ed riv riv Ar Ar Years since household was first displaced Years since household members first arrived in camp Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 263.  IDP and host households have equally low numbers of separated household members. Only 1 in 10 house- holds are a separated member across the displaced and non-displaced populations (Table B.17). These are remarkably few given the enduring conflict situation. This again points toward these camps being well established, in the sense that families have had time to reunify where feasible. Notably, the inhabitants of the El Salam camp have especially few separated members, where only 5 percent of households are missing someone, significantly fewer than in Abu Shouk (p < 0.01). Among IDPs, woman-headed households are equally likely to have separated members as their man-headed counterparts but miss slightly more people if they do (p < 0.05). Among the host population, contrastingly, 20 percent of women-headed households are missing a member, much more than for man-headed ones (p < 0.01). Volume B: Country Case Studies  | 127 264.  Almost all households with separated members can contact them. About 9 in 10 have some way of commu- nication, for both IDPs and hosts, which again points toward separation not being among the most urgent problems (Table B.17). The ages of separated members are very diverse and the sexes balanced across all groups, without signifi- cant differences between them in those respects.   TABLE B.17    Trends in separation for IDP and host populations Overall IDP Man Woman Abu El Host IDP head head Shouk Salam % households with separated members 11 10 11 9 13 5 Average number of separated members (for households with 1.58 1.51 1.35 1.72 1.51 1.51 separated members) % separated members who were women or girls 47 47 44 50 49 38 Average age of separated members 32.0 26.0 28.1 24.0 26.1 25.5 % households that can contact separated members 92 92 95 88 93 85 Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 265.  Only one in three IDPs have returned to their place of origin since being displaced, and fewer than half plan to move back. Given the proximity of the places of origin and the long durations of displacement, it is striking that so few have gone back to their places of origin even once (Figure B.172). Half of the IDP households plan to stay in the camp, while two in five plan to move back and less than one in ten plan to move somewhere else (Figure B.173). Given these stated intentions, it becomes clear that many IDPs need to be supported in finding durable solutions within the camp. There are no discernable differences in future movement plans between the two camps or house- holds with man and woman heads.   FIGURE B.172    Trends in having returned since   FIGURE B.173    Trends in return intentions for IDPs replacement for IDPs 70 45 60 40 35 50 % of households % of population 30 40 25 30 20 15 20 10 10 5 0 0 k m ad ad ll k m ad ad ou la a ou he he la er he he Sa Sh Sh Sa ov an en an en El u u El M om Ab P M om Ab ID W W Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Stay in camp Return to origin Move somewhere else Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 128  |  Informing Durable Solutions for Internal Displacement 266.  IDPs who want to stay in the camps are motivated by better security and access to education and health services. The main reasons IDPs do not want to leave the camps are the security situation (96 percent) and the access to education and health services (81 percent, Figure B.174). The latter are more important for the inhabitants of El Salam (p < 0.05) and households with man heads (p < 0.01). About one in two name the access to livelihood and employment as a factor that keeps them in the camps, although there is an indication that this is less true for people in El Salam than for those in Abu Shouk (p < 0.10). Security considerations are also the main reason (94 percent) that keeps IDPs from moving to another place, be it their place of origin or a new destination (Figure B.175). Worse access to education and health services at the potential destination is an argument for half of those who do not want to leave, and again it is more important for those in El Salam (p < 0.05). Access to assets and livelihood opportunities is the third most import- ant reason for not leaving (44 percent), and it is more important for those that want to stay in Abu Shouk (p < 0.01).  FIGURE B.174    Reasons for staying in current location for IDPs who do not want to relocate 100 % of households who want 80 to stay in the camps 60 40 20 0 l k m ad ad al ou la er he he Sa Sh ov an an El u P M om Ab ID W Security Better access to home/land/livestock Better access to education/health services Better access to livelihood/employment Family reasons Access to aid Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018.  FIGURE B.175    Reasons for not moving to a new location for IDPs who do not want to relocate 100 % of households who want 80 to stay in the camps 60 40 20 0 ll k m ad ad a ou la er he he Sa Sh ov an an El u P M om Ab ID W Conflict or insecurity Drought/famine/flood Low access to education/health services Low access to home/land/livestock/employment Family reasons Lose access to aid Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Volume B: Country Case Studies  | 129  FIGURE B.176    Reasons for moving to a new location for IDPs who want to relocate 80 % of households who want 60 to leave the camps 40 20 0 l k m ad ad al ou la er he he Sa Sh ov an an El u P M om Ab ID W Better security Better access to home/housing/land Better access to education/health services Better access to livelihood/employment Family reasons Access to humanitarian aid Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018.  FIGURE B.177    Reasons for moving away from the camp for IDPs who want to relocate 100 % of households who want 80 to leave the camps 60 40 20 0 l k m ad ad al ou la er he he Sa Sh ov an an El u P M om Ab ID W Insecurity Drought/famine/flood Low access to home/land/livestock/employment Low access to education/health services/aid Family reasons Bad management of the site Tensions with host community Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 267.  IDPs who plan to leave the camps cite access to livelihoods, employment, and assets as the most import- ant pull factors. About 68 percent of those who have plans to leave say that access to livelihood opportunities is what makes them want to move to another location (Figure B.176), while 62 percent also mention security. Better access to homes and land is a factor for equally as many, and one in two are motivated by better access to education and health services. All IDPs clearly do not agree as to what is most important in a potential new location. Family reasons are more important for man-headed than for woman-headed households (p < 0.01). Among the reasons that drive people away from the camps, security is relatively unimportant (30 percent), which speaks to the good perceived security in the camps (Figure B.177). The data indicate, however, that security is a more important factor for households with women 130  |  Informing Durable Solutions for Internal Displacement heads (p < 0.1). Generally, 80 percent of IDPs are unhappy with the availability of employment and livelihood opportu- nities or the access to assets, although this is slightly less important for woman-headed households (p < 0.05): access to services such as education, health services, and aid is also an important factor, and among these aid is the most sought after (41 percent).   BOX B.7    The Darfur conflict since 2003 Darfur (‘Land of the Fur’) is the westernmost province of the Republic of Sudan. The area is inhabited by about 80 different ethnic groups. The largest group, the Fur, lives in the highlands of Jebel Marra. Approximately one-third of the population is classified as ‘Arabs’ (and perceive themselves as Arab, that is, non-African). The most vocal are the Rizzeyqat, the Beni Halba, the Ta’aisha, the Habbaniya, the Ja’aliyin, and the Misseriya. Among ethnic groups considered African are the Fur, the Masalit, and the Zaghawa. The self-perceived Arab groups are often cattle and camel nomads, while the African groups are more often agricul- turalists and agro-pastoralists. All groups are Muslim. However, the boundaries of identity have shifted considerably, enabling a multifaceted web of alliances and conflict over the past centuries.172 The rebellion of 2003 was built on feelings of resentment and on political and social marginalization. Since early 2003 the Darfur Liberation Movement (DLM), later renamed the Sudan Liberation Movement (SLM), and the Justice and Equality Movement (JEM) attacked military and government institutions in the Darfur region. While the SLM was largely Fur based and fighting for social justice and inclusion, the JEM was recruited form Arabized Darfuri elites who wanted to end the political mar- ginalization of the region. The GoS responded to the uprising with a military campaign backed by military and paramilitary units. When in 2003 the JEM and SLM began their rebellion against the central government, the GoS revived the Popular Defense Forces (PDF) para- military unity that was sent against the local population to weaken the popular support for the guerrillas.173 Recruited largely from Northern Darfur Arab communities and equipped with military support and wide-ranging authority, these groups also became synonymous with the Darfur conflict between 2003 and 2007, which resulted in the deaths of several hundred thou- sand of Darfur’s ‘African’ inhabitants and the displacement of another approximately 3 million.174 The conflict saw a renewed peak in 2014 and 2015 during a campaign by the Rapid Defense Forces. While first deployed in Kordufan, these PDFs, which are linked to the National Intelligence Service, recently led the counterinsurgency against the rebel movements in Darfur.175 The conflict in Darfur became a showcase for state failure in Sudan and the complexities of nation building beyond North-South divides. The Darfur conflict overshadowed the nation-building initiative in South Sudan as well the recurring conflicts in Blue Nile, Kordufan, and Eastern Sudan in terms of international attention.176 The misinterpretation of the conflict as one between Africans and Arabs led to overlooking the persistent local political structures and the disintegration of the rebel movement. The conflict reached a peak of international attention in 2009 when the International Criminal Court (ICC) issued an arrest warrant for Sudanese President al-Bashir. In 2010 the war was declared over but continues on a lower scale today. 172. de Waal. 2005. “Who Are the Darfurians?” 173. Kindersley. 2018. “Politics, Power and Chiefship in Famine and War: A Study of the Former Northern Bahr El-Ghazal State, South Sudan.” 174. Flint and de Waal. 2008. 175. Human Rights Watch. 2015. “‘Men with No Mercy’ Rapid Support Forces Attacks against Civilians in Darfur, Sudan.” 176. Flint and de Waal. 2008. Volume B: Country Case Studies  | 131 Several consecutive peace agreements were signed, all of which were unable to bring sustainable peace. In 2006, the Abuja Agreement was signed by the SLA faction under Minni Minawi (SLA/MM); however, parts of it rejected the agreement. In 2011, the Doha Document for Peace in Darfur facilitated by the Government of Qatar was signed by the Liberation and Justice Movement, an umbrella organization of rebel factions formed in 2010. The GoS and other factions that did not sign the Doha Accord continue engagement under the auspices of the African Union High-level Implementation Panel (AUHIP). In 2016, GoS, JEM, and SLA/MM signed the ‘Roadmap Agreement’. It is supposed to foster renewed peace talks and humanitarian access. Since the beginning of the conflict in 2003, approximately 3 million people have been displaced in Darfur, and as many as 2.1 million are still in need of humanitarian assistance.177 Despite peace negotiations, and due to the disintegration and splintering of rebel movements, customary conflicts increased, spurred by large-scale proliferation of weapons and the cross-border dynamics with Chad. These dynamics keep the conflict alive, with regular large-scale military intervention by the GoS to break the rebel movements. Hence, between 2014 and 2017 at least 1 million people were displaced anew in Darfur, while at the same time in some areas the conflict decreased and people began to return to their homes.178 Standard of Living 268.  Poverty rates are alarmingly high, particularly among IDPs. The share of people that consume less than the international poverty line of US$1.90 PPP (2011) per day per capita is high across all surveyed groups (Figure B.178). Poverty is highest among IDPs, where four in five are poor, while this is true for three in five hosts on average. Notably, the rate is higher among the people that live in the parts of Al Fashir that are close to the camps, as compared to the rest of the city (p < 0.05). When comparing the camps, the people in El Salam are slightly poorer than the those in Abu Shouk (p < 0.1). 269.  Poverty among IDPs is deeper than it is for the host community. The poverty gap, the average consumption shortfall relative to the poverty line among the poor, is 42 percent for IDPs and 27 percent for the native Al Fashir popu- lation (Figure B.179). In other words, those that consume less than US$1.90 PPP (2011) per day per capita in the camps actually consume only US$1.10 PPP (2011) per day per capita on average. This figure is at US$1.39 PPP (2011) per day per capita for the hosts, so they are on average considerably closer to escaping poverty. 270.  The high poverty incidence may in part be attributed to the recent high inflation and price levels of sta- ple foods, as well as data collection during the lean season. As noted earlier, the first months of 2018 preceding data collection saw a soaring inflation of the Sudanese pound, with inflation rates of over 50 percent. The prices for the most important staple foods, such as millet, sorghum, and wheat, were around 200 percent above their two-year aver- age.179 These basic crops and simple products made from them still constitute on average a third of the consumption of the people in Al Fashir and the IDP camps.180 These recent developments negatively impact purchasing power. Another important factor is that the months of data collection April–July are mostly in the lean season where availability is low and employment opportunities are usually scarce. Depending on price developments and considering the low poverty gaps of the host population, the level of poverty incidence may thus be a temporary lower bound. 177. UN OCHA. 2017a. “Sudan Humanitarian Needs Overview 2017.” 178. Ibid. 179. Famine Early Warning Systems Network. 2018. “SUDAN Price Bulletin July 2018.” 180. In particular, these products are sorghum, millet, wheat, rice, flours of these crops, noodles, bread, and biscuits. 132  |  Informing Durable Solutions for Internal Displacement   FIGURE B.178    Poverty headcount ratio for IDP and  FIGURE B.179    Poverty gap relative to US$1.90 PPP host populations (2011) poverty line for IDP and host populations 100 2 US$ PPP per person % of population 80 27% 1.5 42% 44% 41% 40% 46% 60 per day 40 1 20 0.5 0 IDP Neighboring Al Fashir Remaining Al Fashir Man head Woman head Abu Shouk El Salam 0 IDP Host Man head Woman head Abu Shouk El Salam Overall IDP Overall IDP Mean income Poverty line Below $1.90 per day Below $3.20 per day Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 271.  The gap in wealth between IDPs and hosts cannot be explained by differences in demographic factors or endowments, which indicates generally worse livelihood opportunities for the displaced. IDP households can afford to consume significantly less than host households, even after controlling for demographic factors such as the number of children and women in the household, which tends to be different for IDPs and might therefore explain some of their disadvantage in wealth (Table B.18, specification 1). The gap also persists after controlling for the house- hold head’s years of education, the main economic activity he or she is engaged in, and the socioeconomic networks, which might also accounted for some of the difference (Table B.18, specifications 2, 3, and 4). Even as there likely are omitted variables that cannot be controlled for, such as skills that are not related to formal education or social networks, this strongly indicates that IDPs do not have the same opportunities to put their human capital to use.181 Another find- ing that points in this direction is that, comparing households with their own businesses between IDPs and hosts, there is an indication that owning a business is less lucrative for IDPs than it is for hosts.182 272.  In line with higher levels of poverty, food insecurity is considerably higher among IDPs. About 64 percent of IDPs are highly food insecure, that is, they have to take measures to cope with food shortages more than 10 times per week (Figure B.180).183 In comparison, 33 percent of hosts face high food insecurity. Among the IDPs, the problem is more prevalent in Abu Shouk (p < 0.01) and for woman-headed households (p < 0.05). In particular, the coping strategies adopted include turning to less preferred foods, while 50 percent of IDP household even have to limit adult eating to ensure that children get enough (Figure B.181). These findings are in line with food insecurity having gotten constantly and considerably worse in IDP camps in the Darfur region over the last years, mostly due to weak harvests and rising prices.184 Intriguingly, only 20 percent of households report receipt of food aid by the UN, NGOs, or the government. 181. Potentially omitted variables that could explain the gap include any skills that are not related to formal education, or social networks. Our data do not allow to control for these factors, which must be taken into account when interpreting these results. 182. In particular, the negative coefficient of the interaction between the IDP indicator and the dummy for the household head working on his or her own account can be interpreted in this way. It is significant at the 10 percent level. 183. Food insecurity is defined as an individual facing food shortage at least once in the previous seven days and using a combination of coping strategies to overcome the shortage. It is calculated using the Reduced Coping Strategies Index (rCSI) adapted by WFP/VAM, FAO, and the Global IPC (Integrated Phase Classification) team, among others. rCSI is a weighted index that combines information on frequency and severity of coping strategies used in a single score for household food security. 184. WFP. 2016. “Darfur Food Security Monitoring November 2016”; WFP. 2017. “Darfur Food Security Monitoring May 2017”; WFP. 2017. “Darfur Food Security Monitoring November 2017.” Volume B: Country Case Studies  | 133  TABLE B.18    Determinants of wealth Dependent variable: total (core) consumption per person per day Specification 1 Specification 4 Specification 5 (with basic Specification 2 Specification 3 (with (with HHH: own demographic (with household (with social household head account and IDP Independent variables controls) head education) relations) occupation) interaction) IDP −0.510*** −0.480*** −0.445*** −0.439*** −0.410*** Household size −0.112*** −0.119*** −0.118*** −0.118*** −0.119*** Male children (0–14) −0.029 −0.017 −0.014 −0.013 −0.012 Female children (0–14) −0.045** −0.032 −0.043* −0.042* −0.041* Male youth (15–24) −0.013 −0.006 −0.019 −0.017 −0.016 Female youth (15–24)  0.021  0.020  0.019  0.017  0.018 Female adults (25–64) −0.054** −0.056** −0.068** −0.069** −0.068** Male elderly (>64) −0.046 −0.025  0.039  0.037  0.034 Female elderly (>64)  0.095**  0.074  0.084  0.084  0.086 Female household head −0.083*** −0.009 −0.014  0.008  0.011 Household head age −0.000  0.005  0.001 −0.001 −0.001 Sq. household head age −0.000 −0.000  0.000  0.000  0.000 HHH: years of education  0.023***  0.020***  0.020***  0.021*** Ease of borrowing money  0.131***  0.144***  0.144*** HHH: farmer  0.105  0.106 HHH: wage earner  0.039  0.040 HHH: own business  0.215***  0.284*** −0.107* IDP × HHH: own business R-squared  0.306  0.326  0.343  0.355  0.356 Observations  2,990  2,895  2,474  2,474  2,474 Source: Authors’ calculations. Significance levels: 1 percent (***), 5 percent (**), and 10 percent (*). The estimated model contained a constant, which is omitted here. 134  |  Informing Durable Solutions for Internal Displacement   FIGURE B.180    Food insecurity for IDP and host   FIGURE B.181    Trends in coping strategies when populations hungry for IDP and host populations 100 90 % of population 80 80 60 70 % of households 40 60 20 50 40 0 IDP Host Man head Woman head Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile 30 20 10 0 Less Borrow Limit Restrict Fewer preferred food or portions adult meals Overall IDP foods money eating High food insecurity IDP Host Medium food insecurity Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Low food insecurity Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 273.  The dwellings IDPs occupy in the camps are permanent structures and similar to their houses before the conflict. About 99 percent of IDP households live in tukuls, traditional dwellings with circular mud walls and a roof, or other permanent mud or wood structures (Figure B.182). This constitutes a similar housing situation to their places of origin, and this was the case for 95 percent. These findings underline the permanent nature of the camps; less than 1 percent of households live in tents. There are reports, however, that at least in El Salam camp the tukuls are partly located in flood-prone areas and are likely to be damaged, especially during the rain reason.185 Among the host popu- lation in Al Fashir, 85 percent live in tukuls or other mud/wood structures, while there is also a 15 percent fraction that live in concrete or brick houses, which virtually does not exist in the camps. Even at their original places of residence, merely 2 percent of IDP households lived in structures of concrete or brick. 274.  Overcrowding in the camp houses is a minor shortcoming, less so than among the host population. In 8 percent of IDP households, there are more than four people per room (Figure B.183), which is what the UN defines as having insufficient living space.186 In the host population the problem is worse, where the fraction living in overcrowded conditions is twice as high (p < 0.01). For IDPs, woman-headed households are slightly more likely to live with insufficient living space than man-headed households (p < 0.05). Unsurprisingly, the problem is worse among the poorest quintile. 275.  IDPs largely lived in their own houses before they were displaced and now live in housing provided by the humanitarian community. Before displacement 90 percent of households lived in their own dwellings, and 8 per- cent lived with relatives or friends (Figure B.184). In the camps, 87 percent now live in houses that are provided to them and are thus dependent on the donor organizations. Notably, this contrasts with the dependency on food aid dis- cussed above, which only 20 percent receive. At 9 percent, woman-headed IDP households are more likely to own their dwellings than man-headed ones (p < 0.01), while they were in very similar ownership positions before displacement. Rented housing only plays a role for the host population (19 percent), while 63 percent of them own their dwellings. 185. IOM. 2015. “North Darfur—Al Salam IDP Camp Profile 2015.” 186. UN-Habitat. 2016. “Monitoring Framework, SDG Goal 11.” Volume B: Country Case Studies  | 135  FIGURE B.182    Type of dwelling for IDP and host   FIGURE B.183    Overcrowding: individuals per room, populations IDP and host populations 100 100 % of households % of households 80 80 60 60 40 40 20 20 0 0 Host IDP now IDP at origin Man head Woman head Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile IDP Host Man head Woman head Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile Overall IDP Overall IDP Tent Tukul/gottiya Overcrowded: > 4 Individuals Mud/wood hut Concrete/brick house 2–4 Individuals Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 1–2 Individuals Less than 1 Individual Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018.  FIGURE B.184    Ownership of dwelling for IDP and host populations 100 % of households 80 60 40 20 0 Host IDP current IDP at origin Current At origin Current At origin Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile Man head Woman head Overall IDP Provided by UN/NGOs Squatting Relatives/friends Work-provided Rented Owned Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018 276.  Most IDP households have no access to electricity for lighting, while the host population is largely con- nected to the power grid. In the camps 9 percent of IDP households have access to electricity for lighting, which is still an improvement over the 2 percent that did before displacement (p < 0.01; Figure B.185). Solar panels or biogas also are unavailable (less than 1 percent of households). Instead, most rely on handheld lamps or torches without power con- nections (63 percent) or have no source of lighting (23 percent). Interestingly, the access to electricity for those living in the camps does not improve with wealth. Among the host population 75 percent of host households have access to electricity-powered lighting, yet 10 percent still have no source of lighting at all. Man-headed IDP households are more likely to have direct access to power than woman-headed ones (p < 0.01). 136  |  Informing Durable Solutions for Internal Displacement   FIGURE B.185    Source of lighting for IDP and host   FIGURE B.186    Access to improved water and sanita- populations tion, IDP and host populations 100 100 % of households 80 80 % of households 60 60 40 20 40 0 20 Host IDP current IDP at origin Man head Woman head Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile 0 ID cu ost or t in he d ad Sa uk m ile 2 3 qu Q4 e at en an ea Q Q til ig la nt El ho in P H P rr om h ui S W an tq u st M Ab es ID he or ic Po Overall IDP R Improved source of water No lighting Grass/firewood Improved sanitation facility Lamp/candle/torch Solar power/biogas Electricity Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 277.  IDPs have better water quality and sanitation access than at their places of origin; they are also twice as likely to have access to an improved source of drinking water than the host population. About four in five IDP households have access to an improved source of drinking water, a standard that was met by 56 percent at their places of origin, and which is only the case for two in five host households (Figure B.186).187 The high percentage of the host community that does not meet this criteria can be traced back to 60 percent getting drinking water from mobile transporters such as carts or tanker trucks, which are unlikely to constantly deliver high quality water. IDPs in the camps largely have access to a tube well, a protected dug well, or a public tap, or they drink mostly bottled water (40, 17, 15, and 13 percent, respectively). For both host and IDP populations, 90 percent of households have access to an improved sanitation facility, that is, one that hygienically separates human excreta from human contact.188 This is also a better situ- ation than the IDPs faced before displacement, when only 50 percent of households had access to improved sanitation. 278.  Alarmingly, a large number of IDPs report sharing sanitation facilities. Sharing sanitation facilities may con- siderably affect the quality of water sources and sanitation in the camps. In particular, latrines that are shared by too many individuals pose health risks and discourage their use.189 About 55 percent of IDP households share their sani- tation facilities with other households, which compares to only 17 percent among hosts. Among those IDPs sharing their toilet facilities, 63 percent share them with 10 or more households, which is alarming. There are also reports that each water source in El Salam camp is shared by at least 600 individuals, which may also result in considerable hygiene problems.190 187. Improved water sources are piped water supply into the dwelling, piped water to a yard/plot, a public tap/standpipe, a tube well/borehole, a protected dug well, a protected spring, and rainwater. Unimproved water sources are unprotected dug well, an unprotected spring, a cart with a small tank/drum, a water tanker truck, and surface water. World Health Organization and UNICEF. 2006. “Core Questions on Drinking Water and Sanitation for Household Survey.” 188. Improved sanitation is defined as a household having some type of flush toilet or latrine, or ventilated improved pit or composting toilet (World Health Organization and UNICEF). 189. World Health Organization and UNICEF. 2006. 190. IOM. 2015a. “North Darfur—Al Salam IDP Camp Profile 2015.” Volume B: Country Case Studies  | 137 279.  For both IDPs and host community households, distances from homes to important services are typically short and considerably shorter than what IDPs faced before their displacement. A source of drinking water, a health center, a primary school, and a market all are reachable at half an hour walking time on average, for both IDP and host households (Figure B.187). This contrasts with the places of origin of IDPs where none of these facilities were reachable within less than an hour on average, with health facilities even being almost two hours away. A 2016 study on Abu Shouk also reported that many IDPs find water, health, and education facilities more readily available in the camps than at their places of origin.191  FIGURE B.187    Time (one way) to amenities for IDP and host populations 120 100 80 Minutes 60 40 20 0 Host IDP current IDP at origin Host IDP current IDP at origin Host IDP current IDP at origin Host IDP current IDP at origin Water Health facility School Market Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018 280.  Literacy rates among IDPs are similar to those of the host population and lower for IDP women and peo- ple in households with a woman head. For both the camp population and those living in Al Fashir city, 70 percent of the people above 15 years of age can read and write (Figure B.188). The sex-based discrepancy in the camps is considerable: while 78 percent of men can read and write, only 62 percent of the women can. Also, people living in a woman-headed household are less likely to be literate as compared to where a man is the household head (p < 0.01). Interestingly, literacy does not increase with wealth in the IDP camps. 281.  Primary and secondary school attendance rates are substantially lower for IDPs, while there are no dis- cernible sex-based differences. For IDP children of primary school age only 75 percent attend school, while this rate is 84 percent for the host children (Figure B.189). About 62 percent of host children of secondary school age also attend school, while only 50 percent of IDP children in that age bracket go to school. Secondary enrollment rates are particu- larly low among IDP children from woman-headed households (p < 0.01). There is, however, no discernible difference in attendance rates between boys and girls in any of these subpopulations (not pictured). Overall, all these attendance rates must be considered high; however, given the displacement and conflict context. In addition, it must be noted that they may slightly underestimate the true figures. This is because data were collected during summer holidays, and there were reports of faulty response to attendance questions because of that. 191. Ali and Mahmoud. 2016. “From a Temporary Emergency Shelter to an Urbanized Neighborhood: The Abu Shoak IDP Camp in North Dārfūr.” 138  |  Informing Durable Solutions for Internal Displacement  FIGURE B.188    Literacy rates, 15 years and above for   FIGURE B.189    Net attendance rates, primary and IDP and host populations secondary school for IDP and host populations 100 100 in the age groups % of population 80 80 % of population > 15 years old 60 60 40 40 20 20 0 IDP Host Man head Woman head Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile 0 Men Women Total Men Women Total Man head Woman head Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile Overall IDP Host IDP Primary school attendance among primary school aged Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Secondary school attendance among secondary school aged Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 282.  Accordingly, IDP adults are also less educated than the host population, while women attain particularly little education. In the camps, 33 percent of people above the age of 18 have not completed any level of schooling, which is only true for 20 percent of hosts (p < 0.01; Figure B.190). At the other end of the education spectrum, an impressive 27 percent of host adults have completed a university degree, while only 13 percent of IDPs have that privi- lege. The sex-based discrepancy is also large: 42 percent of IDP women have not completed any education level, which is double as the percentage of IDP men. The sex-based difference also exists among hosts, however, where 26 percent of women and 15 percent of men have not completed any schooling. Again, IDPs living in a household headed by a woman are underprivileged and have a higher chance of not completing any education than those in a man-headed household (p < 0.01). Interestingly, among IDPs the proportion without any education does not decrease with wealth. The richer are, however, less likely to end up with only primary education (p < 0.01 for the richest vs. the poorest quin- tile) and more likely to hold a university degree (p < 0.05).  FIGURE B.190    Adult educational attainment for IDP and host populations 100 % of population >18 years old 80 60 40 20 0 Men Women Total Men Women Total Man head Woman head Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile Host IDP No education Primary & intermediate Secondary University Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Volume B: Country Case Studies  | 139 Employment and Livelihoods 283.  Employment levels are similar for adult IDPs and hosts. For both the displaced and the host population, the clear majority of those above 24 years are employed in some way (69 percent and 63 percent, respectively; Fig- ure B.191). The unemployment rate, which refers to not working but actively looking for employment, is negligible among adults. This is common in high-poverty contexts since those who must work to support their needs find some- thing to do, if necessary, in the informal labor market. Rather, 20 percent of working age are inactive, without ambition to find employment, across IDP and host populations. 284.  IDP women are more likely to work than host women. Only one in three women in the host population and one in two of the IDP women are working. Women in the host and IDP population are thus 51 percent and 22 percent less likely to be working than men, respectively. However, labor force participation among women is significantly lower than among men. About 3 in 10 women are neither working, nor unemployed, nor in education. A higher propensity to work among IDP women may partly be linked to women’s centers, which exist in both camps to support women in their pursuit to find employment.192 285.  Compared to youth in the host population, IDP youth are more likely to be working and less likely to be in education. About 44 percent of the IDPs between the ages of 15 and 25 are working and 37 percent are in education (Figure B.191). Among youth in the host population, on the other hand, only 25 percent are working while 55 percent are in education.  FIGURE B.191    Labor force participation and employment status among working-age population 100 % of population within working age (15–65) 80 60 40 20 0 Men Women Total Men Women Total Men Women Total Men Women Total Youth Adults Youth Adults Host IDP Not in labor force, not enrolled Not in labor force, enrolled Not in labor force, unemployed Not in labor force, employed Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018 286.  Most working IDPs were own-account agricultural workers before their displacement; today a majority are salaried employees. Before their displacement, 54 percent of the IDPs that were working at the time were in agri- culture while 35 percent had salaried jobs (Figure B.192). This is in stark contrast to their current situation in which only 5 percent are own-account agricultural workers, and a majority have wage employment or work in a non-farm business owned by the household (56 percent and 26 percent, respectively). While host households are more likely to own a 192. IOM. 2016. “Abu Shouk IDP Camp Celebrates the Official Opening of Its First Women’s Centre”; IOM. 2015a. “North Darfur—Al Salam IDP Camp Profile 2015.” 140  |  Informing Durable Solutions for Internal Displacement business (p < 0.01), the distribution of type of workers is similar: 51 percent and 16 percent are salaried employees or work in a non-farm business owned by the household, respectively. In general, the quality and sustainability of work is difficult to assess. A study conducted in 2008 describes many of the newly pursued livelihood strategies by IDPs in Darfur as maladaptive and unsustainable.193 287.  Better-off IDPs are more likely to run a family business and less likely to be wage workers. Only 50 per- cent of the working IDPs in the highest quintile are wage workers, compared to more than 60 percent in the lowest quintile of the consumption distribution. Instead, better-off IDPs are more likely to be running a family business or to be working in one. By itself, this would suggest that self-employment can improve livelihoods at least for some of the IDP households. However, the ratio of those running a family business to those helping in one is much lower among IDPs, suggesting that some of these might be used to provide employment of last resort. IDP women, in particular, are much more likely than men to help in the family business (p < 0.01), while men are more likely to own one (p < 0.01).  FIGURE B.192    Primary employment activity for IDP and host populations 100 % of employed 80 population 60 40 20 0 Host IDP current IDP at origin Current At origin Current At origin Man head Woman head Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile Men Women Overall IDP Salaried labor Run non- farm business Help in non-farm business Own account agriculture Apprenticeship/training Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 288.  IDPs were mostly engaged in agricultural activities at their places of origin, while their current sources of livelihood are more diverse. About 95 percent of IDP households relied on agriculture as their main source of livelihood before displacement, while only 50 percent do now. About 35 percent now live off wages and 10 percent off the income of the family/household business (Figure B.193). This is still a larger dependency on agriculture than among the host community, where only 15 percent support themselves mainly through farming. A clear majority there lives off wages and income from a family/household owned business (56 percent and 26 percent, respectively). The share of households for which remittances constitute the main source of livelihood is only about 2 percent. 289.  Aid only plays a marginal role in the livelihoods of IDPs and appears to be difficult to access. Self- dependence among IDPs is high; a mere 6 percent of the camp households rely mainly on aid and 2 percent on remit- tances (Figure B.193). Both for aid by the government and NGOs, IDPs state that the most important reason for not 193. Young and Jacobsen. 2013. “No Way Back?” explores newly adapted livelihood strategies of camp and non-camp IDPs in two localities in West and North Darfur. The authors conclude that the new activities are often dangerous, harming the environment, and/or illegal, thereby increasing vulnerabilities by using up limited natural resources and stirring up potential further conflict. Volume B: Country Case Studies  | 141 obtaining aid is their lack of information (47 percent and 45 percent of households, respectively, not pictured). One- third also says that aid is captured by other, more powerful people. Given the high poverty rates, this indicates that aid is generally hard to come by in the camps. Most IDPs somehow must make a living by themselves.  FIGURE B.193    Main sources of livelihoods for IDP and host populations 100 % of households 80 60 40 20 0 Host IDP current IDP at origin Man head Woman head Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile Overall IDP Agriculture Wages and salaries Owned business enterprise Aid Remittances Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 290.  Few IDP households own productive assets. Before displacement, IDPs had much more agricultural land, live- stock, or other productive assets than they do now: 90 percent of IDP households owned livestock at their places of origin, 86 percent had access to agricultural land, and 31 percent owned other productive assets (Figure B.194).194 This compares to only 31 percent having access to agricultural land, 21 percent owning farm animals in the camps, and a mere 5 percent owning productive assets. The share of households in Al Fashir that own farmland and livestock is similarly low. Hence, reduced reliance on agricultural assets tends to be the norm in an urban environment such as Al Fashir. However, 37 percent of host households own assets that can be employed productively such as cars, plows, or computers.  FIGURE B.194    Access to productive assets for IDP and host populations 100 % of households 80 60 40 20 0 Host IDP current IDP at origin Man head Woman head Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile Overall IDP Agricultural land Livestock Productive assets Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 194. Productive assets entail cars, trucks, motorcycles, rickshaws, plows, computers/laptops, and refrigerators. 142  |  Informing Durable Solutions for Internal Displacement Social Cohesion, Public Participation, Safety Perceptions 291.  Relations between the IDP and their host community are mostly perceived as good or very good on both sides. 88 percent of the host population evaluates the relationship to be good or very good, an assessment that is shared by 79 percent of the IDPs (Figure B.195). However, there are also 11 percent of the displaced who see the rela- tionship to be bad or very bad, which compares to only 5 percent among hosts (p < 0.01). It is worth noting that IDPs in Abu Shouk tend to be warier of their relationship with the host community (p < 0.10).  FIGURE B.195    Perceptions of relationships between host and IDP communities for IDP and host populations 100 80 % of population 60 40 20 0 IDP Host Man head Woman head Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile Overall IDP Very good Good Neither good nor bad Bad Very bad Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 292.  Almost 90 percent of households have never attended a public meeting where community affairs were discussed. The share of households that has attended at least one public meeting in which community affairs (for example, division of water or land resources) are discussed is only 13 percent among IDPs and 12 percent among the host community (Figure B.196). Notably, in El Salam, the proportion that has never attended such a meeting is even higher than in Abu Shouk (p < 0.05). Likewise, the share of households that has never approached or interacted with a community leader to discuss their problems or needs is 85 percent among IDPs and 89 percent among hosts (Fig- ure B.197). Only 5 percent and 9 percent of IDP and host households have a family member that participates in a com- munity, social, or political organization, respectively. Low rates of participation in public affairs, while surely undesirable, seem to be the norm in Al Fashir. 293.  IDPs feel considerably less safe in their neighborhoods than hosts. By day, a clear majority feel at least mod- erately safe walking around, both in and out of the camps (Figure B.198). However, after dark 87 percent of hosts still feel safe walking outside while this is only true for 60 percent of IDPs. About 17 percent feel very unsafe, something which only 4 percent of hosts share. There are no discernable differences in perceived safety between the camps or across wealth levels. As to the types of threat, the most prominent danger is being robbed: 17 percent of IDP households have a member who has been robbed over the course of the last 12 months. Physical abuse and beatings are rare, however, having affected only 2 percent. Volume B: Country Case Studies  | 143   FIGURE B.196    Frequency of attending a public   FIGURE B.197    Frequency of interacting with a com- meeting for IDP and host populations munity leader for IDP and host populations 100 100 % of households % of households 80 80 60 60 40 40 20 20 0 0 IDP Host Man head Woman head Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile IDP Host Man head Woman head Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile Overall IDP Overall IDP Never 1 to 4 times More than 4 times Never 1 to 4 times More than 4 times Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018.  FIGURE B.198    Perceptions of safety, walking in the neighborhood for IDP and host populations 100 % of population 80 60 40 20 0 By night By day By night By day By night By day By night By day By night By day By night By day By night By day By night By day By night By day By night By day By night By day IDP Host Man Woman Abu El Salam Poorest Q2 Q3 Q4 Richest head head Shouk quintile quintile Overall IDP Very safe Moderately safe Neither safe nor unsafe Moderately unsafe Very unsafe Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 294.  IDPs are less likely to have legal identification documents than hosts, mostly because obtaining them is too expensive.195 Only 43 percent of IDPs have some sort of legal identity documentation, which compares to 58 per- cent among hosts (Figure B.199). The poorest IDPs are less likely to have documentation than the richest (p < 0.01). In accordance with this finding, the most important reason for the lack is that people cannot afford the money to obtain the documents (Figure B.200). This is also a more severe problem among IDPs (p < 0.01), who as described above are poorer than hosts. 195. Legal identification documents include birth certificates, nationality certificates, and passports. 144  |  Informing Durable Solutions for Internal Displacement   FIGURE B.199    Ownership of legal identification   FIGURE B.200    Reasons for not having legal identifica- for IDP and host populations tion for IDP and host populations 80 100 % of population without 70 80 % of population 60 documentation 50 60 40 30 40 20 10 20 0 0 IDP Host Man head Woman head Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile IDP Host Man head Woman head Abu Shouk El Salam Poorest quintile Q2 Q3 Q4 Richest quintile Overall IDP Overall IDP Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Lack of money Administrative challenges Lost during displacement Documentation is unimportant Other Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Targeting Analysis 295.  IDPs have a larger share of support-dependent and productive but poor households compared to host communities, and the profile of IDPs is similar across camps. About 9 percent of IDP households are support- dependent, 70 percent productive but poor, and 21 percent self-reliant. IDPs are more likely to be support-dependent and productive but poor than households living in host communities, who are more than twice as likely to be self- reliant However, the profile of IDPs is similar across camps, with a slightly larger share of self-reliant households in Abu Shouk camp (Figure B.201; Figure B.202).  FIGURE B.201    Vulnerable population by status of  FIGURE B.202    Vulnerable IDP population by camp the household 100 100 80 % of households 80 % of households 60 60 40 40 20 20 0 0 El Salam Abu Shouk Host community IDP Self-reliant Self-reliant Productive but poor Productive but poor Support-dependent Support-dependent Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Volume B: Country Case Studies  | 145 Typology of IDPs 296.  The typologies of IDPs identified two groups in Sudan with different profiles. Group 1 represents 39 per- cent of the IDP population and Group 2 represents 61 percent (Figure B.203; for details about the methodology see Volume C). Before displacement, the two groups had different sources of income, and households from Group 1 were more likely to be displaced in 2003 and 2004. Households from Group 1 are also more likely to live in shelter provided in the camp, and thus are closer to services and more likely to have improved water sources. There are also differences in the current conditions of the two groups. Group 1 has a higher and deeper poverty incidence, and is more likely to face food insecurity and to rely on assistance from development partners or NGOs. As for the future, most households in Group 2, who prefer to stay in the camp, are guided by security reasons, while a majority of IDPs in Group 1 want to relocate based on employment conditions and other considerations. The timeline for moving is clearer for IDPs in Group 2, and they are more likely to have all the information they need to inform their decision.  FIGURE B.203    Visualization of groups from the clustering analysis Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Note: Group 1 is represented by the black triangles and Group 2 by the blue circles. Cause Profile 297.  Group 1 and 2 come from the same districts, but households from Group 1 were more likely to be dis- placed by the Darfur conflict in 2003 and 2004. Both groups of IDPs have the same profile in terms of district of ori- gin, around 74 percent come from Al Fashir, 15 percent from Kebkabiya, and the rest from other districts (Figure B.204). Most households from both groups stayed in North Darfur, indicating the need for finding a camp close to their place of origin and the risks associated with prolonged travel. Households in Group 1 were more likely to be displaced during the peak of the Darfur conflict in 2003–2004 (74 percent vs. 65 percent from Group 2), and less likely to cite insecurity as the reason for displacement (77 percent vs. 90 percent of households from Group 2, Figure B.205). 298.  The two groups had different sources of income before displacement, with Group 1 relying more on agriculture. Most IDPs had access to agricultural land and an agricultural livelihood before displacement.196 However, 196. Having access to land does not necessarily imply ownership. 146  |  Informing Durable Solutions for Internal Displacement a larger share of households from Group 1 had access to agricultural land (92 percent vs. 87 percent from Group 2) and thus were more likely to rely on agriculture. Nearly 90 percent of households in Group 1 had an agricultural livelihood before displacement, compared to 80 percent from Group 1 (Appendix F). The difference in livelihood before displace- ment between both groups is robust after controlling for region effect and household characteristics (Appendix F).   FIGURE B.204    District of origin  FIGURE B.205    Year and reason of displacement 100 100 90 80 80 % of households 70 % of households 60 60 50 40 40 20 30 20 0 10 Group 1 Group 2 0 Al Fashir Kebkabiya Displaced in Relocation due Kutum Other 2003–2004 to insecurity Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Group 1 Group 2 Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Needs Profile 299.  The household composition of both groups is similar, yet Group 1 households are more likely to be headed by economically active and literate men. Households in both groups have the same number of members, proportion of children in the household, and age-dependency ratio (Appendix F).197 Households from Group 1 are less likely to be headed by women (42 percent vs. 56 percent for Group 2), and their heads more likely to be literate (65 per- cent vs. 55 percent for Group 2). Moreover, the share of inactive household heads is larger among Group 2 (25 percent), compared to Group 1 (14 percent). The different profile of the household head between both groups is significant after controlling for other household characteristics and regional effects (Appendix F). 300.  Households from Group 1 tend to live in shelters provided in the camp, and thus are closer to services and more likely to have improved water sources. Only 80 percent of households from Group 2 were provided with shelter, against most of households in Group 1 (94 percent). Group 2 is also more likely to be living in worse conditions and without support from development partners. This group of households is more likely to be located far from ser- vices, and to not have electricity and improved water sources (Appendix F). As a result, Group 2 is more likely to live in an overcrowded dwelling.198 197. The age dependency ratio is defined as the proportion of children and old age dependents to working-age population (15–64). 198. Overcrowded dwellings are defined as those that have four or more people per room. Volume B: Country Case Studies  | 147 301.  IDPs in Group 2 have more heterogeneous income sources, while Group 1 largely relies on assistance from development partners or NGOs. The main source of livelihood for households in Group 2 is agriculture (34 per- cent) and wages or own business enterprise (35 percent, Figure B.206). Contrary to this, the main source of livelihood for IDPs in Group 1 is other sources of income (45 percent), and they are more likely to receive assistance from devel- opment partners or NGOs (27 percent), compared to IDPs in Group 2 (17 percent). The differences in sources of income between both groups are robust after controlling for regional effects and other household characteristics (Appendix F). IDPs in Group 1 seem to have greater access to safety nets and networks as reflected in a larger share of household receiving both assistance and shelter.   FIGURE B.206    Current source of livelihood 60 50 % of households 40 30 20 10 0 Received Livelihood: Livelihood: Livelihood: assistance from agriculture wages or own other UN/NGOs business Group 1 Group 2 Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018   FIGURE B.207    Current poverty and food insecurity status 100 90 80 70 60 50 40 30 20 10 0 Poverty Poverty gap High food Medium food Low food incidence (% (% of the insecurity insecurity insecurity of population) poverty line) (% of HHs) (% of HHs) (% of HHs) Group 1 Group 2 Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 148  |  Informing Durable Solutions for Internal Displacement 302.  IDPs in Group 1 are more vulnerable as they have a higher and deeper poverty incidence and are more likely to face food insecurity. About 8 out of 10 IDPs live below the standard international monetary poverty line.199 Poverty incidence is 5 percentage points higher among households from Group 1, and the poor in this group are 5 percentage points further from the poverty line compared to IDPs from Group 2. The differences in poverty incidence are significant after controlling for region effects and characteristics of the household head (Appendix F). In addition, 67 percent of households from Group 1 were identified with high risk of being food insecure and only 15 percent with low risk, compared to 57 percent and 29 percent of households from Group 2, respectively (Figure B.207). Compared to Group 2, households from Group 1 are possibly at a disadvantage in accessing income-generating activities, and thus are poorer and more likely to rely on aid.   FIGURE B.208    Perception of current conditions 100 90 80 70 % of households 60 50 40 30 20 10 0 Satisfied with health clinic Good relations with & employment opportunities neighbors & the community Group 1 Group 2 Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 303.  Both groups have a good perception of their surroundings, but Group 2 is more satisfied with their cur- rent living conditions. The general perception of both groups (around 99 percent in each) is that they have a good relationship with other IDPs in the camp and also with the community around them (Figure B.208). In terms of the perception of the current situation, households from Group 2 are twice as likely to be satisfied with health conditions and employment opportunities compared to IDPs in Group 1 (24 percent vs. 12 percent). The difference in perception is possibly associated with a lower poverty incidence and less dependency on aid for households in Group 2. Solutions Profile 304.  Most households in Group 2 prefer to stay in the camp due to security reasons, while a majority of IDPs in Group 1 want to relocate based on employment conditions and other considerations. The return intention of households in both groups is different. More than 70 percent of households in Group 2 want to stay in their current set- tlement, compared to less than 20 percent of IDPs from Group 1. Among IDPs with intentions to relocate, households are more likely to report the place of origin compared to moving to another location. About 70 percent of households from Group 1 would like to return to their district of origin and only a small fraction (14 percent) would prefer to move to another location (Figure B.209). Employment conditions and other considerations are more important for IDPs in 199. The poverty line corresponds to a daily value of US$1.90 PPP per day. Volume B: Country Case Studies  | 149 Group 1 than for IDPs in Group 2. In contrast, 9 out of 10 households in Group 2 reported security as the primary factor guiding a decision to stay or move, against less than 5 out of 10 IDPs in Group 1 (Figure B.210). A desire to relocate seems to be associated with worse living conditions and more dependency on assistance.  FIGURE B.209    Return intention 80 70 % of households 60 50 40 30 20 10 0 Stay in camp Return to origin Move somewhere else Group 1 Group 2 Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018.   FIGURE B.210    Reasons for moving or staying   FIGURE B.211    Timing of moving 100 100 80 80 % of households % of households 60 60 40 40 20 20 0 0 Don't know In less than In more than Security Employment Other 12 months 12 months Group 1 Group 2 Group 1 Group 2 Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. Source: Authors’ calculations using Sudan IDP Profiling Survey, 2018. 305.  The timeline for moving is clearer for IDPs in Group 2, and they are more likely to have all the information they need to inform their decision compared to households in Group 1. Among those households looking to relo- cate, the timing is more uncertain for IDPs in Group 1, with 93 percent of the households not having a clear timeline for moving, compared to only 61 percent of those from Group 2 (Figure B.211). Even though IDPs in Group 1 are surer they would like to relocate, this group is more likely to perceive having an incomplete amount of information (55 percent) against households in Group 2 (45 percent). Besides, Group 2 is more likely to receive information from radio, TV, or the Internet (42 percent vs. 29 percent, Appendix F). The information IDPs need could be related to the source used and type 150  |  Informing Durable Solutions for Internal Displacement of information required. Group 1 might be at a disadvantage since they seek information on employment and other conditions, which might be harder to get, and they seem to rely on informal sources, unlike households in Group 2. Policy Implications of IDP Typology 306.  The two groups of IDPs can be differentiated based on their displacement year, location, and return inten- tions. Households in Group 1 were more likely to be displaced during the peak of the Darfur conflict in 2003–2004 and are less likely to be headed by women (42 percent vs. 56 percent for Group 2). In addition, most households in Group 1 were provided with shelter in their respective settlement, compared to 80 percent of households from Group 2. These observable differences could prove useful when targeting policy responses to the different groups of IDPs in Sudan. 307.  Supporting IDPs in Group 1 implies improving their skills to help them diversify their sources of income. Before displacement, households from Group 1 were more likely to have an agricultural livelihood and are currently poorer and more likely to rely on assistance for livelihood and dwelling. Skills and human capital are especially perti- nent for this group due to the loss of physical capital during displacement. In addition, their decision to stay or move is guided by employment and other considerations. Thus, policy efforts for Group 1 should aim to upgrade their skills to avoid a dependency on assistance and support an income diversification strategy to improve their living standards. 308.  Households in Group 2 require gender-responsive programs, increasing their access to safety nets and better living conditions. More than half of the households in Group 2 are headed by women. Policy efforts aimed at Group 2 should consider gender-based vulnerabilities related to domestic work and caring labor, in addition to GBV and discrimination in income-generating activities. Group 2 also seems to have less access to safety nets, perhaps because most of them are located far from main services. Assistance from development partners and NGOs should aim to reach remote areas and target households headed by women. Moreover, programs dedicated to improving living conditions of IDPs should concentrate on households from Group 2 as they are less likely to have improved water sources and electricity, which are critical for health, school performance, and productivity. 309.  The two types of IDPs have different information needs and durable solutions. Most households in Group 1 want to relocate based on employment and other considerations, but do not have a clear timeline, possibly because they need more information to firm their plans. To settle anew, IDPs in Group 1 require information about employment prospects in other districts from reliable sources, as well as conditions for them to re-engage successfully in productive activities. Households in Group 2 want to stay in their current settlement, and many have all the information they need to make this decision. Overall, IDPs perceive a good relationship with their surroundings, but better conditions are required to facilitate the integration of households from Group 2 in their communities. Also necessary improvements in living conditions and access to services, such as improved water sources and electricity, which ultimately are key in bringing about a durable solution to their displacement. Conclusions Informing Durable Solutions 310.  Despite the long times since displacement, IDPs in the camps by Al Fashir are poorer, less educated, face higher food insecurity, have less access to productive assets, feel less safe, and have less identification docu- ments than the host population. About 80 percent of IDPs live below the international poverty line of US$1.90 PPP (2011) per day per capita, 60 percent face high food insecurity, 60 percent adults have no or only primary education, 25 percent households have access to a productive asset, 35 percent feel unsafe in the camps at night, and 60 percent have no legal identification documents. Volume B: Country Case Studies  | 151 311.  Poverty and food security are among the most urgent problems and limit the scope of action for IDPs. While poverty is also high among the host populations, it is worse for IDPs. About 80 percent of IDPs live below the international poverty line, and most are further below it than hosts. In addition, food insecurity is also more common among IDPs. Not only does this clearly remove the present situation from being a potential durable solution, poverty and hunger are also associated with limited freedom of decision making and possibility of movement. For instance, IDPs are less likely to have legal identification documents, and the biggest obstacle to obtaining them are cost. This limits their ability to claim property rights at the places of origin as well as travel possibilities. 312.  The IDP camps are permanent settlements and perceived as such, where relations to the host communi- ties are good. Housing in the camps is largely in permanent traditional houses where overcrowding is low and IDPs are mostly working for their main sources of income. About half of the IDPs have no plans to move away. They have equally few separated family members as the host community, with whom they can largely communicate, and do not want to leave the camps due to the access to education and health services. A clear majority describe relations with the host community as good, a view that is also shared by the latter. Therefore, the camps do have the potential to become a permanent residence for many. 313.  IDPs and the host communities are in similar situations in many respects, which reflects the hardship that the non-displaced endure. Given the sustained conflict in the region, the situation of the urban Al Fashir population is also dire, which undermines their capabilities to host the displaced. Host community members are more likely to live in overcrowded houses than the people in the camps, over 40 percent of adults have only primary school or no education, and poverty is also high in Al Fashir city. In addition, there is at least one dependent household member per working-age adult Thus, host communities need policy interventions focused on improving their living conditions and enhancing hosts access to services and income generating activities. 314.  The urban environment contributes to the drastic change in living conditions for most IDPs, which entails loss of human capital. Before their displacement, IDPs were mostly own-account agricultural workers who, in search of better security, found themselves in an urban environment in the camps. The new surroundings brought about a drastic change in living and working conditions and proximity to services. Since most Sudanese IDPs were farmers before displacement, many had to change their professions because of the lack of available land and livestock. This also means a drastic loss of human capital. Yet the urban environment also means that they are now much closer to services such as water sources, schools, markets, and health facilities. 315.  While IDPs in Sudan largely generate their own income and live in safer conditions, they need durable solutions that foster sustainable income-generating opportunities to escape poverty. IDPs in the camps report living in safer conditions and not relying on monetary or food aid since they largely generate their own income. How- ever, poverty levels are alarmingly high. While the high food insecurity must be overcome in the immediate term, a crucial intervention should focus on the creation of employment, income-generating opportunities and access to productive assets. This could improve wealth without fostering dependencies. 316.  Almost all IDPs are protracted cases and many want to stay in the camps, which supports further invest- ment in camp infrastructure and their integration into the city. Half of the IDPs want to stay in the camps, while most of the others want to return to their places of origin, which given their long time of absence and the conflict 152  |  Informing Durable Solutions for Internal Displacement situation, might be problematic or impossible. This foreseeable stability provides strong arguments for investing further in the camp infrastructure. While housing is already at a relatively good standard, IDP households still need access to electricity in their homes. Most hosts are connected to the city power grid. Also, more water sources and sanitation facilities would reduce are needed. 317.  The low participation of IDPs in community decision making is worrying, yet it is no different among hosts. About 9 in 10 households have never attended a public meeting to discuss community affairs, and similarly few have interacted with a community leader or participated in organizations for communal or political matters. Since this is true also among hosts, it is unclear whether, or in what way, this needs addressing. Refugees from Somalia, South Sudan, and Sudan in Ethiopia Introduction 318.  The analysis on refugee groups in Ethiopia complements the IDP results in Somalia, South Sudan, and Sudan. Along with data on IDPs, a survey in Ethiopia represents refugees living in camps from four countries, including Eritreans, Somalia, South Sudan, and Sudan (Box B.8). While IDP profiles draw detailed insights on the condition and policy responses for internal displacement situations, refugee profiles provide an implicit comparison to the needs of a different type of forcibly displaced group. While IDPs and refugees are often driven due to similar factors (largely related to conflict in the countries studied in this report), they can come from different regions in the origin country, face dif- ferent needs in the current location, and might ultimately settle in different durable solutions. This chapter provides a profile of refugees and proposes policy responses with a focus on the refugees’ current country of refuge, Ethiopia. 319.  The Ethiopia comparative study addresses multiple dimensions of poverty of refugees and hosts. The analysis is based on household surveys conducted with the four main refugee groups in the country (Eritreans, Somalis, South Sudanese, and Sudanese) and with members of host communities. The emphasis on both refugees and host communities acknowledges the mutual—and in some cases, interdependent—development needs of both groups. The Ethiopia case study makes use of original micro-level data on (a) a comprehensive set of social and economic indicators that assess poverty incidence, standard of living, and livelihood sources, among others; and (b) perceptions on the refugee-host community relationship, and the future intentions and prospects of refugees. It analyzes and com- pares the different context and situation of the four refugee groups and the regions that host them. 320.  Ethiopia has been suffering from multiple refugee crises (some more protracted, some more recent) that put a strain on the coping capacity of national and local authorities. Contextual factors and dynamics in the four main regions that host refugees are unique as they are geographically and ethnically distinct, although they are all peripheral and relatively underserved areas. The four refugee groups are also remarkably diverse. They have been dis- placed for different reasons and at different times, and are integrated to different degrees within Ethiopian economy and host communities. The tremendous surge in refugees in the country in the last decade from 85,000 in 2007 to almost 1 million refugees in early 2017—coupled with the more recent spike in the number of IDPs—makes the present analysis timely.200 In addition, the Government of Ethiopia (GoE) is currently designing and implementing policies and 200. UNHCR Statistics. 2018. http://popstats.unhcr.org/en/overview (Accessed on April 23, 2018). Volume B: Country Case Studies  | 153 programs following the GoE 2016 strategicinitiative (‘nine pledges’), which is aimed at expanding refugee rights and service delivery, and comprehensively addressing socioeconomic limitations of both refugees and host populations.   BOX B.8    Skills Profile Survey (SPS) of refugees and host communities Household surveys were conducted with refugees from South Sudan, Somalia, Eritrea, and Sudan living in camps in Ethiopia, and with host community members within a 5 km radius of a camp. IDPs were not included and the survey is only representative of refugees living in camps—who, nonetheless, are the majority. A sample frame was the list of all refugee camps in the five main regions that host refugees: Tigray and Afar (hosting mostly Eritreans), Gambella (South Sudanese), Benishangul Gumuz (Sudanese), and Somali (Somalis). Because each region hosts a predominant majority of one refugee nationality, the strat- ification of the sample is practically based on nationality. A total of 900 refugee households and 500 host community households per region were to be surveyed, and all the refugee camps in the sample frame are to be included. A volatile security situation in the country imposed some changes during fieldwork. In Gambella, host community house- holds were not surveyed, and only 439 of the intended 900 refugee households were surveyed there. The remaining interviews with South Sudanese refugees in Gambella were substituted by oversampling in Benishangul Gumuze, as 25 percent of the refu- gee population in this region is South Sudanese. Similarly, the escalation of violent conflict in Oromia and Somali rendered some of the camps in Somali inaccessible. The result is that camps in peaceful areas were oversampled. In addition, due to a sparse host population in Somali, the final number of host households surveyed in this region was 303 against the intended sample of 500. Despite the changes, the survey captured roughly a similar number of refugee households of the four main refugee nationalities. 321.  In the last decade and a half, Ethiopia has achieved tremendous progress in terms of growth and pov- erty reduction.201 Between 2004 and 2014, the economy recorded an annual average growth rate of 10.9 percent. The percentage of people living below the international poverty line of US$1.25 PPP per day (pre-2011) decreased substantially from 56 percent to 30 percent in 2000–2011. Remarkably, poverty reduction has been more pronounced in poorer regions, some of which host refugees (including Tigray and Benishangul Gumuz, among others), leading to some degree of spatial convergence in the overall level of economic development in the country. In addition, through- out this progress, Ethiopia largely maintained its traditional low levels of inequality. Ethiopian households improved their material condition, living standard, and access to services on several economic and human development indica- tors, including health, education, nutrition, and livelihoods. Poverty eradication has been driven by agricultural growth, accompanied by state-led expansion of services, and effective safety nets for the poorest. Nonetheless, Ethiopia’s econ- omy has been diversifying: between 2004 and 2014, the agricultural share within the economy fell from 52 percent to 40 percent, while the service sector rose from 37 percent to 46 percent. Conflict and Forced Displacement: Dynamics, Scale, and Profile 322.  Regionally, there are three main theaters of crises that have caused refugees to pour into Ethiopia over several decades (Figure B.212). First, conflict and instability in Somalia are examples of how interstate dynamics con- flate with domestic unrest in Ethiopia, and clan-based conflicts and high levels of violence in Somalia. Since the 1960s, Ethiopia has been engulfed in a border conflict with Somalia over the Ogaden region of Ethiopia (or ‘Somali National Regional State’, as per official name),202 including a full war being fought in 1977–78. The Somali region of Ethiopia 201. World Bank 2015. 202. In 1995, the Ethiopia’s state turned into an ethnic-based regional system by proclaiming a federal state composed of nine regional states and two federally administered cities (Addis Ababa and Dire Dawa). Afar, Amhara, Oromia, Somali, and Tigray are single ethnic states, while Benishangul Gumuz, Gambella, Harar, and Southern Nations, Nationalities and People’s Region (SNNPR) are multiethnic states. (International Crisis Group 2009). 154  |  Informing Durable Solutions for Internal Displacement has also been claimed by a low-level domestic insurgency led by the ethnic-based Ogaden National Liberation Front (ONLF) since 1994, with a peace deal between the central government and the ONLF signed as recently as October 2018. In the last decade, Ethiopia and Somalia cooperated to counteract violent extremism in the region. From the east, Somali refugees escaped repeated cycles of internal violence occurring from the late 1980s onward, including during the 2008 drought.203 Second, in the north, following Eritrea’s peaceful independence from Ethiopia (1993), an interstate war between the two countries over a disputed border took place (1998–2000), with instability and further violence occurring as recently as 2012. In the last decade, refugees from Eritrea mostly escaped political persecution, military conscription, and economic hardship.204 The third front of instability is to the west of Ethiopia, at the border with Sudan and South Sudan. The former has had several internal armed conflicts since the 1950s, resulting in continuous cycles of refugees. One such conflict led to the independence of South Sudan (2011), which, in turn, has experienced a full- blown civil war since December 2013. The war has had a severe toll in terms of refugees fleeing to Gambella (Ethiopia), a region with traditional ethnic tensions between the Nuer and Anuak groups.205 323.  As a result of these regional and domestic conflicts, Ethiopia has been one of the most important refugee-hosting countries for decades. The country hosts the sixth largest refugee population in the world and the second largest refugee population in Sub-Saharan Africa, after Uganda. As of May 2018, Ethiopia hosted 920,000 refugees. Nearly half of them arrived in the last four years from South Sudan (443,000), and the other half came at different times from Somalia (256,000), Eritrea (169,000), and Sudan (44,000). Refugees from these four nations tend to settle close to their country of origin (Figure B.213). The nearly 8,000 remaining refugees came from some 15 coun- tries, including Yemen.206 In 2017, nearly 110,000 new refugees arrived in Ethiopia, overwhelmingly from South Sudan (75,000) and Eritrea (25,000).207 These trends (Figure B.212) show how forced displacement in Ethiopia is both a human- itarian emergency and a protracted crisis.   FIGURE B.212    Refugee population in Ethiopia by country of origin 900 800 700 Number of refugees 600 Thousands 500 400 300 200 100 0 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 South Sudanese Somalis Eritreans Sudanese Source: Authors’ calculations based on UNHCR data. 203. International Crisis Group, “Ethiopia: Prospects for Peace in Ogaden”; Carter and Rohwerder, “Rapid Fragility and Migration Assessment for Ethiopia”; Richards and Bekele, “Conflict in the Somali Region of Ethiopia: Can Education Promote Peace-Building?” 204. International Institute for Strategic Studies (IISS); Carter and Rohwerder 2016. 205. Feyissa 2014; Asfah Gemechu 2016; World Bank 2016; Carter and Rohwerder 2016. 206. UNHCR 2018a. “Ethiopia Fact Sheet (May 2018).” 207. Ibid. Volume B: Country Case Studies  | 155 324.  Forced displacement of IDPs is a rising emergency too. As of the end of 2016, Ethiopia had 258,000 IDPs orig- inating mostly from the unrest in Ogaden/Somali. Following widespread anti-government protests and ensuing vio- lence and instability, conflict-induced IDPs skyrocketed to an estimated 1.1 million by April 2018 (plus another 488,000 IDPs due to environmental factors).208 This dramatic spike was due to widespread anti-government protests and state repression that have taken place since late 2015 in several regions, including Oromia, Tigray, and Amhara, reinforcing long-standing ethnic grievances of some ethnic groups and regions. The issue of internal forced displacement in Ethi- opia is also a protracted one, which is intertwined with highly variable environmental conditions.209 325.  Since the conflict contexts are diverse, the displacement situations are similarly heterogenous. Most ref- ugees settle in areas of Ethiopia that border their country of origin and reside in camps. Refugee populations have marked differences in terms of size, demographics, causes, and length of displacement, as well as origins and prospects. In turn, development needs and durable solutions apply differently to the different refugee groups. Border regions that host refugees are often isolated and lagging, some of which suffer from domestic conflicts too (for example, Somali, Oromia, Afar, and Gambella).210 Similarly, there are ethnic, socioeconomic, and security-related differences among host- ing regions in Ethiopia, in addition to the differences between refugee groups. Thus, the present report adopts a differ- ential analytical approach to the displacement situations in the country. 326.  Over 90 percent of the 443,000 South Sudanese refugees settled in camps along the southwestern border of Ethiopia, in Gambella, a traditionally fragile and underserved region.211,212 Here, the South Sudanese refugees (mostly ethnic Nuer) outnumber the host population (307,000 as of 2007 census), causing enormous strains on food security, service delivery, and access to livelihoods. The two main Ethiopian ethnic groups in Gambella are the Nuer (47 percent, mainly pastoralists) and the Anuak (21 percent, mainly farmers), who at the national level represent a mere 0.2 percent and 0.1 percent, respectively.213 The presence of refugees has exacerbated existing tensions between the two groups over land and water rights. However, the influx of refugees and the associated flow of humanitarian assistance have also benefitted host communities through infrastructure projects, and expanded services and local markets.214 327.  Nearly all of the 256,000 Somali refugees reside in eight camps in the conflict-affected Somali region (that is, Ogaden) of eastern Ethiopia. The largest group of Somali refugees (208,000) live confined in Dollo Ado, a small area with a host population of 141,000.215 While Somali refugees arrived in waves over several decades, most of 208. IOM, “Ethiopia 2018: Humanitarian Compendium”; IOM, “Displacement Tracking Matrix (DTM) – Ethiopia. Round 10: March to April 2018”; OCHA, “Ethiopia: Conflict Displacement Situation Report #3”; International Crisis Group, “Ethiopia: Prospects for Peace in Ogaden”; Tesfaye, “The Point of No Return in Ethiopia.” 209. “Internal displacement in Ethiopia is multi-causal and complex. The confluence of numerous drivers and triggers of new displacement is so complex that any attempt to distinguish between displacement caused by conflict or disaster is rendered pointless. The interaction between high levels of existing vulnerability in rural populations; severe droughts, sometimes followed by heavy rains and floods; ongoing conflict; already high numbers of displaced people; and overstretched government capacity creates a high-risk environment in which new displacements are likely to continue.” (IDMC 2018). 210. While the four regions where fieldwork took place are traditionally poorer than the national average, the host communities that were surveyed for this study were mainly peri-urban with poverty trends and development indicators close to the national level. 211. Ethiopia currently hosts 17 percent of the total 2.4 million South Sudanese refugees in the region. In 2016, South Sudan had the third largest, after Syria and Afghanistan, and fastest growing refugee population in the world. 212. UNHCR, “South Sudan Situation: Refugee Population in Gambella Region”; UNHCR, “South Sudan Situation (31 December 2017)”; Sarzin, “Background Paper for the Ethiopia Forced Displacement Strategy Note”; World Bank and UNHCR, “Forced Displacement and Mixed Migration in the Horn of Africa.” 213. Conflict between Anuak and Nuer across Sudan/South Sudan–Ethiopia border dates back to the 19th century. Civil war in Sudan in the 1980s caused a wave of ethnic Nuer refugees from Sudan to pour into Ethiopia, permanently altering Gambella’s ethnic composition. World Bank, “South Sudan Poverty Assessment 2017”; Feyissa, “The Spillover Effect of South Sudan in Gambella”; Asfah, “The Recent Attack in Gambela and Its Implications for Humanitarian Operations.” 214. Girma, “Assessing the Impact of South Sudanese Refugees on the Host Communities of Itang Woreda.” 215. Sarzin 2017a. 156  |  Informing Durable Solutions for Internal Displacement the current refugees arrived after the 2008 drought and the Ethiopian-Somali joint military campaign against violent extremism (2006–09). Somali refugees share the same ethnic group of host communities in the Somali region of Ethi- opia: 97 percent of the 4.4 million Somali region’s population (2007) is ethnic Somali, who in turn, are 6.2 percent at the national level. In the Somali region, historic grievances against the central government fueled a protracted insur- gency by the ONLF and brutal state responses, including a counterinsurgency campaign in 2007. In addition, inter- and intra-clan conflicts traditionally beset the Somali region, increasingly over ownership of resources and governance arrangements. Currently, the region remains neglected, chronically poor, and unstable. Rural population (85 percent of the total) is mostly pastoralist (60 percent), with presence of agro-pastoralists (25 percent) and farmers (14 percent).216   FIGURE B.213    Map of Ethiopia, refugee camps and heat map of conflict events since 1997 Source: ACLED. 328.  Half of the 169,000 Eritrean refugees have settled along the northern border, in the Afar and Tigray regions. Compared to other refugee groups, only one-third of Eritrean refugees live in camps, with many thousands living in individual accommodations and/or benefitting from the ‘Out-of-Camp’ scheme (see below). Eritreans also 216. Richards and Bekele, “Conflict in the Somali Region of Ethiopia: Can Education Promote Peace-Building?”; International Crisis Group, “Ethiopia: Prospects for Peace in Ogaden”; World Bank, “South Sudan Poverty Profile 2015.” Volume B: Country Case Studies  | 157 account for most of the 21,000 refugees who reside in Addis Ababa.217 Many Eritreans do not intend to permanently settle in Ethiopia, but seek to reach Europe via Sudan and Libya: 65 percent of Eritrean refugees arriving in camps in Tigray leave within the first year. The regions of Tigray and Afar are home to 4.3 million and 1.4 million individu- als, respectively. Both are also ethnically homogenous: 97 percent of Tigray’s population is ethnically Tigray, and over 90 percent of Afar’s population is ethnically Afar.218 329.  The vast majority of the 44,000 Sudanese refugees reside in four camps in the Beneshangul-Gumuz region in western Ethiopia. Sudanese refugees come from protracted conflicts and chronically marginalized regions of Sudan, including Darfur, Abyei, Southern Kordofan, and Blue Nile State (from where most of the Sudanese refugees settling in Ethiopia come from). Beneshangul-Gumuz is a multiethnic region inhabited by 780,000 individuals (2007) with a predominantly farming economy. Despite favorable climatic conditions, it suffers from food insecurity due to limited social and economic infrastructure, and more recently environmental degradation.219 Legal Framework, Policy, and Program Responses 330.  Ethiopia formally adheres to international or regional instruments of refugee protection and has adopted national policies consistent with international standards. It signed and ratified the 1951 UN Convention Relating to the Status of Refugees, the 1967 Protocol, and the 1969 Organization of African Unity Convention Governing Specific Aspects of Refugee Problems in Africa. Ethiopia is also a party to the 2010 Kampala Convention on the protection of IDPs. At the national level, the Refugee Proclamation No. 409/2004 establishes protection standards and related rights for refugees, providing the legal framework for Ethiopia’s traditional open door policy to refugees (Art. 9). The procla- mation expresses a commitment for safe reception, promoting peaceful coexistence and returning refugees when conditions in countries of origin are safe. Institutionally, the Administration for Refugees and Returnee Affairs (ARRA) is responsible for the camps’ management, security, and some services (health, education, and WASH). Increasingly, some refugees access basic services through national providers, as well as some host communities benefit from ARRA- provided services.220 331.  Despite adequate protection mechanisms, refugees in Ethiopia still have limited socioeconomic rights. The proclamation allows competent authorities to restrict refugees’ freedom of movement to designated settlements (Art. 21), which has been the legal basis for Ethiopia’s overt use of camps to host refugees. Camps are separated from the social and economic life of host communities and are mostly dependent on aid. In the last decade, the restriction to freedom of movement has been partially lifted. While Ethiopia remains one of the most camp-reliant refugee hosting countries globally, the introduction of an ‘Out-of-Camp’ scheme in 2010 allowed Eritrean refugees who had the means to support themselves to live in urban areas.221 In terms of livelihoods, refugees rely primarily on humanitarian aid. Under the 1995 Constitution, only Ethiopian citizens are granted the right to work, and there are no provisions under Ethiopia’s law for the local integration of refugees. While self-employment is limited to the privileged few and business 217. UNHCR 2017; Sarzin 2017b. 218. World Bank 2017; Sarzin 2017a; Mallet et al., “Journeys on Hold.” 219. UNHCR, “UNHCR Global Trends: Forced Displacement in 2017”; Sarzin. 2017b. “Stocktaking of Global Forced Displacement Data”; World Bank, “South Sudan Poverty Profile 2015.” 220. Sarzin 2017b; World Bank 2016. 221. As part of the 2016 pledges, the government aimed to bring the percentage of urban refugees to 10 percent. The commitment to expand the out- of-camp policy recognizes that it has enhanced the self-reliance of Eritrean refugees who are living outside the camps, increased opportunities to pursue educational and employment opportunities, and reduced pressures to embark on perilous journeys to Europe. (ARRA 2017). 158  |  Informing Durable Solutions for Internal Displacement licenses are not available to refugees, they can only seek limited employment opportunities in camps or in the informal sector of surrounding areas.222 332.  In 2016, the government articulated a strategic approach in a series of pledges aimed at improving rights and expanding services to benefit both refugees and host communities. The nine pledges (Box B.9) include potential provisions to ease the refugees’ restrictions in matters of freedom of movement; labor rights; and access to services, livelihoods, and resources. The groundbreaking plan has a long-term vision that includes interventions in several social and economic sectors beyond the assistance to refugees. In synergy, target key areas can strengthen refugees’ self-reliance by moving away from aid dependency and contribute to greater socioeconomic inclusion of refugees. The pledges are part of the GoE’s adoption of the CRRF—an initiative launched at the UN General Assembly meetings in September 2016. Under the CRRF, the international community commits to supporting host countries and refugee groups through a more comprehensive response to displacement crises. Specifically, the GoE’s adoption of the CRRF entails out-of-camp support to refugees and refugee integration within host communities, gradually moving away from the current in-camp assistance to refugees.  BOX B.9    GoE’s nine pledges The nine pledges are GoE’s strategy to sustainably address refugee crises beyond the humanitarian level and with the involve- ment of a broader array of stakeholders. In line with the CRRF, they represent a groundbreaking approach that includes policy provisions to benefit both refugees and host communities. The nine pledges are the following: (1) Expand the ‘Out-of-Camp’ policy to benefit 10 percent of refugees. (2) Provide work permits to refugees and those with permanent residence identification. (3) Provide work permits to refugees in the areas permitted for foreign workers. (4) Increase enrollment in all levels of education to all qualified refugees. (5) Make available 10,000 ha of irrigable land to allow 100,000 people (refugees and local population) access to crop production. (6) Allow refugees who have lived in Ethiopia for 20 or more years to locally integrate. (7) Build industrial parks, reserving some jobs for refugees. (8) Strengthen provision of social services. (9) Provide access to other benefits including birth certificates, bank accounts, and driving licenses. Source: ARRA, 2017. 333.  The World Bank currently supports the GoE’s efforts to address the displacement crises. In addition to the rollout of the pledges, support to the implementation of the CRRF also includes the reform and expansion of ARRA, 222. Zetter and Ruaudel, “KNOMAD Study Part-II Refugees Rights to Work—An Assessment.” Volume B: Country Case Studies  | 159 from an agency exclusively responsible for refugee protection to a potential larger role of coordination of refugee policies within Ethiopia’s broader economic development agenda. The World Bank’s portfolio on displacement in the country includes the ongoing Development Response to Displacement Impact Project in the Horn of Africa (P152822), aimed at improving access to services, enhancing environmental management of host areas, and expanding economic opportunities for both refugee groups and host communities in Ethiopia, Djibouti, and Uganda; the Economic Oppor- tunities Program (P163829) with a refugee component to create jobs, build skills, and support work permits and busi- ness license for refugees; and the Additional Financing to Education Project. Some of the funding for these programs are drawn from the IDA18 US$2 billion sub-window dedicated to refugees and to supporting host countries.223 Demographic Profile 334.  Differences in displacement dynamics between the four main refugee groups (South Sudanese, Somalis, Eritreans, and Sudanese) are reflected in their demographic profiles. Refugee populations display unique demo- graphic features based on the drivers and immediate causes of their displacement. Demographics of the four groups are also related to the length of displacement, to the severity of violence to which refugees were exposed, and more generally to the intensity and status of the conflicts in the four countries. 335.  Sudanese, South Sudanese, and Somalis overwhelmingly link their displacement to security issues, while for Eritreans state persecution is the most important driver of displacement (Figure B.214). Armed conflict is the most prevalent reason for fleeing that Sudanese, South Sudanese, and Somali refugees report. For Somalis, both rising crime and violence and drought are also particularly relevant. Most Eritrean refugees report being displaced to Ethi- opia due to political persecution (48 percent) and rising crime (30 percent), while another combined 11 percent fled due to family reasons and lack of employment. Only 5 percent of Eritreans maintain that the opportunity to migrate to another country was the leading factor for displacement: this percentage seems low, as in other studies migrating has come up as a more frequent explanation The most important causes of displacement for Eritreans are either marginal or negligible for the other three groups. 336.  All refugee groups list improved security as the leading reason for settling in the current location of dis- placement in Ethiopia (Figure B.215). This point directly links the reasons for fleeing to the reasons for settling in a specific location as refugees. Unsurprisingly, conflict and security are the most recurrent factors of forced displacement dynamics, both as push and pull factors. Conflict and security are related to forced displacement in terms of (a) what the causes are for an individual or group to flee (that is, push factors), and (b) what his/her motives are for choosing and settling in a new place to live (that is, pull factors). In addition to security, Eritrean respondents provide more diverse answers. Due to the more open legislation and policies by Ethiopia toward Eritrean refugees compared to the other groups, more than 30 percent of Eritreans state that the possibility to join family members and employment opportu- nities were factors in their decision to settle in the current location in Ethiopia. These two reasons are barely mentioned at all by the other three groups. In addition, approximately a quarter of Eritrean, South Sudanese, and Somali refugees state that access to humanitarian aid was driving their decision to settle in a specific area, whereas this factor is not relevant for the Sudanese. 223. Sarzin 2017a and 2017b; World Bank 2016. 160  |  Informing Durable Solutions for Internal Displacement  FIGURE B.214    Reasons for displacement   FIGURE B.215    Reasons for settling in the current location 100 100 80 % of refugees 80 60 % of refugees 60 40 40 20 20 0 South Somalis Eritreans Sudanese 0 Sudanese South Somalis Eritreans Sudanese Sudanese Armed conflict Increased crime or violence Discrimination Climate events Better security Humanitarian aid Lack of employment Family Asset or service access Livelihood opportunities Migration Family Source: Authors’ calculations using SPS 2017. Source: Authors’ calculations using SPS 2017. 337.  Demographic profiles of the four groups present differences in terms of age and sex. Eritrean refugees have a higher percentage of men who are at a military age (over 31 percent combined adults and youth ages 15–24), compared to the other three groups (14 percent South Sudanese, 18 percent Somalis, and 20 percent Sudanese). This could be explained by the fact that Eritrean men at home are subject to mandatory, indefinite, and harsh military ser- vice, and have therefore a higher incentive to flee than similar age groups in other countries.224 338.  One more remarkable difference across groups concerns the pronounced sex imbalance in favor of women among South Sudanese refugees. While there are 16.4 percent South Sudanese refugees who are adult women, there are only 6 percent adult men (Figure B.216). Other groups, including host communities, have an approxi- mately similar sex balance in their adult population.225 In turn, the low percentage of adult men among South Sudanese refugees results in an overwhelming percentage of South Sudanese refugee women-headed households in Ethiopia (91 percent), compared to Eritrean, Somali, and Sudanese refugee households. The majority of the latter three groups’ households are headed by men (Figure B.217). 339.  In terms of age group, refugees are predominantly young, leading to high dependency rates. All groups except Eritreans have higher percentages of children than in host communities. About 60 percent of the refugee pop- ulations of Sudan, South Sudan, and Somalia are under 15 years of age, as compared to 50 percent of the host com- munity population (Figure B.216). On the other hand, host communities have a higher proportion of adults (especially men) than the four refugee groups. Young populations lead to high dependency ratios, which put an increased burden on working-age individuals (Figure B.218). In turn, a relative shortage of men of working age among refugees may indicate higher rates of dependency from external aid, for example, and lower rates of economic self-reliance. The dependency ratio is highest for South Sudanese and Somali refugees: one working-age member is responsible to 224. Human Rights Watch 2015, 201. 225. The gender imbalance among South Sudanese refugees could be partially attributed to the intensity and volatility of the civil war in South Sudan. Adult men may be more systematically recruited by armed groups compared to Somalia and Sudan where conflict dynamics are protracted. Such hypothesis could be confirmed by survey respondents from South Sudan, who list ‘recruitment’ as an important factor for family separation (30 percent), compared to the other three refugee groups, for which ‘recruitment’ is negligible (Figure B.220). However, that is not the full picture. Such gender imbalance among South Sudanese refugees exists among South Sudanese population too (see South Sudan case study). Volume B: Country Case Studies  | 161 support approximately two dependents compared to the 1:1.2 ratio among host populations. This situation speaks to the extreme vulnerability of refugees when considering the overwhelmingly prevalent women-headed households of South Sudanese refugees, for example.  FIGURE B.216    Demographics of refugee population in Ethiopia 60 50 40 % of population 30 20 30.3 30.8 30.8 32.0 28.7 32.1 28.4 24.9 24.8 26.3 10 23.5 19.9 0 Men Women Men Women Men Women Men Women Men Women Men Women Refugee Host community South Sudanese Somalis Eritreans Sudanese Children (0–14) Youth (15–24) Adults (25–64) Elderly (>65) Source: Authors’ calculations using SPS 2017.   FIGURE B.217    Sex of household head   FIGURE B.218    Dependency ratio 100 2.5 2.1 80 2.0 1.9 1.9 % of households Dependency ratio 1.7 60 1.5 1.2 1.2 40 1.0 20 0.5 0 0 Refugees Host community South Sudanese Somalis Eritreans Sudanese Refugees Host community South Sudanese Somalis Eritreans Sudanese Overall Refugees Overall Refugees Man Woman Source: Authors’ calculations using SPS 2017. Source: Authors’ calculations using SPS 2017. 340.  Rates of separation from family members during forced displacement are different for the four refugee groups. These differences may be related to the drivers and circumstances of displacement. More than 25 percent of refugees report that they have been separated from immediate family members due to displacement (Figure B.219). This estimate is higher for Eritrean and South Sudanese refugees (39 percent and 38 percent, respectively) than for Sudanese and Somalis (27 and 4 percent, respectively). Arguably, the drivers of forced displacement in the former groups (intense violent conflict in South Sudan and repressive regime in Eritrea) could help explain their higher rates 162  |  Informing Durable Solutions for Internal Displacement of family separation than the latter groups. The intensity of the civil war in South Sudan, the emergency character of displacement, and recruitment by armed groups may cause refugees to flee faster than in the Somalia context, for example. As a result, more South Sudanese families may get separated during displacement, compared to Somalis and Sudanese who may have more time to organize and flee. For different reasons, Eritrean refugee families may suffer from higher rates of separation because either politically active men or men at military age are at higher risk of incarceration and conscription, respectively. Ultimately, Eritrean men may be forced to flee even without family members; the latter may be at a lower risk and thus may have lower incentives to flee. 341.  As the four refugee groups exhibit different percentages of family separation, the reasons why family members got separated are also different among the four refugee groups. Most of these separated household members were left behind in the country of origin at the time of displacement. Eritrean and Sudanese survey respon- dents list this as the only reason for separation. Almost half of Somali and a quarter of South Sudanese refugees, instead, report that family members were displaced to another location. In South Sudan, recruitment by armed groups also plays an important role in separating families, as nearly one-third of South Sudanese male family members either joined or were recruited by armed groups (Figure B.220).226 In turn, separation contributes to refugees’ vulnerabilities.   FIGURE B.219    Percentage of  FIGURE B.220    Reasons for separation refugee population separated from their 100 household members % of separated population 90 45 80 70 40 60 35 50 % of population 30 40 25 30 20 20 15 10 10 0 Women Women Women Women Women All All All All All Men Men Men Men Men 5 0 se is ns se l al al Refugee— South Somalis Eritreans Sudanese er ne ea ne m ov da itr da So overall Sudanese e— Er Su Su e h ug ut ef So Deceased Recruited in armed forces R Displaced to another location Stayed behind Source: Authors’ calculations using SPS 2017. Source: Authors’ calculations using SPS 2017. 342.  In Somalia, South Sudan, and Sudan, a disproportionately high percentage of separated members are adult men compared to women. For South Sudan, 48 percent of the separated members are adult men and 30 per- cent are adult women, while for Sudan 36 percent of the separated members are adult men and 24 percent are adult women (Figure B.221). For Somali refugees, the sex discrepancy of separated family members is even more tilted toward men: over 33 percent of the separated members are adult men compared to only 6 percent of adult women. For both Somali and Eritrean refugees, there is also a higher proportion of separated members who are children and youth than in the other two groups. Eritrean refugees experience an opposite trend than the previous three groups. As Eritrean 226. For South Sudanese IDPs, the data do not show that recruitment plays a role in separating families (see South Sudan case study). Volume B: Country Case Studies  | 163 men cross into Ethiopia to avoid forced military service, a disproportionately higher percentage of separated members are adult women—24 percent of adult women are separated as compared to 15 percent of adult men.   FIGURE B.221    Demographics of separated population 80 % of separated household members/households 60 40 20 0 Men Women Men Women Men Women Men Women Men Women Refugees— South Sudanese Somalis Eritreans Sudanese overall Children (0–14) Youth (15–24) Adults (25–64) Elderly (>65) Source: Authors’ calculations using SPS 2017. Standard of Living 343.  Refugees in Ethiopia are much poorer than host community households, but poverty rates across the four refugee groups are quite heterogenous, ranging from 38 percent (Eritreans) to 73 percent (Suda- nese). Every two in three refugees live below the international poverty line of US$1.90 PPP (2011) per day per capita (Figure B.222). This compares with around one in four host community members living below the poverty line. Some refugee groups (South Sudanese and Eritreans) fare better compared to their respective country of origin at the national level, while some others (Sudanese, Somalis) fare worse. South Sudanese and Sudanese ref- ugees in Ethiopia have the highest incidence of poverty (each group with over 70 percent), but the two countries of South Sudan and Sudan have remarkably different poverty rates (82 percent and 47 percent, respectively). Pov- erty incidence in Eritrea is also high (66 percent), although this figure and the one for Sudan are quite outdated. However, Eritrean refugees in Ethiopia have the lowest poverty rate among the refugee population in the country with a poverty rate that is closer to that of the host population. Poverty incidence of women-headed households is higher than men for all groups, except for Somali refugees for whom the percentages between women and men are approximately equal. 344.  Refugees will need a substantial increase in their consumption to overcome poverty. On average, refu- gees who are poor consume 28 percent below the poverty line (Figure B.223). Like the trends in poverty incidence, South Sudanese and Sudanese refugees suffer from the highest poverty gap, followed by Somali and Eritrean ref- ugees. On average, the former two groups consume 34 percent below the poverty line, while Somalis consume 23 percent below the poverty line. Similar to the trends observed for poverty, Eritrean refugees are the best-off group among the four: Eritreans who are poor on average consume 10 percent below the poverty line. In addition, households headed by women both among all refugees and within all four groups face a higher poverty gap than households headed by men. 164  |  Informing Durable Solutions for Internal Displacement   FIGURE B.222    Poverty incidence of refugee groups, countries of origin, and host communities 100 Poverty incidence 80 (% of population) 60 40 20 0 Woman head Overall Refugee Host community Man head Woman head Man head Woman head Overall Man head Woman head Overall Man head Woman head Overall Man head Overall Refugee South Sudanese Somali Eritrean Sudanese Poverty incidence South Sudan—2016 Somali—2016 Eritrea—2003 Sudan—2009 Source: Authors’ calculations using SPS 2017.  FIGURE B.223    Poverty gap 20 Poverty gap in birr (mean 18 6% 8% relative to PPP P line) 13% 10% consumption shortfall 16 28% 22% 23% 24% 23% 31% 29% 34% 29% 39% 34% 14 34% 12 10 8 6 4 2 0 Refugee Host community Man head Woman head Man head Woman head Overall Man head Woman head Overall Man head Woman head Overall Man head Woman head Overall Overall Refugee South Sudanese Somalis Eritreans Sudanese Mean income Poverty line Source: Authors’ calculations using SPS 2017. 345.  Hunger trends mirror poverty trends for the four groups and their demographics and in relation to the host population, resulting in high dependence on food aid. On average, 65 percent of refugees are highly food insecure compared to 25 percent of host community members experiencing high food insecurity (Figure B.224).227 Refugees in Ethiopia, specifically South Sudanese, Sudanese, and Somali refugees, are in dire need of food aid to over- come high food insecurity in the camps, which is around 80 percent for the former two groups, and 51 percent for the 227. Food insecurity is defined as an individual facing food shortage at least once in the previous seven days and using a combination of coping strategies to overcome the shortage. It is calculated using rCSI adapted by WFP/VAM, FAO/FSNAU (Food Security and Nutrition Analysis Unit for Somalia), and the Global IPC team, among others. rCSI is a weighted index that combines information on frequency and severity of coping strategies used in a single score for household food security. Volume B: Country Case Studies  | 165 Somalis. Mirroring poverty trends, Eritrean refugees fare similarly to host community members and have the lowest food insecurity rates among refugee groups. In addition to higher poverty incidence, women-headed households are also more food insecure than men, and as a result they are more vulnerable. Predictably, food security improves with income: more than 33 percent of refugees among the richest quintile experience low food insecurity, compared to 5 percent for the poorest quintile and an average of 14 percent for all the other quintiles. These high rates of food insecurity are despite the available food aid in the refugee camps. On average, 41 percent of the food consumption of refugees come from food aid (Figure B.225). This percentage is around 50 percent for the poorest refugee groups (South Sudanese and Sudanese), while it is lower for Eritreans and Somalis (37 percent and 27 percent, respectively).   FIGURE B.224    Food insecurity   FIGURE B.225    Aid as a share of food consumption 100 100 80 80 % of population % of population 60 60 40 40 20 20 0 0 Refugee Host community Man head Woman head Poorest quintile Q2 Q3 Q4 Richest quintile South Sudan Somalia Eritrea Sudan Refugee Host community Man head Woman head Poorest quintile Q2 Q3 Q4 Richest quintile South Sudanese Somalis Eritreans Sudanese Overall Refugee Overall Refugee High food insecurity Medium food insecurity Source: Authors’ calculations using SPS 2017. Low food insecurity Source: Authors’ calculations using SPS 2017. 346.  The housing situation of refugees in Ethiopia—assessed in terms of housing ownership, conditions, and overcrowding—goes hand in hand with poverty and hunger trends. Even more than food aid, housing needs of refugees in Ethiopia are almost entirely provided by the UN or non-profit organizations through temporary shelters. Because these sources are unsustainable, strengthening self-reliance is key to achieving durable solutions. As for other standard of living measures, refugees fare worse than host community members on housing issues too. Ethiopia’s overreliance on camps is confirmed by survey respondents among refugees, 95 percent of whom live in temporary shelters compared to nearly 80 percent of host community members who live in owned dwellings (Figure B.226). An overwhelming majority of the refugee population (82 percent) live in unimproved housing,228 and 68 percent of the refugee population live in overcrowded housing.229 In contrast, 54 percent of the host community population live in unimproved housing and 45 percent of them live in overcrowded housing (Figure B.227; Figure B.228). Unimproved and crowded housing conditions increase the likelihood of disease transmission for a wide range of respiratory diseases including pneumonia, tuberculosis, and many allergies. Both refugees and host community populations need improve- ment in housing conditions as well as in food security. 228. Unimproved housing is defined as a structure that is not made of wood, concrete, or block and/or that is not intended for habitation. 229. Housing is defined as overcrowded when there are four or more individuals per room (UN-HABITAT 2016). 166  |  Informing Durable Solutions for Internal Displacement   FIGURE B.226    Ownership of dwelling 100 80 % of population 60 40 20 0 Refugee—current Host community Man head Woman head Poorest quintile Q2 Q3 Q4 Richest quintile Origin Current Origin Current Origin Current Origin Current Refugee—origin South Somalis Eritreans Sudanese Sudanese Overall Refugees Temporary shelter (UN/NGOs) Squatting Relatives/friends Work-provided Rented Owned Source: Authors’ calculations using SPS 2017.   FIGURE B.227    Housing conditions 100 80 % of population 60 40 20 0 Refugee—origin Refugee—current Host community Man head Woman head Poorest quintile Q2 Q3 Q4 Richest quintile Origin Current Origin Current Origin Current Origin Current South Somalis Eritreans Sudanese Sudanese Overall Refugees Unimproved housing Improved housing Source: Authors’ calculations using SPS 2017. 347.  According to their nationality, refugee groups experience differences in terms of housing. Trends on own- ership of dwellings are significantly low for all groups, ranging from 2 percent to 11 percent (Figure B.226). This is simply explained by the fact that surveys were conducted in camps where temporary shelters are the norm. As more Eritreans live in urban settings, their rate of either rented or owned dwelling may be higher than the data from survey results indicate. Trends in overcrowding are fairly homogeneous too, ranging from 65 percent of South Sudanese living in Volume B: Country Case Studies  | 167 overcrowding housing to 50 percent of Sudanese living in such conditions (Figure B.228). More differentiation concerns housing conditions. As for poverty and food security, Eritreans fare better than the other three groups, in line with lower poverty incidence, higher consumption rate, and greater disposable income. About 50 percent of Eritrean refugees live in improved housing, compared to 27 percent of Somalis, 16 percent of Sudanese, and 4 percent of South Sudanese (Figure B.227). 348.  The housing situation of refugees was better before displacement than during it. Even though some ref- ugee groups (South Sudanese and Eritreans) have lower poverty rates compared to their respective country of origin, all refugee groups have worse housing situations during displacement than before being displaced. While 94 percent of the refugee population owned their dwellings before displacement, only 5 percent of refugees own their dwellings in Ethiopia. Similarly, only 25 percent of refugees before displacement lived in overcrowded housing, while 65 percent are currently in that situation. South Sudanese refugees experience the greatest gap, as over 90 percent used to live in non-overcrowded housing before displacement compared to 29 percent South Sudanese refugees who are currently able to live in non-overcrowded housing. Among refugee groups, South Sudanese are also the ones with the highest percentage of refugees living in both overcrowded and unimproved housing. Interestingly, a higher percentage of Somali refugees live in improved housing than they used to before displacement—a trend that links with ownership of dwellings which, albeit low in absolute terms, is highest for Somali refugees when compared to other groups.   FIGURE B.228    Overcrowding (4 or more persons per room) 100 80 % of population 60 40 20 0 Refugee—origin Refugee—current Host community Man head Woman head Poorest quintile Q2 Q3 Q4 Richest quintile Origin Current Origin Current Origin Current Origin Current South Somalis Eritreans Sudanese Sudanese Overall Refugees Overcrowded Not overcrowded Source: Authors’ calculations using SPS 2017. 349.  Similarly, refugees have much lower access to electricity or solar power/biogas than host community households (37 percent vs. 66 percent, respectively), but higher access compared to before being displaced. Access to electricity or solar power/biogas increases with household income. Interestingly, only the richest quintile has access to electricity (17 percent), while the bottom three quintiles have access exclusively to solar power/biogas. As women-headed households are poorer, a lower percentage of women-headed households have access to elec- tricity as compared to households headed by men (Figure B.229). There are also substantial and significant differences between the average hours of electricity available to refugee and host community populations, and whether the two 168  |  Informing Durable Solutions for Internal Displacement populations can charge their phones using the electricity. While host community members receive around 16 hours of electricity per day, refugees receive electricity for only half as long. Similarly, while around 70 percent of host com- munity members who receive electricity can charge their mobile phones using electricity, only half as many percent of refugees can charge their mobile phones. On the other hand, except for South Sudanese refugees, a higher percentage of refugees have current access to electricity or solar power/biogas as compared to before being displaced. This finding is opposite to the trends in housing, where the situation before displacement was generally better than during dis- placement. Thus, improved access to electricity and fuel points to the relative good quality of Ethiopia’s refugee camps.   FIGURE B.229    Source of lighting 100 % of households 80 60 40 20 0 Refugee—origin Refugee—current Host community Man head Woman head Poorest quintile Q2 Q3 Q4 Richest quintile Origin Current Origin Current Origin Current Origin Current South Somalis Eritreans Sudanese Sudanese Overall Refugees No lighting Grass/firewood Lamp/candle/torch Solar power/biogas Electricity Source: Authors’ calculations using SPS 2017. 350.  Access to services for refugees (water and sanitation, health, and education) compares, or in some cases even exceeds, access to the same services for host community members, pointing to a high quality of the camps. Trends in service delivery run opposite to poverty incidence, food insecurity, and housing situation, all of which saw refugees faring worse than host communities. Both refugees and host community members have nearly complete access to improved water sources230 (98 percent and 96 percent, respectively), which for refugees is better than it was in their respective countries of origin (62 percent, Figure B.230). Refugees have also far better access to improved san- itation facilities (69 percent) as compared to host community members (51 percent) and as compared to their access before displacement (28 percent, Figure B.231).231 Similarly, refugees in Ethiopia have greater access to health services (in particular, hospitals) compared to host communities. For example, the majority of refugee women delivered babies at hospitals (72 percent) or maternity clinics (18 percent) during the last 24 months, compared to women among host community who gave birth in hospitals (36 percent) and maternity clinics (52 percent). 230. Improved water sources are piped water supply into the dwelling; piped water to a yard/plot; a public tap/standpipe; a tube well/borehole; a protected dug well; a protected spring; and rainwater. Unimproved water sources are unprotected dug well; an unprotected spring; a cart with a small tank/drum; a water tanker truck; and surface water (WHO 2006). 231. Improved sanitation is defined as a household having some type of flush toilet or latrine, or ventilated improved pit or composting toilet, provided they are not shared. The additional criterion of not sharing the toilet with other households reduces the percentage of population with access to improved sanitation. Following this definition, a lower percentage of the refugee population, 42 percent, have access to own improved sanitation, that is, do not share their toilet with other households (WHO 2006). Volume B: Country Case Studies  | 169  FIGURE B.230    Access to improved sources of water 100 % of population 80 60 40 20 0 Refugee—origin Refugee—current Host community Man head Woman head Poorest quintile Q2 Q3 Q4 Richest quintile Origin Current Origin Current Origin Current Origin Current South Somalis Eritreans Sudanese Sudanese Overall Refugees Unimproved Improved Source: Authors’ calculations using SPS 2017.   FIGURE B.231    Access to improved sanitation 100 80 % of population 60 40 20 0 Refugee—origin Refugee—current Host community Man head Woman head Poorest quintile Q2 Q3 Q4 Richest quintile Origin Current Origin Current Origin Current Origin Current South Somalis Eritreans Sudanese Sudanese Overall Refugees Improved, not accounting for sharing Improved, accounting for sharing Source: Authors’ calculations using SPS 2017. 351.  Overwhelmingly, refugee children in Ethiopia attend primary school (79 percent), whereas enrollment rates for secondary school are remarkably low (13 percent). Refugee children have higher net primary enrollment rates than the national average in their countries of origin. In fact, net primary enrollment rates for refugee children (79 percent) are even slightly higher than net primary enrollment rates for children in the host communities (Fig- ure B.232). In contrast, net secondary enrollment rates for refugee children are lower both as compared to net enroll- ment rate for children of host community households and as compared to net secondary enrollment rates in their respective countries of origin (except for South Sudan). In addition, a majority (82 percent) of secondary age students who are not in secondary school are attending grades in primary school. While forced displacement clearly contributes 170  |  Informing Durable Solutions for Internal Displacement to school disruption, the fact that host communities also have a similar percentage of children of secondary age attend- ing primary school (65 percent) suggests that there are structural barriers to secondary school attendance in Ethiopia other than displacement.232   FIGURE B.232    Net primary and secondary enrollment rates 100 secondary age children 80 % of primary and 60 40 20 0 Refugee Host community Boys Girls South Sudanse Somalis Eritreans Sudanese Overall Refugees Net primary enrollment Net secondary enrollment National—primary National—secondary Source: Author’s calculations based on SPS 2017, World Bank World Development Indicators, MICS 2014 (Sudan), HFS South Sudan Wave 3 (2016) and HFS Somalia Wave 1 (2016). 352.  Most refugee children between 6 and 18 years of age (72 percent) attend schools run by NGOs, but there is an ample difference across the four refugee groups.233 While nearly all South Sudanese refugees attend NGO-run schools (96 percent), most Somalis (65 percent) and Sudanese (57 percent), and approximately 35 percent of Eritreans attend Ethiopian public schools. This estimate points to not only differences in terms of camp management and dynamics of displacement (that is, South Sudanese represent the latest arrival of refugees and are the largest group) but also a variation in attaining durable solutions by the different refugee groups. Durable solutions entail an incorporation of refugee children into the government or local school system. Thus, as far as education is concerned, Somalis and Sudanese are the two groups that are faring best. 353.  Proximity to services also represents an important measure to assess quality of service delivery. In fact, the longer the time incurred to access services (that is, proximity to water sources, health facilities), the smaller the rate of access to the specific service; limited access also puts a strain on households and limits the time available for other economically productive activities. In Ethiopia, refugees have shorter access to services while in displacement than before being displaced, resulting in some improvement in quality of life. Such improvement is related to the quality of 232. Interestingly, Eritrean children faced the most difficulty to obtain documents to enroll in schools: 20 percent of Eritrean children between the ages of 6 and 18 who were currently out of school cited a lack of documentation as the reason for not attending school. This finding may be counterintuitive as Eritreans are the most integrated among refugee groups in Ethiopia on a number of indicators. Nonetheless, it could be the case that other refugees are not be required to produce documents to enroll in school (for example, school is inside the refugee camp), whereas Eritreans who are more integrated and have more freedom of movement may have to go through more elaborate bureaucratic processes to enroll in school. A higher degree of integration may then involuntarily come with more barriers to access services (in this case education). 233. This contrasts with children in host community households who overwhelmingly attend government schools. Volume B: Country Case Studies  | 171 refugee camps in Ethiopia, which are meant to be temporary. Thus, greater proximity cannot be taken as a sustainable indicator of improvement in the life of refugees. Irrespective of income status, nationality, and sex of the household head, it takes refugees approximately the same amount of time as host community members to reach the nearest source of water, hospitals, and primary school (Figure B.233, Figure B.234, Figure B.235).   FIGURE B.233    Mean time (minutes) taken (one way) to fetch water 100 Mean time (minutes) taken to fetch water 80 60 40 20 0 Refugee—origin Refugee—current Host community Man head Woman head Poorest quintile Q2 Q3 Q4 Richest quintile Origin Current Origin Current Origin Current Origin Current South Somalis Eritreans Sudanese Sudanese Overall Refugees Source: Authors’ calculations using SPS 2017.   FIGURE B.234    Time (minutes) to walk one way to health facility 180 Time (minutes) to walk one way to hospital 160 140 120 100 80 60 40 20 0 Refugee—origin Refugee—current Host community Man head Woman head Poorest quintile Q2 Q3 Q4 Richest quintile Origin Current Origin Current Origin Current Origin Current South Somalis Eritreans Sudanese Sudanese Overall Refugees Source: Authors’ calculations using SPS 2017. 172  |  Informing Durable Solutions for Internal Displacement   FIGURE B.235    Time (minutes) to walk one way to primary school 120 Time (minutes) to 100 walk one way to primary school 80 60 40 20 0 Refugee—origin Refugee—current Host community Man head Woman head Poorest quintile Q2 Q3 Q4 Richest quintile Origin Current Origin Current Origin Current Origin Current South Somalis Eritreans Sudanese Sudanese Overall Refugees Source: Authors’ calculations using SPS 2017. Employment and Livelihoods 354.  An overwhelming majority of refugees rely on aid (including cash, food, and non-food) from the govern- ment or humanitarian organizations for their livelihood. While refugees relied primarily on agriculture and wages and salaries for their livelihood before displacement, currently 83 percent of refugees obtain their livelihood through aid (Figure B.236). This is in stark contrast to the host populations, who mainly derive their livelihood from agriculture (39 percent), and wages and salaries (28 percent), with 25 percent being occupied in services or retail. Among refugees, richer households and households headed by men are slightly more likely to depend on wages and salaries for their livelihood (each group with an estimated 15 percent) than the poorest quintile and women-headed households (2 and 4 percent, respectively). With respect to nationality, all groups overwhelmingly depend on aid except for the Somalis: while 66 percent of Somali refugees are aid dependent, 20 percent obtain their livelihood from salaries and wages, and 15 percent from other sources (including services, retail, and agriculture). Thus, Somalis’ greater economic self-reliance compared to other groups may explain their higher rate of home ownership and housing conditions enjoyed by Somali refugees compared to Sudanese, South Sudanese, and partially Eritreans. 355.  High dependency on aid is a result of low labor force participation rates among refugees. Only 22 percent of working-age refugees (15–64) are currently employed, compared to 66 percent of the host community working-age population who is employed (Figure B.237). Among refugees, the working-age population in the richest quintile is more likely to be employed than the poorest (37 vs. 9 percent), confirming the trend that refugees with higher income are more economically self-reliant. Refugee status comes with legal challenges and restrictions around being allowed to work, which explains the low labor force participation rate, as over 70 percent of them are inactive (neither employed nor actively looking for employment). In contrast, 35 percent of host community members in working age are inac- tive. On a positive note, 27 percent of all inactive refugees are currently enrolled in school or college, and are thereby developing skills. There are significant differences in the proportion of working-age population inactive but enrolled in schools across nationalities. Unexpectedly, Sudanese and South Sudanese—the two groups with higher poverty incidence—have higher rates of inactive refugees who are attending school (41 percent and 34 percent, respectively) compared to Somali and Eritreans (12 percent and 11 percent, respectively), both of whom have over a 60 percent rate of inactivity without school enrollment. For Eritreans, this trend may be explained by the fact that most members of Volume B: Country Case Studies  | 173 this group only stay in Ethiopia on a temporary basis, and thus do not invest in education—an explanation that can- not be applied to Somalis. On the other hand, Somali refugees have the highest rate of employment among refugees (28 percent).   FIGURE B.236    Source of livelihood currently and before displacement 100 % of households 80 60 40 20 0 Refugee—current Host Man headed Woman headed Poorest quintile Q2 Q3 Q4 Richest quintile Origin Current Origin Current Origin Current Origin Current Refugee—origin South Somalis Eritreans Sudanese Sudanese Overall Refugees Agriculture Manufacturing Retail Services Wages and salaries Income from assets Aid Remittances Source: Authors’ calculations using SPS 2017.   FIGURE B.237    Labor force participation and employment status 100 80 age population % of working 60 40 20 0 Refugee Host Men Women Poorest quintile Q2 Q3 Q4 Richest quintile Overall Men Women Overall Men Women Overall Men Women Overall Men Women South Somalis Eritreans Sudanese Sudanese Overall Refugees Inactive, not enrolled Inactive, enrolled Employed, enrolled Employed, not enrolled Unemployed Source: Authors’ calculations using SPS 2017. 356.  There are important differences between men and women participating in the labor force, which are likely to persist in the foreseeable future. Men of the working-age population have a higher rate of employment than women (27 percent vs. 19 percent, respectively). More importantly, most women are inactive and are not pursuing an education opportunity (55 percent) compared to men in the same situation (31 percent). In fact, this trend is true for 174  |  Informing Durable Solutions for Internal Displacement all four nationalities of refugee groups, ranging from 47 percent of South Sudanese refugee women, who are inactive and not attending school or college, to 72 percent of Eritrean refugee women in the same condition. While there are more women than men who are not currently part of the labor force, it is conceivable that more women than men will not be part of it in the future either, because they are not working toward attaining an education. Thus, there are specific gender-based barriers to women in accessing the labor force both in the present and in the foreseeable future. 357.  On average, refugees have similar levels of educational attainment comparable to those of host pop- ulations. Both refugee and host community populations have a large and comparable percentage of working-age population without any education (41 percent and 44 percent, respectively, Figure B.238). However, Ethiopians in the host community are more likely to have secondary and university education (19 percent and 6 percent, respectively) as compared to the refugee population (13 and 2 percent, respectively). Also, among refugees, a higher percentage of men have some education as compared to women across all nationalities. There are significant differences between the educational attainment of refugees of different nationalities, with Eritreans being the most likely to have some educa- tion, followed by South Sudanese and Sudanese refugees. Somali refugees have the lowest educational attainment but as observed above, have the highest labor force participation and employment rates.   FIGURE B.238    Highest educational attainment for working-age population 100 80 age population % of working 60 40 20 0 Refugee Host community Men Women Poorest quintile Q2 Q3 Q4 Richest quintile Overall Men Women Overall Men Women Overall Men Women Overall Men Women South Somalis Eritreans Sudanese Sudanese Overall Refugees No education Primary & intermediate Secondary University Technical Others Source: Authors’ calculation using SPS 2017. 358.  Barriers to labor force participation and employment for refugees need to be removed. About 28 percent of the working-age population does not participate in the labor force because they are enrolled in school or college and are therefore not currently active in the labor market (Figure B.239). Another 20 percent of working-age refugees (predominantly women) are not participating in the labor force because they are taking care of their households. Even those who enter the labor market face barriers in securing employment. Of those who are looking for jobs, around half report a lack of regular work opportunities and around a quarter feel that they lack adequate skills and experience (Figure B.240). Around half have been looking for a job for more than a year, and almost all (97 percent) of the refugee population not participating in the labor force in the last seven days was not engaged in any economic activity for more than a year. These protracted periods of displacement and economic inactivity potentially erode skills and make it harder for refugees to find employment. Refugees report that lack of connections and discrimination also played a role. Volume B: Country Case Studies  | 175   FIGURE B.239    Reasons for not participating in the   FIGURE B.240    Reasons for not securing employment labor force (refugees) (refugees) 30 35 % of population inactive 25 30 % of unemployed in labor force 25 population 20 20 15 15 10 10 5 5 0 0 n ) es ills fy er nce ts ns io es d l re b k d el or le jo ol ci en iti at sk ca io nw iti pe l n w rie in ro a or m ct n rtu e ld U rtu im to (s En d e pe cu at g ho fin nn po un cr qu po ed do ex se co op is o th yo de op w tt D of of ou O lo of o a ec o d ck ck al N In H To ite ck xp La La ot m La te N Li no s oe D Source: Authors’ calculations using SPS 2017. 359.  Refugees need access to productive assets to develop livelihood opportunities for themselves. Before dis- placement, most refugees had access to agricultural land, at least one productive asset,234 and at least one livestock.235 Currently, a very small percentage of refugees have access to any of the above (Figure B.241). Refugees of all national- ities have faced a significant decline in their access to agricultural land, productive assets, and livestock since displace- ment, and South Sudanese refugees have faced the greatest decline. Only Somali refugees fare slightly better—a trend that mirrors Somalis’ higher labor force participation and lower dependence on aid compared to other nationalities. While 53 percent of Somali refugees held any livestock before displacement, the same percentage of Somali refugees still hold at least one livestock. However, the quantity of livestock held has declined significantly: while Somali refugees held 6.68 livestock units236 on average before displacement, the current livestock that is held is estimated at 0.2 live- stock units on average. This lack of access to productive assets further limits the ability of refugees to create employ- ment opportunities for themselves and hampers self-reliance. 360.  Refugees who are employed are working primarily as salaried laborers or are engaged in non-farm self-employment. While self-employment in agriculture was a predominant activity for most refugees in their coun- tries of origin, especially Somalis and Sudanese, only 5 percent of refugees are currently engaged in own-account farming (Figure B.242). There are remarkable differences between refugee groups according to their nationality. More than half of all South Sudanese refugees that are in employment are merely helping in the non-farm businesses of their families. Instead, the main employment activity of the other three groups is salaried labor, followed by non-farming self-employment. 234. Productive assets include moefer and kember, axe/sickle, plow, weaving equipment, builder’s equipment, carpenter’s equipment, welding equipment, wood cutting equipment, block production equipment, refrigerator, private car, and Bajaj. 235. Livestock include cattle, horses, donkey/mules, pigs, sheep, goats, poultry, camels, and beehives. 236. Livestock units are used for aggregating the numbers of different categories of livestock for regional and global comparisons and are obtained by converting the body weight into the metabolic weight. The livestock unit coefficients used are those corresponding to the regions of Near East North Africa: cattle—0.70, buffalo—0.70, sheep—0.10, goats—0.10, pigs—0.20, asses—0.50, horses—0.40, mules—0.60, camels—0.75, chickens—0.01 (Chilonda and Otte 2006). 176  |  Informing Durable Solutions for Internal Displacement   FIGURE B.241    Access to productive assets 100 80 % of population 60 40 20 0 Current Current Current Current Current Origin Origin Origin Origin Origin Refugee Host South Somalis Eritreans Sudanese Sudanese Overall Refugees Agricultural land Productive assets Livestock Source: Authors’ calculations using SPS 2017.   FIGURE B.242    Primary employment activity 100 % of employed 80 population 60 40 20 0 Origin Current Host Origin Current Origin Current Poorest quintile Q2 Q3 Q4 Richest quintile Origin Current Origin Current Origin Current Origin Current Refugee Men Women South Somalis Eritreans Sudanese Sudanese Overall Refugees Wages and salaries Own non-farm business Help in non-farm business Own account agriculture Apprenticeship or training Source: Authors’ calculations using SPS 2017. 361.  Both salaried and self-employed refugees are predominantly engaged in the services sector. A remark- able 71 percent of employed refugees are working in the services sector with only 12 percent working in agriculture, 10 percent in manufacturing, and 6 percent in education. This contrasts significantly with the host community popu- lation, as 48 percent of the employed host community population is currently employed in services and 41 percent in agriculture (Figure B.243). When asked details about the work refugees and host community members would want to do in the future, an overwhelming majority of both refugees and host community members aspire to continue or start working in the services sector. After services, more men among refugees aspire to work in the manufacturing sector as compared to women (16 percent vs. 7 percent). Similarly, after services, 25 percent of Eritreans aspire to work in the manufacturing sector, an aspiration which is not largely shared by refugees of other nationalities. Volume B: Country Case Studies  | 177   FIGURE B.243    Sector of employment 100 80 % of employed population 60 40 20 0 Origin Current Aspiration Current Aspiration Origin Current Aspiration Origin Current Aspiration Origin Current Aspiration Origin Current Aspiration Origin Current Aspiration Origin Current Aspiration Refugees Host Men Women South Somalis Eritreans Sudanese Sudanese Overall Refugees Agriculture Manufacturing Services Education Public sector/defense Source: Authors’ calculations using SPS 2017. Social Cohesion, Public Participation, and Security 362.  Both refugees and host communities perceive that they are in good relationship. Only a relatively small percentage of both groups perceive that relations are bad: 18 percent of refugees and 10 percent of host communities report so (Figure B.244). Somalis have much better relations with host community members followed by Eritreans and Sudanese, although over 85 percent of all three groups report to have good relations. South Sudanese refugees are the group reporting the lowest percentage in terms of good relations, perhaps because they are the newest group and have not yet adapted: 35 percent think that refugees and host community populations have bad relations with each other.237 The more positive perceptions by Somalis tie with their longer permanence as refugees, common ethnic iden- tity, and similar clan system. These commonalities entail a higher degree of integration, from economic self-reliance and higher participation in the labor force to better housing conditions and lower poverty incidence. In addition, the higher the income level for all refugees, the more positive relations are, ranging from 67 percent of poorest quintile who report good relations to 95 percent of the richest quintile who believe relations are good. Women are more likely than men to report negative relations: nearly 25 percent of women perceive relations are not good, compared to 5 percent of men. 363.  Host communities’ sentiments toward refugees are remarkably varied when the three surveyed regions that host the refugees are considered (Figure B.245; Figure B.246). In the Somali region, relations between host communities and Somali refugees are the best. Over 90 percent ‘agree’ that relations are good, 71 percent of which ‘strongly agree’ that relations are good. This confirms that Somali refugees are well integrated within their host commu- nity population. In the Tigray and Afar regions that host Eritreans, the picture is slightly more mixed, although relations are overall perceived positively. In contrast, host community sentiments in the Benishangul Gumuz region (which hosts 75 percent Sudanese and 25 percent South Sudanese refugees) are not very positive: about 35 percent report that relations are not good, and only 25 percent ‘strongly agree’ that they are good. In addition, approximately 60 percent of host community members in Benishangul Gumuz report that Ethiopians feel that refugees should be repatriated, that crime has risen, and it is more difficult to secure employment. 237. Another reason could be that there is some saturation effect: problems become apparent when more and more refugees arrive. Thus, latecomers are perceived in negative terms as they are linked to these experiences.   FIGURE B.244    Interpersonal relations between refugees and host community 100 80 % of population 60 40 20 0 Refugee—current Host community Man headed Woman headed Poorest quintile Q2 Q3 Q4 Richest quintile South Sudanese Somalis Eritreans Sudanese Overall Refugees Refugees and Ethiopians have good relations with each other Strongly disagree Slightly disagree Neither agree nor disagree Slightly agree Strongly agree Source: Authors’ calculations using SPS 2017.   FIGURE B.245    Host community feelings: good relations 100 80 % of host population 60 40 20 0 Somali Tigray/Afar Beninshangule Strongly agree Slightly agree Neither agree nor disagree Slightly disagree Strongly disagree Source: Authors’ calculations using SPS 2017.   FIGURE B.246    Host community feelings: other measures 100 % of host population 80 60 40 20 0 Somali Tigray/Afar Beninshangule Somali Tigray/Afar Tigray/Afar Beninshangule Somali Beninshangule Ethiopians want refugees to The arrival of refugees has made The arrival of refugees has return to their homes it more difficult for people in this brought insecurity to the area community to find work Strongly disagree Slightly disagree Neither agree nor disagree Slightly agree Strongly agree Source: Authors’ calculations using SPS 2017. 178 Volume B: Country Case Studies  | 179 364.  Public participation of refugees is not high, as 50 percent of refugees never participate in the public sphere nor have interactions with community leaders (Figure B.247; Figure B.248). However, there are remark- able differences between refugee groups according to nationality. Eritreans and Sudanese have a 90 percent rate of participation in public meetings, and nearly 80 percent and 60 percent rate of interaction with community leaders, respectively. Somalis and South Sudanese report an opposite picture: over 75 percent of the former and nearly 50 per- cent of the latter never participate with similar trends in interactions with leadership. The low public participation of Somali refugees is striking: by several measures, Somalis are the better-off group and the most integrated one econom- ically. This finding shows how the relationship between community participation and social and economic integration is not uniform, but multifaceted. Host communities have higher rates of participation than refugees as only 20 percent fail to participate publicly and 30 percent never interact with community leaders. In terms of sex, both men and women equally participate in the public sphere.   FIGURE B.247    Participation in public meetings, last   FIGURE B.248    Interaction with community leader, 12 months last 12 months 100 100 % of population % of population 80 80 60 60 40 40 20 20 0 0 Refugee—current Host community Man headed Woman headed Poorest quintile Q2 Q3 Q4 Richest quintile South Sudanese Somalis Eritreans Woman head Q2 Q3 Q4 South Sudanese Somalis Eritreans Sudanese Refugee—current Host community Man head Poorest quintile Richest quintile Sudanese Overall Refugees Overall Refugees Never 1 to 4 times 5 or more times Never 1 to 4 times 5 or more times I am a community leader Source: Authors’ calculations using SPS 2017. 365.  Over half of the refugee population in Ethiopia feels very safe or moderately safe in the refugee camps, while host communities feel overwhelmingly safe. While 30 percent of the refugees feel neither safe nor unsafe, 15 percent of the refugee population feel unsafe at home or walking around in the refugee camps during day or night (Figure B.249).238 This sense of security varies significantly across sex of household head, household income, and nation- ality. A higher percentage of refugees living in households headed by women feel unsafe as compared to refugees living in households headed by men (19 percent vs. 5 percent). Similarly, wealthier refugees feel safer. According to nationality, differences are striking. A large majority of Somali refugees (78 percent) feel very safe in the camps, com- pared with 18 percent of Eritreans, 8 percent of Sudanese, and only 2 percent of South Sudanese. While a majority of Eritrean and Sudanese refugees also feel moderately safe in the camps, a large percentage of South Sudanese refugees feel either unsafe (28 percent) or neither safe nor unsafe in the camps (49 percent). 238. SPS (2017) asked three questions: ‘In general, how safe from crime and violence do you feel when you are alone at home?’, ‘How safe do you feel when walking around alone after dark?’, ‘How safe do you feel walking around during the day?’. A combined scale for safety was created using these three questions. Cronbach alpha for the scale was 0.70. 180  |  Informing Durable Solutions for Internal Displacement   FIGURE B.249    Feelings of safety and security 100 80 % of population 60 40 20 0 Refugee—current Host community Man headed Woman headed Poorest quintile Q2 Q3 Q4 Q5 South Sudanese Somalis Eritreans Sudanese Overall Refugees Very unsafe Moderately unsafe Neither safe nor unsafe Moderately safe Very safe Source: Authors’ calculations using SPS 2017. Movement and Return Intentions 366.  Based on their nationality, refugees have different ideas about their preferred durable solution.239 Almost half of the refugee population does not plan to move from their current locations in camps at any point in the future— although they do not specifically mention wanting to integrate locally (Figure B.250). Nearly 35 percent report wanting to be resettled, while 16 percent prefer to return. Most South Sudanese refugees intend to stay in their current location followed by Sudanese and Somalis, while only 25 percent of Eritreans want to stay where they are. In contrast, Sudanese refugees’ preferred solution is to return to Sudan (50 percent), while only a small minority of the other three groups intend to return. As it is evident from the literature, most Eritreans want to move to another country (over 60 percent), an intention shared by 40 percent of Somalis, 22 percent of South Sudanese, and by virtually no Sudanese (3 per- cent). The United States is the ideal destination for 70 percent of the refugees who want to move to a new country. In addition, about 60 percent of all refugees would still want to move to a new country even if the right to settle and work freely in Ethiopia was granted to them. The nationality of the refugee group plays an important role here too in differentiating refugees’ intentions: the rate of those who would prefer to resettle anyway, even in the presence of local integration mechanisms, ranges widely from 80 percent of Somalis to 40 percent for Sudanese. 367.  Refugees highlight the main reasons for their preferred intention and what factors may affect their decision. With respect to return, the same reasons why refugees were displaced in the first place from their respective country of origin apply as reasons for not wanting to return. These include armed conflict and security concerns and political or identity-based persecution, as well as lack of employment opportunities (Eritreans) and drought and associated famine (Somalis). Irrespective of nationality, the refugees who want to stay list the following reasons not to move: better security, better access to education and health services, and access to humanitarian aid (Figure B.251). In contrast, family reasons and other economic indicators do not feature prominently. Concerning refugees who want to move from the current location, lack of employment opportunities (40 percent) is the main reason for wanting to move, followed by lack of access to home/ land/livestock (26 percent) and to humanitarian aid (17 percent). Refugees also mention the political situation and the secu- rity conditions on transit routes as the most important information they would need to make an informed decision. Finally, refugees largely point to security as the most important factor to settle in the preferred location (Figure B.252). 239. Refugees were not specifically asked about their preferred durable solution, but only what their intention was. Conceivably, reported intentions can be framed as refugees’ preference for a durable solution although this caveat should be noted when analyzing and using this data. Volume B: Country Case Studies  | 181   FIGURE B.250    Movement and return plans of refugees 100 80 % of population 60 40 20 0 l ad ad ile 2 3 4 ile se is ns se l ra Q Q Q al nt nt ne ea ne he he ve m ui ui da itr da So O an an tq q Er Su Su st M om es he or h W ic ut Po R So Stay here Move to another camp Move out of camp Return Move to a new country Source: Authors’ calculations using SPS 2017.   FIGURE B.251    Reasons to stay 60 % of refugee population 50 40 30 20 10 0 Overall South Sudanese Somalis Eritreans Sudanese Better security Home, land, livestock access Services access Livelihood access Family Humanitarian aid Source: Authors’ calculations using SPS 2017.   FIGURE B.252    Main support needed to settle in preferred location 100 80 % of refugee population 60 40 20 0 Overall South Sudanese Somalis Eritreans Sudanese Security Mine clearance Original home or land New housing or shelter Farm or grazing plot access Agricultural inputs Funds and assets for business Skills training Connections for job Education market access Health services Aid Source: Authors’ calculations using SPS 2017. 182  |  Informing Durable Solutions for Internal Displacement Targeting Analysis 368.  Nearly 40 percent of households are support-dependent, which is more common among South Suda- nese and Eritrean refugees, and they concentrate in Jewi, Kule, and Tierkidi camps. About 37 percent of refugee households are support-dependent, 45 percent productive but poor, and 18 percent self-reliant. The share of support- dependent households is larger among South Sudanese and Eritrean refugees, followed by Somali and Sudanese (Fig- ure B.253). As a result of better conditions and a lower incidence of poverty in host communities, a larger share of self-reliant households is found among this group. Refugees from Sudan have more households classified as productive but poor (66 percent). Moreover, the majority of support-dependent refugees (around 60 percent of households) can be found in Jewi, Kule, and Tierkidi camps.   FIGURE B.253    Vulnerable population by status and country of origin 100 80 % of households 60 40 20 0 Host South Sudan Eritrea Somalia Sudan & other Self-reliant Productive but poor Support-dependent Source: Authors’ calculation using the SPS, Ethiopia (2017). Typology of Refugees 369.  The analysis for Ethiopia identified two groups of refugees with different profiles. Group 1 represents 78 percent of the refugee population in Ethiopia, while Group 2 represents 22 percent (Figure B.254). The displace- ment situation of each group is different according to the conflict context in each country of origin. Group 1 is mainly composed of households displaced by armed conflict from South Sudan, while refugees from Group 2 come mostly from Eritrea and were displaced for various reasons. Before displacement, Group 1 was more inclined to an agricultural livelihood and Group 2 to depend on aid and remittances. Currently, Group 1 has larger households which are less likely to have Ethiopian relatives and are usually headed by women without education. Group 2 is more likely to have better access to services and to rely on aid and remittances, yet Group 1 is poorer, poverty deeper, and more likely to face food insecurity. As for the future, most households in Group 1 prefer to stay in the camp for security reasons, while most refugees in Group 2 intend to move to another country guided by access to land, ser- vices, and employment. Volume B: Country Case Studies  | 183  FIGURE B.254    Visualization of groups from the clustering analysis Source: Authors’ calculations using SPS 2017. Note: Group 1 is represented by the blue circles and Group 2 by the black triangles. Cause Profile 370.  Group 1 is mainly composed of households displaced by armed conflict from South Sudan, while refu- gees from Group 2 come mostly from Eritrea and were displaced for various reasons. Displacement situations in Ethiopia result from a combination of protracted conflicts in neighboring countries (Somalia, Eritrea, and Sudan) and more recent crises (South Sudan and Yemen). Most households in Group 1 come from South Sudan (59 percent) and Somalia (28 percent), while those from Group 2 from Eritrea (59 percent) and Somalia (28 percent) (Figure B.255). About 9 out of 10 households in Group 1 were displaced in or after 2011, compared to 6 out of 10 households in Group 2 (Figure B.256). The date is possibly associated to the independence of South Sudan and subsequent civil war in the country. Refugees in Group 1 were twice as likely to be displaced due to armed conflict (79 percent vs. 36 percent), whereas insecurity, and discrimination and prosecution was 3 and 6 times more likely, respectively, for refugees from Group 2. The displacement situation of both groups is different according to the conflict context in South Sudan, Eritrea, Somalia, and Sudan. 371.  Before displacement, Group 1 was more inclined to an agricultural or agro-pastoralist livelihood, while households in Group 2 were more likely to rely on aid and remittances. About 69 percent from Group 1 had access to agricultural land before displacement, compared to 48 percent of refugees in Group 2.240 As a result, nearly 6 out of 10 households in Group 1 had an agricultural livelihood before displacement, compared to around 4 out of 10 from Group 2. The main sources of income for around 1 out of 3 households in each group were wages, salaries, and own business. Refugees in Group 2 were three times more likely to rely on aid, remittances, and other sources of income before (Figure F3 in Appendix F). The differences in main source of income before displacement between both groups are robust after controlling for region effect and other household characteristics (Table F10 in Appendix F). 240. Having access to land does not necessarily imply ownership. 184  |  Informing Durable Solutions for Internal Displacement   FIGURE B.255    Country of origin   FIGURE B.256    Year and reason for displacement 100 100 90 80 80 70 % of households % of households 60 60 50 40 40 30 20 20 10 0 0 Group 1 Group 2 In/after 2011 Armed conflict Insecurity Discrimination/ persecution South Sudan Somalia Year | Reason for displacement Eritrea Sudan & other countries Group 1 Group 2 Source: Authors’ calculations using SPS 2017. 372.  Access to services and relationships with neighbors were similar for both groups, yet Group 2 was twice as likely to be satisfied with living conditions before displacement. Access to improved water sources, sanitation, and electricity was similar for Groups 1 and 2 before displacement. However, 48 percent of households in Group 2 lived in an overcrowded dwelling before displacement, compared to 29 percent from Group 1 (Figure F2 in Appendix F). A good relationship with neighbors was a common characteristic of most households in both groups, although 73 per- cent of refugees from Group 1 were satisfied with their life against 42 percent of households in Group 2. Needs Profile 373.  Group 1 has larger households that are less likely to have Ethiopian relatives and are usually headed by women without education, compared to refugees in Group 2. Households in Group 1 have 5.9 household mem- bers and an age-dependency ratio of 2:1, compared to 5 members and a dependency ratio of 1:2 among refugees in Group 2 (Table F11 in Appendix F).241 A larger dependency ratio comes from differences in the number of working-age members between both groups, since they have a similar share of children and elderly in the household. In addition, households in Group 2 are more likely to have a direct connection with the country as 37 percent of them have an Ethi- opian relative, against only 8 percent of households in Group 1. There is a large sex imbalance in terms of the headship of the household between the groups. About 71 percent and 50 percent of households in Groups 1 and 2, respec- tively, are headed by women. Those in Group 2 are almost twice as likely to have some formal education compared to household heads from Group 1. The different profile of the household head between both groups is significant after controlling for other household characteristics and regional effects (Table F12 in Appendix F). 241. The age dependency ratio is defined as the proportion of children and old age dependents to working-age population (15–64 years). Volume B: Country Case Studies  | 185   FIGURE B.257    Current assistance and food  FIGURE B.258    Current poverty status insecurity 80 100 70 90 60 80 % of households 70 50 % of households 60 40 50 30 40 20 30 20 10 10 0 Poverty incidence Poverty gap 0 (% of population) (% of the poverty line) Received Received High food remittances assitance insecurity Group 1 Group 2 Group 1 Group 2 Source: Authors’ calculations using SPS 2017. Source: Authors’ calculations using SPS 2017. 374.  In the current settlement, Group 2 is more likely to have electricity and improved sanitation but more likely to live in overcrowded dwellings. Most households (94 percent) live in shelter provided within the camp and thus are close to main services and most have improved water sources. However, Group 2 is more likely to have elec- tricity (36 percent vs. 1 percent for Group 1), improved sanitation (80 percent vs. 64 percent for Group 1) and less likely to share the toilet facilities (57 percent vs. 71 percent for Group 1; Figure F4 in Appendix F). Moreover, Group 2 is also less likely to live in an overcrowded dwelling (49 percent) compared to refugees in Group 1 (62 percent).242 Differences between both groups are significant after controlling for other household characteristics (Table F13 in Appendix F). 375.  Group 2 continues to rely more on aid and remittances, yet Group 1 is poorer, poverty deeper, and more likely to face food insecurity. About 32 percent of refugees in Group 2 received remittances compared to only 2 per- cent of households in Group 1. In addition, 90 percent of households in Group 2 also received assistance from NGOs, development partners, or the government, compared to 74 percent of refugees from Group 1 (Figure B.257). The share of population consuming below the standard international monetary poverty line is different between groups of ref- ugees (Figure B.258).243 Poverty incidence is 27 percentage points larger for refugees in Group 1 compared to those in Group 2. Among the poor, poverty is also deeper for the population in Group 1 (30 percent vs. 14 percent for Group 2). Hence, households in Group 1 are more than twice as likely to face food insecurity. The differences in poverty incidence are significant after controlling for region effects and characteristics of the household head (Table F12 in Appendix F). Eritreans—most households in Group 2—enjoy more rights compared to refugees from other countries, and thus have an advantage in accessing income-generating activities, which could explain a lower poverty incidence for this group of households. 242. Overcrowded dwellings are defined as those that have more four or more people per room. 243. The poverty line corresponds to a daily value of US$1.90 PPP per day. 186  |  Informing Durable Solutions for Internal Displacement   FIGURE B.259    Current perceptions Feel safe from crime & violence Good living conditions today Expect better conditions in the future 0 10 20 30 40 50 60 70 80 90 100 % of households Group 1 Group 2 Source: Authors’ calculations using SPS 2017. 376.  Refugees in Group 2 have a better safety perception in their current location and are more optimistic about the future. Around half of households in both groups have a similar perception of good current living condi- tions (Figure B.259). Nevertheless, Group 2 is more likely to feel safe from crime and violence (93 percent vs. 74 percent in Group 1), and they are also more optimistic about the future (60 percent vs. 31 percent in Group 1). Safety concerns could be driven by a previous exposure to armed conflict, since most households in Group 1 were displaced for this reason. A better perception about the future in Group 2 relative to Group 1 is probably associated with a lower poverty incidence and greater access to aid and support from development partners, as well as a better perception of safety. Solutions Profile 377.  Most households in Group 1 who prefer to stay in the camp are motivated by security reasons, while most refugees in Group 2 intend to move to another country and are guided by access to land, services, and employment. The profile of both groups of refugees is different in terms of return intentions. About 60 percent of households in Group 1 want to stay in their current location, compared to only 10 percent of households in Group 2 (Figure B.260). Among those who would want to relocate in Group 2, the majority prefer to move to a new country (77 percent), and only some (11 percent) want to return to their country of origin. Group 1 is mostly motivated by security reasons (48 percent vs. 9 percent from Group 2), while access to land, services, and employment opportunities are more important for Group 2 (67 percent vs. 28 percent from Group 1, Figure B.261). For both groups, security is the main reason for not returning to their country of origin, yet Group 2 is more likely to also be motivated by other factors such as discrimination or prosecution (18 percent vs. 5 percent in Group 1) and access to land, services, and employ- ment (13 percent vs. 3 percent from Group 1). Insecurity was less likely to be the cause of displacement for households in Group 2, and thus they are less likely to currently consider insecurity as a key factor to stay or move. Furthermore, a desire to move to a new country seems to be associated with a higher satisfaction before displacement and a better perception about the future. Volume B: Country Case Studies  | 187  FIGURE B.260    Return intention 100 80 % of households 60 40 20 0 Stay in Move to Return to country Move to a current camp another camp of origin new country Group 1 Group 2 Source: Authors’ calculations using SPS 2017. 378.  Despite being more certain of wanting to move and having all the information required, refugees from Group 2 are less likely to have a clear timeline. Uncertain of their current situation, many refugees are unable to plan for the future. Among those households looking to relocate, the majority from both groups do not have a clear timeline for moving. However, a larger share of refugees from Group 1 reported their intention to move in less than 12 months (39 percent vs. 23 percent from Group 2, Figure B.262). Households in Group 2 are surer they would like to relocate and perceive to have all the information they need to make this decision (78 percent) against refugees in Group 1 (63 per- cent). Besides, Group 2 is more likely to receive information from radio, TV, the Internet, or written media (10 percentage points more) while households in Group 1 rely on family, friends, or leaders (16 percentage points more, Figure F5 in Appendix F). A higher exposure to armed conflict for refugees in Group 1 might explain the use of informal sources and the need for more security information (21 percent vs. 9 percent from Group 2).   FIGURE B.261    Reasons for moving or staying  FIGURE B.262    Timing of moving 80 70 70 60 60 % of households % of households 50 50 40 40 30 30 20 20 10 10 0 0 Security Access to land, Other In less than In more than Don't know reasons services, and reasons 12 months 12 months employment Group 1 Group 2 Group 1 Group 2 Source: Authors’ calculations using SPS 2017. Source: Authors’ calculations using SPS 2017. 188  |  Informing Durable Solutions for Internal Displacement Policy Implications of the Typologies 379.  Each group of refugees can be identified from their current location, country of origin, and return inten- tion. Refugees in Group 1 are mainly South Sudanese displaced by armed conflict, and more than half of them are in the camps of Jewi, Kule, and Tierkidi in Gambella, the southwestern border of Ethiopia. Households in Group 2 are mainly Eritreans, and more than half of them settled in the Mai-Aini and Adi Harush camps along the northern border in the Tigray region, and in the Kebribeyah camp in the Somali region of eastern Ethiopia. Identifying these two groups of refugees is relevant to target different policy responses given their different needs and durable solutions. In addition, most households in Group 1 want to stay in their current settlement, while most of those in Group 2 want to move to a new country. 380.  Refugees in Group 1 would benefit from gender-responsive programs, increased access to safety nets, and interventions that focus on armed conflict–related trauma. About 7 out of 10 households are headed by a woman in Group 1, and so policy efforts should consider gender-based vulnerabilities related to domestic work and caring labor, in addition to GBV and discrimination. Group 2 also seems to have less access to safety nets since they are less likely to receive aid, perhaps because South Sudanese enjoy less rights compared to other refugees. Assistance from development partners and NGOs could perhaps aim to target South Sudanese households headed by women. The GoE should continue with the implementation of their strategic approach to improve the rights and expand ser- vices to benefit households in Group 2, the majority of whom want to stay in their current location and require bet- ter conditions to ultimately bring a durable solution to their displacement. Moreover, supporting a conducive legal framework to enable self-reliance is crucial for this group of households as they have a larger number of working-age members. In addition, most refugees in Group 1 were displaced by armed conflict and witnessed trauma. Cognitive and non-cognitive approaches that can help overcome conflict-induced and displacement-induced trauma are particularly relevant for refugees in Group 1. 381.  Policy efforts for households in Group 2 require improving their skills and supporting their relocation. Skills and human capital are pertinent for displaced populations due to the loss of physical capital during their dis- placement. Additionally, protracted periods of displacement and economic inactivity make it harder for refugees to find employment. Households in Group 2 were more likely to rely on assistance before displacement. Currently, they are also highly dependent on aid, despite refugees in this group enjoying more rights—due to the Eritrean composition of the group. In addition, the decision to stay or move for Group 2 is mainly guided by access to land, services, and employment. Therefore, policy efforts should aim to upgrade their skills and support an income diversification strategy to improve their living standards. Furthermore, households in Group 2 want to relocate but do not have a clear timeline, despite claiming to have all the relevant information. Thus, a strategic approach to find a durable solution for this group of refugees should aim to facilitate their relocation process, beyond providing relevant information. Conclusions Informing Durable Solutions 382.  Countries such as Somalia, South Sudan, and Sudan, which have large IDP populations, also have large numbers of refugees in Ethiopia. The refugees, largely driven by conflict and staying in border areas, face differing conditions than IDPs in important respects. Seeking refuge across international borders leads to differing legal status, including in terms of documentation and rights to work. Local conditions also have a bearing on levels of consump- tion and service access, leading to differences in the living conditions of IDPs and refugees. The Ethiopia comparative Volume B: Country Case Studies  | 189 refugee study complements the results and recommendations for IDPs with results and recommendations for refugees in Ethiopia, including from Somalia, South Sudan, and Sudan. 383.  Over the last few years, Ethiopia witnessed a dramatic increase of refugees (and IDPs), which led to a shift in its approach to forced displacement. The number of refugees has risen tenfold in the last decade, to almost 1 mil- lion refugees. Nearly half a million South Sudanese refugees settled in Ethiopia in the last five years alone. IDP trends are booming too: from an official 258,000 in 2016 the number of conflict-induced IDPs climbed up to an estimated 1.1 mil- lion. While IDPs are not part of the present case study, nonetheless Ethiopia faces compounding humanitarian crises and protracted displacement that put a strain on domestic actors’ coping capacities and question the sustainability of the current reception system. As a result, the GoE unveiled a strategic plan (‘nine pledges’) to address the development aspects of the refugee situations in the medium to long term. Currently under implementation, the multi-sectoral ini- tiative aims to improve the material condition of refugees, to expand their rights, and to support synergies and local integration mechanism with host communities. 384.  The micro-level study on refugees in Ethiopia and host communities is meant to support the implemen- tation of the new strategy. It provides comprehensive findings on several measures of poverty at the household level, putting in relationship different groups of refugees in Ethiopia and their respective host communities. The analysis found that not only refugees witnessed trauma and life disruption (for example, family separation), and incurred mate- rial loss during displacement (that is, land, livestock, and assets), but also their standard of living, livelihood, and access to services’ prospects are currently highly dependent on aid. Uncertainty and inability to plan for the future represent the common denominators among refugees, which is the reason why a conducive legal framework and related policies to enable refugees’ self-reliance are key. 385.  Eritreans, Somalis, South Sudanese, and Sudanese were displaced due to different drivers related to conflict and fragility, and each group experiences different displacement dynamics while in Ethiopia. While all groups except Eritreans reported that armed violence and a general sense of insecurity were the chief drivers of displacement, specific conflict drivers include localized conflicts and a fallout in law and order (Somalis and Suda- nese), environmental degradation leading to strain in resources (Somalis), political persecution (Eritreans), and a full blown civil war (South Sudanese). The timeline of displacement is quite different from group to group: while South Sudanese are the newest group (as they are also coming from the newest country), the other three groups have had some repeated patterns of displacement in the last three decades. Geographically, refugees are compartmentalized according to nationality: each group settled in the border area in proximity of the respective country of origin. As a result, refugees have specific development needs that should be put in relation to the respective Ethiopian region of displacement. Border regions in Ethiopia are traditionally poorer and more marginalized than the center, although the economic gap center-periphery has been closing during the last decade. 386.  Overall, Eritreans are better off compared to other refugee groups, with indicators that are similar to host community ones. In fact, according to nationality, refugees fare differently with respect to standard of living, livelihood and employment, and ties to host communities. Among refugee groups, Eritreans are the ones that enjoy more rights compared to others, and, as a result, display higher standards of living and much lower poverty rates. In turn, less than 30 percent of Eritreans experience high food insecurity, while the other three groups have rates ranging between 60 percent and over 80 percent. Similarly, trends in housing conditions and overcrowding highlight Eritreans’ better 190  |  Informing Durable Solutions for Internal Displacement situation with respect to other groups. On the other hand, South Sudanese are the poorest group on many indicators. Arguably, the length of displacement plays an important role when assessing not only refugees’ standard of living but also social capital and relationships with host communities. Longer interactions with refugees–host communities seem to predict better relationships going forward. 387.  Indicators on livelihood and access to services highlight the nearly complete dependence of refugees on aid. Over 90 percent of Eritrean, South Sudanese, and Sudanese households rely on aid as their greatest source of live- lihood, while Somalis are the least aid dependent and they have the lowest rate of food consumption coming from aid. Arguably, among the four refugee groups, Somalis are the most integrated group with host communities. From higher rates of labor force participation to slightly higher rates of home ownership, stronger feelings of safety and security, and better host populations’ perceptions, Somalis display specific displacement dynamics that are different from those of other groups. Access to services for all refugees compares and, in some cases, is even better than host communities—a situation that underlines the relative high quality of camp management compared to domestic service delivery mecha- nism. These findings directly feed into the need to support policies of job creation, enhanced and sustainable access to markets and services, and conducive regulatory environment and governance, among others. Virtuous practices—that organically take into consideration refugees as well as host communities—are functional to economic self-reliance and to move away from dependency on external sources of aid. These findings also call for the need for more granular understanding of livelihood sources among refugee groups: for example, it would be important to understand why and how Somalis differ from the other three refugee groups in terms of aid dependency. 388.  Ultimately, refugees showcase perceptions of uncertainty which lead them to an inability to plan for the future. As the main driver of displacement was personal security, likewise, all refugee groups regard security as the utmost priority when making considerations about future intentions, prospects, and/or durable solutions. Overall, one in two refugees prefer to stay in the current location, which is an indication of their inability to plan. Interestingly, South Sudanese—the poorest among the four—have the highest percentage of respondents who do not want to move. This fact suggests that the trauma they have suffered from is still fresh and ongoing, and that staying put is the most favorable option despite a poor standard of living. Eritreans and Somalis have, instead, large percentages of refugees who want to move to a new country, while over one in two Sudanese want to return to Sudan. These contradicting intentions—which run opposite to standard of living across the four groups—are indicative of the fundamentally dif- ferent situations that refugee groups currently face. 389.  More analysis is needed to understand the ad hoc development needs and durable solution prospects of the four refugee groups against host communities. Given the above findings, it is key to better understand each refugee groups’ dynamics relative to the respective host population, who, in turn, live in regions of Ethiopia that have remarkable differences and are far away from one another. The SPS collected data on refugees and host communities in Tigray and Afar, Somali, and Benishangul Gumuz. The present paper analyzed and compared refugee groups among them, and against an average host community. There is scope to zoom in on each specific host community to make more meaningful comparisons between each refugee group and its respective host community. Such findings would give more explanatory power to each of the four displacement situations in Ethiopia, and offer more specific policy insights. Cross-Country Analysis Agricultural IDPs Introduction 390.  Large numbers of displaced populations flee from rural areas to urban centers.244 Conflict, violence, and cumulative effects of climate phenomena such as drought often drive IDPs from rural areas to the closest safe location. IDPs tend to concentrate in urban areas, which often have camps or settlements for the displaced. Following drought and rural violence during 2016 to 2017, forced displacement contributed to make Mogadishu, the capital of Somalia, the most densely populated city in Africa.245 Once populations make a transition to urban areas, it is uncommon for them to return to rural settings.246 The rapid, unplanned urbanization puts a strain on jobs, infrastructure, and urban service delivery for both the displaced and the non-displaced. 391.  Displaced populations coming from rural areas often have an agricultural background.247 Rurally triggered displacement is often associated with livestock raiding and the capture of land, in addition to the loss of homes and assets. Many IDPs who have an agricultural or pastoralist background thus lose out on livelihood-generating assets. A lack of transferrable skills makes adjusting to urban labor markets especially difficult, adding to the challenges of livelihood and asset loss. Proposed development agendas include providing support to agricultural programs where feasible, to create opportunities for low-skilled hosts as well as IDPs with an agricultural background. While this is more feasible in areas conducive to agriculture, the program mix should also include a focus on building transferrable skills that are relevant to urban labor markets.248 392.  This section focusses on whether IDPs with an agricultural background face greater challenges in urban centers. Among the five countries analyzed, Somalia, South Sudan, and Sudan have IDPs largely in urban camps. In Nigeria, the IDPs are in the northeastern states with the entire state represented, while in Ethiopia, the refugees are in border areas close to their country of origin. Comparisons among these cases can lead to insights on the conditions facing agricultural IDPs or refugees. An agricultural IDP or refugee is defined as belonging to a household whose pri- mary source of livelihood at the original residence was in own-account agriculture. IDPs and refugees with any other livelihood background, such as wages and salaries, business, or remittances and aid, are grouped as ‘nonagricultural’. 244. World Bank. 2017. “Forcibly Displaced: Toward a Development Approach Supporting Refugees, the Internally Displaced, and Their Hosts.” 245. Internal Displacement Monitoring Center. 2018. “UnSettlement: Urban Displacement in the 21st Century; City of Flight.” 246. World Bank. 2017. “Forcibly Displaced: Toward a Development Approach Supporting Refugees, the Internally Displaced, and Their Hosts.” 247. Ibid. 248. Ibid. 191 192  |  Informing Durable Solutions for Internal Displacement Livelihood and Employment Shifts 393.  IDPs and refugees relied more heavily on agriculture before displacement than their hosts do now. 249 A large proportion of the displaced population, 42 percent, relied on own-account agriculture as the primary source of household livelihood before displacement. In contrast, 26 percent of hosts rely on agriculture currently. The differences between the livelihood structures are driven by the business sector and other miscellaneous jobs. Hosts are more likely to be involved in business, services, or retail than the displaced at origin. Hosts are also twice as likely to be in other mis- cellaneous jobs (13 percent) than the displaced (6 percent). There is a clear shift in the labor market environment of the displaced, away from agriculture. Those who relied on wages or business might find it easier to integrate as significant proportions of the hosts are employed in these sectors at the IDPs’ current locations (Figure C.1).   FIGURE C.1    Livelihoods of the displaced at origin, compared to host communities now 100 80 % of households 60 40 20 0 Host Displaced Host Displaced Host Displaced Host Displaced Host Displaced Host Displaced Overall Ethiopia Nigeria Somalia South Sudan Sudan Agriculture Wages and salaries Business/trade/services Remittances Aid Other Source: Authors’ calculations using SPS Ethiopia 2017, IDP Survey Nigeria 2018, Somali HFS 2017–18, HFS South Sudan 2017, and IDP Survey Sudan 2018. 394.  The overall shift in the livelihood landscape away from agriculture is mirrored in each country apart from South Sudan. Across the countries, livelihood structures between the displaced at origin and hosts now differ to vary- ing degrees. However, hosts’ lower reliance on agriculture is a common pattern across the cases. The exception is South Sudan, where the urban non-displaced population is as reliant on agriculture as IDPs were at the origin (41 percent each). IDPs’ pre-displacement reliance on agriculture also varies across the countries, from 15 percent in Somalia to 90 percent in Sudan. Apart from Somali IDPs, a large number of the displaced had agricultural backgrounds, with sig- nificant proportions of those in Nigeria, Ethiopia, and South Sudan (ranging from 41 percent to 56 percent) depending on agriculture before displacement (Figure C.1). 395.  Depending on the context, different sectors take precedence over agriculture in the five countries. In Nigeria, internal displacement is largely within the northeast region, from where both IDPs and hosts hail. However, hosts rely significantly more on wages (32 percent compared to 19 percent IDPs). In Somalia, along with agriculture, hosts also rely less on business than IDPs. Instead, they are more reliant on remittances (27 percent compared to 5 percent IDPs). While remittances are prevalent in Somalia, they do not seem to be used as a safety net for displacement—only 249. In the case of South Sudan, ‘host’ refers to urban populations in South Sudan, rather than explicit host communities. In all other countries, the hosts are host communities living near and around the IDPs/refugees. Volume B: Country Case Studies  | 193 7 percent of IDPs rely on remittances currently.250 In South Sudan, the livelihood structures of IDPs at origin and hosts are almost identical, indicating that IDPs, like the hosts, covered in the study come from urban backgrounds. In Sudan, IDPs face the starkest livelihood shift among the five countries. Almost all the IDPs were agricultural in Sudan (90 per- cent), while now they are in an urban environment with heavy salaried labor and business sector reliance. In Ethiopia, the hosts rely on business rather than agriculture (Figure C.1). 396.  Agricultural IDPs’ and refugees’ employment status differs starkly across the five countries. Overall, agri- cultural IDPs and refugees are as likely to be employed as nonagricultural displaced despite facing significantly different labor markets in their new locations. However, employment and labor force trends vary vastly with the country con- text. In Nigeria, agricultural and nonagricultural IDPs are equally likely to be employed (64 percent each). This relatively high employment rate might be because hosts might be renting out agricultural land to the IDPs to work.251 However, among the IDPs not participating in the labor force, agricultural IDPs are more likely to be ‘idle’, neither enrolled in edu- cation nor looking for work. In Somalia, agricultural IDPs seem to have better labor market outcomes than their nonag- ricultural counterparts. They have higher employment rates, and lower rates of being ‘idle’. In South Sudan, agricultural IDPs have lower unemployment, but are more often ‘idle’. In Sudan, where agricultural IDPs face very different labor market environments, agricultural IDPs are in fact more likely to be employed, and less likely to be ‘idle’. In Ethiopia, agri- cultural IDPs have lower employment and higher unemployment than their nonagricultural counterparts (Figure C.2). Thus, despite facing very different and new labor market structures, agricultural IDPs are more engaged as employed workers, less ‘idle’ in Somalia and Sudan, have mixed outcomes in Nigeria and South Sudan. Ethiopia is the exception, where agricultural refugees face unequivocally worse labor market outcomes than nonagricultural refugees.   FIGURE C.2    Labor force participation status of working-age IDPs 100 % of working-age population 80 60 40 20 0 Non-agri Agri Non-agri Agri Non-agri Agri Non-agri Agri Non-agri Agri Non-agri Agri Overall Ethiopia Nigeria Somalia South Sudan Sudan Not enrolled Enrolled Unemployed Employed Source: Authors’ calculations using SPS Ethiopia 2017, IDP Survey Nigeria 2018, Somali HFS 2017–18, HFS South Sudan 2017, and IDP Survey Sudan 2018. 397.  Agricultural IDPs have adjusted to current labor markets in different ways depending on country con- text. Across the five countries, agricultural IDPs and refugees continue to be more dependent on agriculture today than the nonagricultural displaced. In South Sudan, a majority of IDPs depend on aid for their livelihood, but agricultural IDPs are more aid-dependent. In Somalia, where remittances are prevalent, only a negligible proportion of agricultural IDPs 250. See Somali case study. 251. See Nigeria case study. 194  |  Informing Durable Solutions for Internal Displacement rely on remittances, compared to 8 percent of nonagricultural IDPs. While this is a low rate, especially compared to hosts (27 percent rely on remittances), it indicates that agricultural IDPs have even lower access to remittance infrastructure than hosts or nonagricultural IDPs. With their continued dependence on agriculture, agricultural IDPs in Nigeria, Soma- lia, South Sudan, and Sudan are still less likely to rely on wages or business sectors, than nonagricultural IDPs. Ethiopia is an exception, where agricultural refugees have the same livelihood structure as nonagricultural refugees and have very low reliance on agricultural activities (Figure C.3). While IDPs still have some access to agricultural means of livelihood, refugees largely turn to aid as a feasible income source.   FIGURE C.3    Current livelihoods of IDPs and refugees 100 80 % of households 60 40 20 0 Non-agri Agri Non-agri Agri Non-agri Agri Non-agri Agri Non-agri Agri Non-agri Agri Overall Ethiopia Nigeria Somalia South Sudan Sudan Agriculture Wages and salaries Business/trade/services Remittances Aid Other Source: Authors’ calculations using SPS Ethiopia 2017, IDP Survey Nigeria 2018, Somali HFS 2017–18, HFS South Sudan 2017, and IDP Survey Sudan 2018. Poverty, Social Cohesion, and Return Intentions 398.  Agricultural IDPs are often poorer than nonagricultural IDPs. Though displaced populations face a high poverty incidence across the five countries, the agricultural IDPs are poorer (82 percent compared to 75 percent nonag- ricultural IDPs). Depending on country context, agricultural IDPs are either as poor or poorer than nonagricultural IDPs. In Somalia and South Sudan, the agricultural IDPs are significantly poorer, while in Ethiopia, Nigeria and Sudan, agricul- tural and nonagricultural IDPs and refugees have a similar poverty incidence (Figure C.4). The higher poverty associated with an agricultural background might be driven, at least in part, by lower income since before displacement. However, even though IDPs have adapted livelihood structures to their new environments, the lack of transferrable skills in urban environments can make it especially difficult for agricultural IDPs to find sustainable livelihoods to lift themselves out of poverty. Poverty reduction programs integrated with livelihood opportunities and transferable skills programs can be instrumental in helping agricultural IDPs find a stable exit from poverty. 399.  Though IDPs from a rural and agricultural background face very different livelihood prospects in their new environments, they still enjoy good relationships with host communities. Overall, agricultural IDPs were more likely to report relations with surrounding host communities as ‘very good’ or ‘good’, even after accounting for country fixed effects. In each country, a large majority of IDPs report positive relations with hosts, but differences in this perception, though small, persist. Nonagricultural refugees of Ethiopia report worse host relations than the agricultural refugees. In Nigeria, South Sudan, and Sudan, the trend is reversed, with agricultural IDPs significantly less likely to report ‘very good’ or ‘good’ relations. In Somalia, there is no statistical difference in the relationship perceptions of the two groups of IDPs (Figure C.5). Thus, agricultural IDPs seem to have comparable or better social cohesion with hosts than nonagricultural IDPs. Volume B: Country Case Studies  | 195   FIGURE C.4    Poverty headcount ratio for IDPs 100 80 % of individuals 60 40 20 0 Non-agri Agri Non-agri Agri Non-agri Agri Non-agri Agri Non-agri Agri Non-agri Agri Overall Ethiopia Nigeria Somalia South Sudan Sudan Source: Authors’ calculations using SPS Ethiopia 2017, IDP Survey Nigeria 2018, Somali HFS 2017–18, HFS South Sudan 2017, and IDP Survey Sudan 2018.   FIGURE C.5    IDPs’ perception of relations with surrounding host communities 100 80 % of households 60 40 20 0 Non-agri Agri Non-agri Agri Non-agri Agri Non-agri Agri Non-agri Agri Non-agri Agri Overall Ethiopia Nigeria Somalia South Sudan Sudan Very bad Bad Neither Good Very good Source: Authors’ calculations using SPS Ethiopia 2017, IDP Survey Nigeria 2018, Somali HFS 2017–18, HFS South Sudan 2017, and IDP Survey Sudan 2018. 400.  Agricultural IDPs are keener to return to the original residence than nonagricultural IDPs. Overall, about 35 percent of agricultural displaced want to return, compared to 20 percent among the nonagricultural group. The over- all result is largely driven by country-level differences (Table C.1). In Nigeria, agricultural IDPs are twice as likely to prefer a return to the origin (56 percent compared to 26 percent nonagricultural IDPs). Even in South Sudan, where the livelihood structures of the urban and IDPs-at-origin are not very different, the agricultural IDPs have a stronger keenness to return (38 percent compared to 31 percent nonagricultural IDPs). In Somalia and Sudan, the return intention patterns are similar for agricultural and nonagricultural IDPs. Even though Sudanese IDPs have seen the starkest shift away from agriculture in their new labor environments, the protracted nature of displacement (15 years on average, see Sudan case study) might have induced IDPs to gradually adapt their skill set so that differences among agricultural and nonagricultural IDPs are no longer as pronounced. In Ethiopia, agricultural refugees have a stronger preference to return to origin as well as stay; nonagricultural refugees are keener to move to a new location (Figure C.6). Agricultural IDPs’ stronger preference for a return might reflect a desire to restore agricultural livelihoods, for which land and livestock are required, and for which urban settings are not conducive. The self-reported return intentions indicate that rural return programs might be more conducive for agricultural IDPs, while nonagricultural IDPs might be more open to local urban integration programs. 196  |  Informing Durable Solutions for Internal Displacement   TABLE C.1    Differences in overall IDP outcomes with country fixed effects Employment status In labor force Not in labor force Return intention Not Social Employed Unemployed Enrolled Enrolled cohesion Stay Return Agricultural IDP or  0.079  0.145 −0.102* −0.053  0.208* −0.069 −0.248 refugee Nigeria  1.167*** −0.254* −0.476*** −0.822***  0.554***  0.222 −1.725*** Somalia  0.660*** −0.340** −0.731*** −0.054  0.426***  0.477*** −1.178*** South Sudan  0.570*** −0.157** −0.205*** −0.317***  0.197**  0.172*** −1.023*** Sudan  0.905*** −0.301** −0.171** −0.725***  1.363***  0.106 −1.576*** Constant −0.804*** −1.594*** −0.554*** −0.121**  0.675***  0.063 −0.329*** Observations 27,258 27,258 27,258 27,258 10,331  9,917  9,917 Source: Authors’ calculations using SPS Ethiopia 2017, IDP Survey Nigeria 2018, Somali HFS 2017–18, HFS South Sudan 2017, and IDP Survey Sudan 2018. Note: *** p < 0.01, ** p < 0.05, * p < 0.1   FIGURE C.6    IDPs’ return intentions 100 80 % of households 60 40 20 0 Non-agri Agri Non-agri Agri Non-agri Agri Non-agri Agri Non-agri Agri Non-agri Agri Overall Ethiopia Nigeria Somalia South Sudan Sudan Stay here New location Return to origin Source: Authors’ calculations using SPS Ethiopia 2017, IDP Survey Nigeria 2018, Somali HFS 2017–18, HFS South Sudan 2017, and IDP Survey Sudan 2018. Volume B: Country Case Studies  | 197 IDPs in Camps Introduction 401.  IDPs and refugees in camps are often believed to receive better services through international humani- tarian actors than those available locally through country systems.252 While service delivery and security through camps is required for an immediate humanitarian response, other approaches would be more appropriate for a medium- to long-term developmental response. Shifting IDPs and refugees into country systems and markets and allowing them to benefit from and contribute to socioeconomic systems locally, is likely to be more cost-effective and equitable in the medium term.253 If humanitarian actors, particularly the international community, are providing better service access and quality in camps, then shifting to country systems could cause a net loss in welfare.254 402.  This section aims to study how the socioeconomic outcomes of IDPs inside and outside camps differ. Analysis on service delivery and poverty, where camps could provide stronger outcomes, is complemented with out- comes in labor force participation and social perceptions, which can be a challenge for sequestered populations. Since factors such as income opportunities, social cohesion, and family connections influence IDPs’ opportunities to locate within or outside camps, the analysis does not present a causal effect of camps on IDP outcomes. Rather, the aim is to describe whether the camp-based displaced differ on key welfare measures, and to account for pertinent differences in policy recommendations to support IDPs and refugees to achieve durable solutions. 403. The comparison groups differ across the five countries analyzed, allowing for insights on different dimensions. All five countries have data that represent camp-based IDPs (or refugees in the case of Ethiopia) and non-displaced populations. Further, data on IDPs located outside camps is analyzed for Nigeria and Somalia, allowing for a contrast of camp IDPs with those outside camps, in addition to host communities. While in the other countries, IDPs have been displaced for less than five years on average, in Sudan, the displacement is protracted, with the average IDP household having been displaced for fifteen years. The camps analyzed in Sudan have acquired settlement-like features, providing an additional lens to analyze how camps that are semi-permanent affect service delivery and labor market outcomes. Service Access and Living Standards 404.  Contrary to notions of better service provision in camps, service access is worse for camp IDPs in four of the five countries. Camp-based IDPs in Nigeria have lower enrollment rates of school-age children, are farther away from the nearest market, and face more prevalent overcrowding than hosts. In Somalia, both camp and non-camp IDPs have lower school enrollment and improved sanitation facilities than hosts, and camp IDPs are additionally farther away from the closest health facility, primary school, and market. This contradicts the notion that camps provide proxi- mal access to key amenities. In South Sudan, camp IDPs have lower school enrollment but higher quality of water and sanitation facilities, and greater proximity to key amenities like the closest primary school, water point, health facility, or market. However, they still face higher poverty. In Sudan, though camp IDPs have better access to improved water, they 252. World Bank. 2017. “Forcibly Displaced: Toward a Development Approach Supporting Refugees, the Internally Displaced, and Their Hosts.” 253. World Bank. 2017. 254. Ibid. 198  |  Informing Durable Solutions for Internal Displacement have lower school enrollment and are farther from health and educational facilities than the host community. In Ethio- pia, camp-based refugees only have better sanitation facilities, but worse proximity to the closest market and primary school (Table C.2). The lower school enrollment (four in five countries) and farther access to the nearest market (three in five countries) indicate threats to socioeconomic integration with surrounding communities, in goods markets, labor markets, and future work prospects (through education). 405.  Even in cases where camps provide higher quality facilities, overcrowding deteriorates both service access and service quality, bringing overall access to a lower level than hosts. In three of the four countries where overcrowding data were collected, camp-based IDP households are extremely overcrowded, and significantly more so than hosts and non-camp based IDPs. In South Sudan, where facilities are improved and amenities closer among camp IDPs, the difference between overcrowding of hosts and IDPs is also the steepest. To be classified as ‘improved’, sanita- tion facilities should not be shared with other households.255 After accounting for toilet sharing, the rate of improved sanitation falls sharply in camps—more sharply than for hosts. Ultimately, WASH facilities in camps of South Sudan do not necessarily guarantee better quality or access. In Sudan, the overcrowding trend is reversed with IDP households in settlement-like camps being less overcrowded than host households in the city, indicating that overcrowding must be addressed to reach a medium-term solution (Table C.2). 406.  Specific instances of better services are sparse and do not link to lower poverty in camps. Camps in Ethiopia have better sanitation, camp settlements in Sudan have better water sources, and the PoCs in South Sudan have higher quality WASH services and are closer to key amenities. However, IDPs and refugees are significantly poorer than hosts in each of these countries. The difference in the poverty of the displaced and non-displaced is starkest for refugees in Ethiopia, steep in Sudan, and significant even in South Sudan, where 82 percent of the over- all population is poor (Table C.2). Thus, the provision of basic facilities in camps is not linked to higher welfare and lower poverty compared to host communities. Perceptions of host communities about camps and camp services can have an important bearing on social cohesion. Hosts might hold perceptions of the displaced having access to special resources that are denied to them; for instance, in a refugee camp in Ghana, perceptions of discrimination in land and water distributions led to tensions with hosts.256 Sharing information on the true access and conditions of camp-based IDPs and addressing perceptions regarding camp conditions can be an important step toward integra- tion outside and beyond the camp. 255. World Health Organization and UNICEF. 2006. “Core Questions on Drinking Water and Sanitation for Household Survey.” 256. World Bank. 2017. “Forcibly Displaced: Toward a Development Approach Supporting Refugees, the Internally Displaced, and Their Hosts.”   TABLE C.2    Outcomes of camp IDPs, non-camp IDPs, and hosts South Sudan Ethiopia Nigeria (IDPs) Somalia (IDPs) (IDPs) Sudan (IDPs) (Refugees)         Non- Non-     Camp camp Host Camp camp Host Camp Host Camp Host Camp Host Poverty Poor 89.3** 87.8 81.5 75.6 72.5 64.6 91.3*** 75.4 80.7*** 61.7 64.4*** 28.0 and aid- dependence Aid as primary livelihood  8.9***  1.6  1.2 14.3***  7.9**  0.2 75.7***  2.3  4.1***  0.9 82.5***  2.2 Employed 58.1 67.6 60.2 45.7 43.2 52.5 42.0*** 76.1 56.9*** 46.0 22.2*** 61.2 Working-age Unemployed  6.3  2.7  6.1  3.2  2.2  1.6  4.6***  1.5  3.9  4.0  6.3***  2.1 population Not in labor force nor enrolled 23.9 14.3* 17.6 43.2 42.4 36.9 32.4*** 10.4 18.6 22.0 44.4*** 22.6 Enrolled, at school age 42.1* 44.8 49.9 30.7*** 42.2** 62.3 73.0** 78.1 79.2*** 90.4 79.6 75.0 Improved drinking water 92.2 90.2 83.6 78.6 76.0 64.2 99.3*** 82.4 82.8*** 37.1 97.7 95.6 Improved sanitation 78.8 79.8 77.3 87.3** 62.1*** 95.1 78.4*** 56.4 89.0 93.1 67.4*** 48.4 Standard of Near water point 95.6 95.8 96.3 91.0 89.3 89.8 98.8*** 94.3 83.7 80.5 98.3 98.9 living Near health facility 77.3 78.0 75.2 72.6* 76.7 85.7 95.2*** 57.2 68.1*** 84.0 74.2 80.8 Near primary school 84.6 87.1 91.4 78.4** 81.4 89.3 93.1*** 73.6 86.2*** 92.4 75.1*** 92.8 Near market 71.7* 79.1 82.9 78.4** 79.3 90.1 94.1*** 62.4 63.8 71.3 72.2*** 86.6 Overcrowded house 54.2*** 46.7*** 22.9 57.6***  8.8  8.3*** 16.0 58.7*** 32.4 Good relations with hosts 82.6*** 93.0 87.0 87.5 83.1 98.7 78.2 Feel safe from violence 84.1*** 96.3 97.3 96.4 90.1* 96.6 46.8** 53.6 78.6*** 95.3 Perceptions Feel safe walking daytime 90.9*** 98.1 98.4 83.6 88.3 88.4 67.4*** 74.6 94.6 98.0 93.0 95.5 Feel safe walking nighttime 71.1** 78.1** 84.4 84.0* 84.7 93.3 22.3*** 10.6 60.2 86.9 49.1*** 83.2 Return Intend to stay here 22.2*** 72.3 70.4 69.8 58.1 54.3 49.8 intentions Intend to return 75.7*** 26.2 23.4 23.3 34.5 44.0 16.6 Source: Authors’ calculations using SPS Ethiopia 2017, IDP Survey Nigeria 2018, Somali HFS 2017–18, HFS South Sudan 2017, and IDP Survey Sudan 2018. Note: *, **, and *** indicate significance at the 10 percent, 5 percent, and 1 percent levels respectively. Estimates for camp-based IDPs/refugees are compared to estimates for hosts. Where applicable, estimates of non-camp IDPs are compared to hosts. For indicators that are only relevant to IDPs, estimates for camp IDPs are compared to non-camp IDPs where possible. Yellow indicates similar outcomes, red indicates worse outcomes, and green indicates better outcomes, compared to hosts. Volume B: Country Case Studies  | 199 200  |  Informing Durable Solutions for Internal Displacement Labor Force Participation and Social Integration 407.  Labor market outcomes depend heavily on country context, with the camp-based displaced in South Sudan and Ethiopia participating very little in the labor force. While camp-based IDPs have lower educational enrollment in all the four countries, the labor-market patterns are more contextual. In Somalia and Nigeria, camp IDPs, non-camp IDPs, and hosts all have similar levels of employment and unemployment, which together constitute labor force participation. In Sudan, the camp IDPs are more likely to be employed than are host communities. However, camp-based IDPs in all three countries depend more on aid as a primary source of livelihood, than non-camp IDPs and hosts. This could indicate that camp-based IDPs, even if employed, find it more difficult to find self-reliant, revenue- generating livelihoods. In South Sudan and Ethiopia, the camp-based displaced have lower employment and higher rates of being ‘idle’ by neither participating in the workforce nor being enrolled in education. These two countries also have the highest rates of aid dependency as a primary source of livelihood (76 percent in South Sudan and 83 percent in Ethiopia). The opportunities to integrate with the labor markets outside camps depend on country context—security concerns are higher in South Sudan and Ethiopia, with relatively lower proportions of the populations feeling safe from violence, walking in the daytime, and walking during nighttime, than in other countries (Table C.2). Further, in Ethiopia, at the time of the study, refugees faced additional legal challenges around being allowed to work. Program responses that work toward ending ‘continuing limbo’ in camps by supporting labor force rejoining and integration into country systems, can be key to facilitate reliance in the medium term.257 408.  Social integration is worse for camp IDPs than non-camp IDPs in Nigeria, and camp IDPs in four of the five countries feel less safe than hosts and non-camp IDPs. In Nigeria, camp-based IDPs report worse relations with hosts, while in Somalia, camp and non-camp IDPs have similar, largely positive relations with hosts. Apart from Sudan, camp-based IDPs and refugees in all the countries tend to feel either as safe or less safe from violence, walking in the daytime, and walking in the night-time as hosts and non-camp IDPs. In Nigeria, camp-based IDPs are much more likely to prefer a return than IDPs outside camps, who prefer to stay (Table C.2). Social cohesion outcomes might have influ- enced IDPs to locate themselves in or outside camps. Somali refugees in Kenya with low social and economic capital tend to reside in camps near the borders, while others with lineage connections and economic resources reside in areas of Nairobi.258 A low level of initial socioeconomic capital can be further eroded from being sequestered from local markets and culture. Building a sense of safety and facilitating increased social relationships with hosts would help to reach local integration. Inequality and Social Perceptions in Host Communities Introduction 409.  The influx of IDPs influences the host community through multiple economic and social channels, includ- ing job markets, service delivery, and social cohesion. The arrival of large numbers of IDPs or refugees represents a local demographic shock for host communities.259 The demand for essential services such as water, sanitation, and schooling increases, often accompanied by more urgent needs such as food provision. The labor market faces an influx 257. World Bank. 2017. “Forcibly Displaced: Toward a Development Approach Supporting Refugees, the Internally Displaced, and Their Hosts.” 258. World Bank. 2017. 259. Ibid. Volume B: Country Case Studies  | 201 of new labor supply, while the displaced often have different livelihood backgrounds and skills than local workers. The arrival of IDPs and refugees also “alters balances across ethnic or social groups within host communities,” which affects social cohesion both within host communities and between hosts and the forcibly displaced.260 410.  IDP presence can alter the socioeconomic fabric of the host community through positive and negative channels, which interact through the immediate term to the short and middle term. The influx of IDPs can affect certain parts of the host communities in specific ways. For instance, the poorer parts of the host community, or those closest to IDP camps, might find themselves sharing the same resources for water and sanitation, which can put a stress on them in the immediate term. Other parts of the host community, for instance, landowners who can rent out the land to IDPs seeking work as farm labor, might benefit from the IDP influx once the labor market adjusts to the IDP presence. Groups that gain from IDP presence might diverge in their personal trajectories from groups that lose out. Thus, the “inflow of forcibly displaced often transforms dynamics within host communities.”261 411.  Host communities that are more resilient may be better able to accept others.262 If a host community is more heterogeneous, its subgroups can be affected by IDP presence in directions that are more different. Inequality in income can result in a greater negative effect of IDP influx on lower income groups, for instance, in sharing limited amenities or participating in the same labor market. Individuals who are in the workforce may be affected differently than those who are not. Characteristics of household heads including sex and education can further affect the beliefs that household members hold about IDPs. If a community is itself receiving aid, it might not view IDPs as receiving spe- cial aid resources. Thus, divergence along several lines, from income to literacy, can affect how prone a host community might be toward being able to absorb IDP influx and presence. 412.  This section examines the effects of income inequality and other measures on host community percep- tions. Perceptions within the host community, as well as beliefs about IDPs and refugees, are studied. Income inequal- ity is measured as the standard deviation of consumption-based income at the area level. For each country, the area is defined as the highest administrative division below national, that is available in the data. In Sudan, the area variable has a different calibration due to the geographic organization of the host community in the sample. It indicates being inside or outside the city of Al Fashir, the city near which the camps are located. 413.  The samples for the analysis of perception indicators largely contain comparable sets of countries but with caveats. Most indicators, specified through columns 1–6 (Table C.3), contain households from Ethiopia, Nigeria, and Sudan. The indicator in column 3 includes Somalia in addition to these three countries, while the indicator in col- umn 4 only contains Ethiopia and Nigeria. The varying country composition of analysis samples is due to differences in questionnaire design in the countries. South Sudan is excluded from this analysis as the sample does not cover a host community, but the urban resident sample which does not overlap perfectly with the (prewar) states where surveyed IDP camps are located. While this caveat has been made elsewhere in the report, the analysis of this section focusses exclusively on host communities, thus South Sudan is excluded. 260. World Bank. 2017. 261. Ibid. 262. Ibid. 202  |  Informing Durable Solutions for Internal Displacement Within-Host Community Perceptions and Beliefs toward IDPs 414.  Host communities with higher inequality believe more strongly that the arrival of the displaced has worsened job prospects. Income inequality in the area does not have a significant impact on the relations that hosts have among themselves, overall life satisfaction, or even relations with the IDP or refugees. However, it results in a strong belief that the influx of forcibly displaced groups has worsened job prospects (Table C.3). Higher inequality thus seems to be manifested largely through this perceptions about the labor market effect. Lower-income host households in unequal host communities might be more severely affected by competition for jobs than lower-income host house- holds in less unequal communities. 415.  Higher overall prosperity of host communities results in better perceptions within the host community and of the displaced population. Host households in wealthier areas report better relations with neighbors within the host community, have higher levels of overall satisfaction with their lives, have better relations with IDPs or refugees, and are less likely to believe that the influx of the displaced has worsened job prospects. While wealthier host commu- nities believe more keenly that the displaced receive strong aid support, this perception does not seem to manifest as ‘resentment’ (Table C.3). 416.  Communities can be heterogenous on a range of socioeconomic characteristics apart from income. Such heterogeneity can result in IDP influx affecting certain subgroups more, and thus leading to differing perceptions of IDPs. Sex, literacy, participation in the labor market, and receiving aid support are channels through which households within a host community differ. 417.  Areas with higher proportions of woman-headed households report better social relations but worse perceptions of employment opportunities. Host communities with higher proportions of woman-headed house- holds have better relations both within the hosts and with the displaced population. However, they also have lower levels of satisfaction with employment opportunities and believe more strongly that the presence of the displaced has worsened job prospects. They also perceive that the displaced receive strong aid support. Households headed by women seem to be more able to build positive relationships with other households, while suffering from less favorable job prospects, possibly as the job markets might be more difficult for women to penetrate due to mobility, access, or competing demands such as childcare or household work. 418.  Higher literacy of household heads is associated with less favorable social relations and perceptions of employment prospects. Areas with larger proportions of literate household heads report worse relations within the host community and with displaced populations. Further, such communities report lower satisfaction with employ- ment opportunities and a stronger belief that the displaced have worsened job prospects (Table C.3). The dissatisfac- tion associated with household head literacy can indicate that job prospects are not sufficient in host areas, which are often poor and underdeveloped. The influx of IDPs could thus worsen this job market dynamic. If the displaced are roughly as literate or educationally accomplished as hosts, thus attempting to search for similar job opportunities, then their presence can exacerbate negative host perceptions.   TABLE C.3    Links between host community characteristics and host community perceptions 1 2 3 4 5 6 Do not believe that displaced Believe that Good relations Satisfied with have adversely displaced receive with neighbors employment affected job Good relations strong aid and community Satisfied with life opportunities prospects with displaced support Standard deviation of household income in area  −1.446 −12.945  1.405 −179.615***  −3.270  3.133 Average household income level in area   1.163***   1.161** −0.424    8.400***   1.237***  0.730*** Proportion of woman-headed households in area   3.369***  −1.154 −1.053***  −18.483***   3.143***  4.467*** Proportion of literate household heads in area −13.627***  −6.449 −0.845*  −85.680*** −14.361*** −3.211 Proportion of employed household heads in area  −2.617 −12.264* −0.093 −123.905***  −3.866*  1.196 Proportion of households receiving aid in area   2.880   4.946  0.229   58.826***   3.472*  1.694 Observations   4,197   4,214  4,673    3,200   4,199  4,216 Source: Authors’ calculations using SPS Ethiopia 2017, IDP Survey Nigeria 2018, Somali HFS 2017–18, HFS South Sudan 2017, and IDP Survey Sudan 2018. Note: *** p < 0.01, ** p < 0.05, * p < 0.1. All regressions include country fixed effects and area fixed effects. Volume B: Country Case Studies  | 203 204  |  Informing Durable Solutions for Internal Displacement 419.  Employment of household heads is associated with less favorable employment perceptions and atti- tudes toward the displaced. Areas with higher proportions of employed household heads report lower overall satis- faction with life, a stronger perception that the displaced have threatened job prospects, and worse relations with the displaced (Table C.3). Members of host communities who engage more regularly in the labor market might be more exposed to interactions with a potential new labor influx, which can lead to a greater sense of being economically threatened. 420.  Having aid-receiving host households in the area leads to more favorable perceptions of displaced groups. Host communities with higher proportions of aid-receiving households report better relations with the dis- placed and are less likely to believe that their arrival has threatened job prospects (Table C.3). Such households are thus more sympathetic to displaced groups and more able to accommodate them with positive perceptions and relations. Providing humanitarian and developmental support to host communities can reduce the sense of resource strain, and thus enhance relations with the displaced. 421.  The heterogeneity of host community households drives nuanced dynamics with IDPs and warrants a greater developmental investment in host community areas. Wealthier households likely do not compete with IDPs for jobs and could possibly have livelihoods that are complemented by a labor force influx, explaining their better relations with IDPs and their belief that IDPs do not worsen job prospects for locals. However, poorer households, espe- cially in unequal areas, and those who are engaging in the labor market, may be more exposed to the IDP presence. This could be through interaction in a job landscape, or due to sharing the same resources such as health, education, and water facilities. Through both channels, it is possible that poorer households perceive that IDPs will have a greater impact on their socioeconomic outcomes. More literate host community members might already be dissatisfied with the current job opportunities, which can explain their less favorable perceptions of IDP influx on jobs. However, by pro- viding aid to the host community areas and developing them further as a developmental objective in its own right, the integration of IDPs can be bolstered. Households receiving aid have more accommodative perceptions of IDPs, which supports the provision of aid to host communities as well as the displaced. Geographic Dispersion and Duration of Displacement Introduction 422.  IDPs who have been displaced for longer periods of time can have specific needs and opportunities. IDP groups who have been displaced for a long period of time are in a different stages in the displacement trajectory than groups who are recently displaced. IDPs who have been displaced for many years might have faced a prolonged ero- sion of assets or human capital. Alternatively, they may have had more time to adapt to their current locations, integrate socially, and build relevant skills for local labor markets. Thus, their return intentions might also differ from IDPs who are recently displaced. 423.  Geographical proximity to the original residence can influence IDP outcomes. IDPs who are close to their original residence may face different outcomes and challenges than those displaced to farther areas. Going farther can result in a more different ethnic or cultural environment in the new location, influencing social integration as well as economic interaction. Alternatively, IDPs who go far could have been motivated by strong reasons such as family Volume B: Country Case Studies  | 205 connections, or have enough facilities to travel further. They could also have been displaced further over repeated waves of displacement, which can imply a serious deterioration of physical and human capital. 424.  This section examines the effects of distance and time on IDP outcomes. As the two factors interact over the course of displacement, they are analyzed together. A range of welfare outcomes from poverty and living standards to employment and social perceptions are compared over geographical and time dimensions through regressions. The broad causes of displacement, social perceptions, and return intentions are also analyzed through the time-geography lens. IDPs and refugees from all five countries are included in the analysis sample. Geographic Dispersion 425.  IDPs with agricultural backgrounds avoid traveling far, often being displaced close to the original res- idence. IDPs displaced within the same ward/boma are more agricultural than IDPs in a different ward/boma in the same district.263 As IDPs move from being in the same district to a different district to a different state, they have, increasing nonagricultural backgrounds. Refugees, who are in a different country, were least agricultural at the origin (Table C.4). The geographic trend could be due to a desire to stay close to agricultural land in the hope of an eventual retrieval. It could also indicate less resources for travel, thus precluding longer routes and prolonged mobility. 426.  IDPs who live further from the origin have a higher preference for return. The desire to return to the origin is stronger for IDPs who have moved further away. The preference for return increases progressively as IDPs move from being in the same ward/boma through being in a different state in the country though it drops strongly for refugees who prefer to stay in the country of refuge or seek a new country. Despite the desire to return, IDPs displaced further have better social relations with hosts. IDPs in a different ward/boma or a different district report better social relations with hosts than those in the same ward/boma (Table C.4). Since IDPs who are displaced farther are largely nonagricul- tural, they may have similar livelihood backgrounds to hosts, who are largely in urban areas. However, the challenges of being geographically distant from the residence appear to outweigh potential benefits. Having traveled long routes could lead to depletion of physical and human capital, and a loss of connection with social networks. Further, a greater geographic distance could indicate that labor markets and cultural norms are more difficult to integrate into. Displacement Duration and IDP Outcomes 427.  Longer-displaced IDPs report better social relations with hosts, but this does not translate to higher stan- dards of living. Longer-displaced IDPs are more likely to perceive social relations with host communities as good or very good. Having more time to adapt and integrate thus leads to better social cohesion and could also give hosts more space to accommodate IDPs. However, better social cohesion is not linked to higher economic welfare. The duration of displacement does not affect living standards on measures such as poverty, school enrollment, or even housing, which might be expected to improve over time as IDPs move from immediate-term shelters into short- or medium-term dwellings (Table C.4). 263. The analysis draws on administrative divisions to measure dispersion. Different countries use different terms for subnational geographic divisions. In this section, ‘district’ refers to ‘district/LGA/county’ and ‘state’ refers to ‘state/region’. 206   TABLE C.4    Effect of displacement duration and geographic distance from origin on IDP outcomes 1 2 3 4 5 6 7 8 9 10 11 In labor Good Enrolled force Agricultural relations Duration of and of and of   Improved Conflict Climate livelihood Prefers to Prefers to with displacement school working   Poor housing displaced displaced at origin stay return hosts in years age age Duration of displacement in years −0.002  0.014 −0.012 −0.024  0.031** −0.017  0.009  0.032**    0.003 −0.010* Same district/LGA/county  0.003 −0.348 −0.102  0.031 −0.248* −0.254*  0.471***  0.401*** −0.140  0.088 −0.030 Same state/region −0.016  0.220  0.003 −0.215 −0.279***  0.085  0.163*  0.196*  0.383**  0.116  0.036 Different state/region −0.069* −0.346 −0.190 −0.635 −0.563***  0.001  0.263* −0.241  1.346***  0.123  0.039 Different country −0.035  1.218** −0.218 −4.471*** −1.143***  0.624* −0.722**  0.106  2.100* −0.401  0.071 Household head is literate −0.033  0.507*** −0.078 −0.674*** −0.424***  0.021 −0.145*  0.163*  0.009  0.125* −0.031 Household size  0.054*** −0.023 −0.038**  0.142***  0.043*** −0.012  0.019  0.008  0.012  0.016 −0.045*** Household head is a woman  0.080*** −0.200**  0.308*** −0.255  0.027  0.167**  0.001 −0.157 −0.472***  0.222***  0.025 Household head is employed −0.021 −0.137 −0.176*  0.476*** −0.057  0.056  0.082  0.161*  0.535***  0.020  1.351*** Observations  7,972  9,214  9,086  9,086  9,142  8,835  8,835  9,208  9,216 20,066 22,241 Source: Authors’ calculations using SPS Ethiopia 2017, IDP Survey Nigeria 2018, Somali HFS 2017–18, HFS South Sudan 2017, and IDP Survey Sudan 2018. Note: *** p < 0.01, ** p < 0.05, * p < 0.1. All regressions include country fixed effects. Volume B: Country Case Studies  | 207 428.  IDPs who have been displaced for longer tend to have agricultural backgrounds and low labor force par- ticipation. IDPs and refugees who have been displaced for longer are more likely to have derived their livelihoods from agriculture at their origin. Further, they report lower labor force participation in the current location (Table C.4). This could be due to a lower degree of skill transferability as agricultural IDPs. Long spells of inactivity can deteriorate skills and morale making it more difficult to reenter the labor force. Thus, policy responses for long-displaced agricultural IDPs should focus on employment generation and skill-building as key components. 429.  IDPs who have been displaced for longer have traveled farther from the origin. IDPs located in a different district than the origin are displaced for longer than those displaced within the same ward/boma or the same district. IDPs in a different state are even further dispersed, while refugees are displaced for longer than all the other groups (Table C.4). IDPs who were displaced farther might have faced a more widespread cause of displacement, such as drought that affected a large area or fast-spreading violence—making it more difficult to return and thus prolonging the state of displacement. IDPs who are located further from the residence might also have lost contact with networks at the origin, or find it more costly and less safe to return along long travel routes. Conclusion Summary of Findings from Country Case Studies Nigeria 430.  Nearly 2 million people are currently displaced in northeast Nigeria, most of them due to armed conflict. The Boko Haram conflict has led to a steep number of deaths and displacements. More than 53,000 deaths from politi- cal violence have been reported in Nigeria between 2011 and the second quarter of 2018. More than 60 percent of the deaths have been in the northeast states of Borno, Adamawa, and Yobe, with Borno state alone accounting for more than 25,000 deaths during the period. As of August 2018, there are 1.9 million IDPs in Nigeria, which represents a 2 per- cent rise from April 2018, and 5 percent rise from June 2017. Most of these IDPs are in Borno state. Almost two in three IDP households cite armed conflict as the reason for their displacement, and security as a primary need to settle anew. 431.  The Nigeria IDP survey represents IDPs and host community households across northeast Nigeria. The Nigeria IDP Survey 2018 was conducted in six states of northeast Nigeria (Adamawa, Bauchi, Borno, Gombe, Taraba, and Yobe) where most of the IDPs are currently living. The survey collected micro-data on consumption, health, education, living conditions, and displacement-related questions. Within each state, an equal number of IDP and host community households were interviewed, where IDPs consisted of either those in camps or in host community settings. IDPs in camps make up 40 percent of the IDPs, while IDPs in host communities make up 60 percent. Borno represents 80 per- cent of the total sample, as most of the IDPs are currently living there.264 As the survey was conducted over a span of several weeks, it can capture certain seasonal effects, for instance on employment and food security, which might vary at other times of the year. The results that follow should be interpreted keeping this in mind. 432.  IDPs are largely children and live either in host communities or in camps. About 61 percent of all IDPs are living in host communities, while 39 percent are in camps. Borno state, which holds 75 percent of Nigeria’s IDPs, is the only state where the percentage of IDPs residing in host communities is the same as that of IDPs residing in displace- ment sites. Almost three in five IDPs is a child under 15 years old. 433.  Almost all IDPs are poor, food insecure, and doing badly on a range of basic living outcomes. Almost 9 in 10 IDPs in northeast Nigeria are living under the international poverty line of US$1.90 PPP (2011) per day per person, on average consuming less than one-third of the poverty line threshold. Another 7 in 10 IDP households are highly food insecure. IDPs suffer from overcrowding in their dwellings and toilets, are less likely than hosts to use a hospital or clinic for childbirth, and have low enrollment in school. Many IDP children have not been in school for three or more years, possibly since their displacement. A significant proportion of IDPs lost houses that they had owned for many years and now live in worse housing than before. They also lost agricultural land, most of which was owned before displacement. 264. As the majority of the survey was conducted in Borno (2,200 out of 2,800 interviews), the analysis does not compare outcomes across different states due to a low number of observations from the other states. 208 Volume B: Country Case Studies  | 209 434.  Though slightly better off than IDPs, host communities face widespread poverty and severe living con- ditions. Host community households are more likely to own homes and agricultural land than IDPs. They also have higher primary and secondary enrollment rates, are more likely to use a doctor or clinic for childbirth, and are less likely to have crowded toilets. Access to basic amenities such as water, sanitation, school, and markets is similar for IDPs and hosts. However, despite doing better than IDPs on certain welfare measures, host communities also face significant challenges. Almost 8 in 10 host community households are living in poverty, consuming less than 40 percent of the poverty threshold of US$1.90. Almost half of the households are highly food insecure, while one in five working-age members are inactive. The severe living conditions for host communities, and the urgency to accommodate the often disadvantaged and vulnerable IDPs, necessitates investment in host community areas as a whole. 435.  IDPs in camps generally have less favorable outcomes than IDPs living among the host community and are more keen to return home. While IDPs in camps and in host communities have similar levels of poverty, the latter are slightly better off on a number of dimensions. Significantly more IDPs in camps face overcrowding in rooms and toilets than IDPs in host communities. Since most camp-based IDPs occupy tents or temporary houses, the effects of overcrowding are more pronounced for them. Additionally, IDPs in camps are less likely to visit clinics or hospitals for childbirth but would instead give birth at home or use the services of traditional birth attendants. They also have lower education enrollment rates. The worse living conditions in camps are reflected in return intentions. Most camp-based IDPs plan to return home, while most IDPs living among host communities intend to stay. 436.  IDP women face worse educational and labor market outcomes, and often head households with higher dependency ratios. Households headed by women are smaller, but with higher dependency ratios. Almost 60 per- cent of households send women or girls to collect water, and they often endure long queuing times. Poor maternity practices have been adopted as most IDP households had their child delivered at home in the absence of a doctor, nurse, or midwife. Enrollment rates are lower for girls than boys, reflecting in turn the lower educational attainment of women than men. Finally, women are more likely than men to be inactive in the workforce. 437.  Two distinct profiles of IDPs are identified in drawing the typologies for Nigeria. Group 1 represents 74 per- cent of the IDP population and Group 2 around 26 percent. The Boko Haram conflict has led to a steep number of deaths and displacements. The place of origin is similar for both groups of IDPs. Even though most households were displaced by armed conflict, Group 1 is slightly more likely to cite this reason compared to Group 2. Before displacement both groups had similar living conditions, yet Group 2 was more inclined to an agricultural livelihood and Group 1 was more likely to rely on wages, salaries, and their own businesses. Currently, households in Group 1 have less members, a higher dependency ratio, and are more likely to be headed by a woman who is not working. IDPs in Group 2 are more likely to have an agricultural livelihood and to receive assistance, although both groups are equally poor and food insecure. The differences in housing conditions and access to services between groups are determined by their current location, as Group 1 is more likely to live in host communities, whereas Group 2 in settlements or camps. IDPs in Group 2 were more satisfied before displacement, are more dissatisfied today, less likely to feel safe, and are more pessimistic about the future. Households in Group 1 prefer to stay in their current location motivated by security reasons, while IDPs in Group 2 intend to return to their place of origin guided by access to land, services, and employment. 438.  A durable solution for Nigeria’s IDPs will require security for all and a consideration of the needs of dif- ferent groups. IDPs are being provided water and sanitation facilities, are close to many basic amenities, and have 210  |  Informing Durable Solutions for Internal Displacement good relations with their host communities. However, there is a dire need to address the alarmingly high poverty, food insecurity, and poor housing, education, and health conditions. Lack of security is a key concern for IDPs, most of whom were driven from their origins by armed conflict. An option of return remains infeasible, while many of the areas of IDPs’ origin remain conflict stricken. While concerns such as lack of security, low school enrollment, and overcrowding are common to all IDPs, certain groups of IDPs have additional vulnerabilities or disadvantages. Poorer educational rates and labor markets for women, and the lower living standards in camps have an important bearing on the prospect of a durable solution. Somalia 439.  About one in five Somalis are forcibly displaced. From a population of 14 million, Somalia has nearly 2 million internally displaced people. Additionally, over 877,000 Somali refugees live in neighboring countries, making them one of the largest refugee populations in the world. Four consecutive poor rainy seasons, along with ongoing clan-based conflict and violence from armed non-state actors, have led to surging displacements from late 2016 to late 2017. Forc- ibly displaced populations of Somali regions are thus a complex mix of IDPs, returnees, and refugees seeking asylum in other countries. 440.  The survey represents IDPs across Somalia and host communities living in their proximity, as well as the overall national urban and rural populations. In the SHFS 2017–18, IDP populations, host communities, and national resident communities were surveyed and are nationally representative. Boundaries of IDP settlements across the dif- ferent pre-war regions of Somalia were obtained using UNHCR’s Shelter Cluster and United Nations Population Fund’s (UNFPA) Population Estimation Survey 2014 (PESS). To account for the major displacement and migration induced by drought, each selected IDP enumeration area (EAs) was inspected to ensure that it was still inhabited by displaced com- munities. For the host communities sample, all urban EAs adjacent to IDP settlements were pre-selected as a separate sampling frame. As the survey was conducted in about two months, several indicators have likely captured seasonal effects, for instance on employment and food security, which might vary at other times of the year. The results that follow should be interpreted with this caveat. 441.  IDPs, like the rest of the Somali population, are overwhelmingly young. More than 50 percent of IDPs are under 15 years old, and less than 1 percent are above 64 years old. The large proportion of children drives high depen- dency ratios—IDP households have dependency ratios larger than 1, indicating that for each working-age member there is a child who must be provided for. 442.  IDPs face greater poverty and worse living conditions than residents. While almost 70 percent of Somali residents are poor, IDPs form an especially marginalized group, with over 75 percent of IDPs living on less than US$1.90 PPP (2011) per day per person and more than 50 percent of IDP households facing hunger. Large numbers of IDPs must share essential amenities such as toilets, crowding out the improved WASH facilities that settlements offer. IDPs in set- tlements are also farther away than host communities from essential facilities such as primary schools, health centers, and markets. 443.  Along with worse living conditions, IDPs also have lower human capital, which can lead to lifelong wel- fare gaps. IDP children are less likely to attend school than urban residents. Also, IDP adults are less likely to be able to read and write than urban residents. The educational outcomes of the IDP population are closer to rural outcomes and Volume B: Country Case Studies  | 211 lag urban ones. However, most IDP households (three in four) are in urban areas. These gaps in educational attainment are particularly crucial since half the Somali population is under 15 years old. Lags in educational attainment for IDPs can translate to persistent, lifelong gaps not only in education but also in employment and overall well-being, as the young population matures. 444.  IDPs are less able to rely on agriculture as before being displaced, so they need to adjust to new live- lihoods and skills in their new settings. The pre-displacement livelihoods of IDPs differ significantly from urban livelihoods today. IDP livelihoods before displacement consisted of a mix of salaries, small businesses, and agriculture, while urban livelihoods today consist largely of salaries, followed by remittances. Fewer IDPs are making a living from agriculture, and many IDPs are employed in helping with businesses, indicating an adjustment into the employment landscape of their new locations. IDPs today rely on a mix of salaries, small family businesses, and aid for household income. Livelihood enhancing programs thus need to account for likely skills gaps. 445.  IDPs receive relatively low remittances, indicating a lack of safety nets. Only 7 percent of IDP households rely on remittances as their primary source of livelihood. The average IDP household receives half the remittances of the average urban household. IDP households were as likely to rely on remittances after displacement as before, indicating that remittances do not serve as a safety net for displacement. A lack of sustainable safety nets and networks, under- standable for IDPs who have lost homes and faced varying degrees of separation, makes government-provided social protection especially relevant—particularly in helping households deal with shocks. 446.  Sustainable employment opportunities for IDPs are especially important, given the changing livelihood structures and a lack of safety nets. Employment and labor force participation is low, especially among IDP men as compared to urban resident men. Expanding salaried labor opportunities through public work schemes or infrastruc- ture building activities can help accommodate the influx of labor force in urban areas. Approaches that combine cog- nitive and non-cognitive skill building in addition to vocational training or cash transfers can help address psychosocial challenges such as trauma and depression which may impede participation in employment activities. Moreover, gen- der-responsive approaches can address barriers to employment opportunities, such as disadvantageous social norms and the domestic labor burden on women. Interventions to enable women to engage in economic activities should consider protection and prevention of GBV. 447.  Among the displaced population, certain groups are worse off than others. IDPs displaced by climate events are poorer and have worse housing quality than those displaced by conflict. IDPs who are in protracted displacement— mostly in urban areas—have better access to health care. IDP households headed by a woman get only one-sixth the amount of remittances that IDP households headed by a man get. Richer IDPs are more confident than poorer IDPs of being relocated within the next year. Understanding the specific challenges of different types of IDP households can help inform a strategy to reach a durable solution for all. 448.  Most IDPs prefer to stay in their current locations, regardless of why they were displaced. More than 7 in 10 IDP households want to remain in their current location, and 9 in 10 have not visited their original residence since they were displaced. Intentions to stay are likely motivated by security—a majority of IDP households cited security as the reason for choosing their current location, and 8 in 10 report feeling safe or very safe where they currently are. IDPs 212  |  Informing Durable Solutions for Internal Displacement also perceive positive social relations with host communities, with 9 in 10 IDP households agreeing that they have good dealings with their surrounding communities. 449.  Two typology profiles can be distinguished among Somali IDPs. The groups differ in their displacement trajectories, particularly from the cause-based and needs-based lens. Group 1, which accounts for about 40 percent of IDPs, had more agricultural livelihoods at the origin, was more likely to be drought displaced, and had poorer living conditions before displacement and currently. Group 2, while less agricultural and with better housing at the origin, was more likely to be displaced by armed conflict. The households of Group 2 are less poor, less hungry, and in better housing than those in Group 1. The differences among the groups, while still present, are smaller in the solution-based lens. Most members of both groups prefer to stay in the current location rather than return or relocate, though Group 2 is more likely to be guided by security and Group 1 by basic amenities and livelihoods in addition to security. 450.  Successful local integration for IDPs would require substantial investment in strained urban centers, which can currently only offer sub-par living conditions to the displaced. The concentration of IDP settlements in urban cities and towns, in addition to ongoing rapid urbanization, places significant stress on the urban infrastructure of Somalia. Strengthening the viability and resilience of Somalia’s urban and peri-urban areas is critical. This will entail investing in services such as water, housing, sanitation, and education, and infrastructure such as roads and telecom- munications, to help cities absorb massive population influxes. This is also essential to prevent a decline in service and livelihood access of existing host communities, and thus preserve good IDP-host community relations. 451.  Support to rural resilience and recovery is also important for providing IDPs with the option of a safe and voluntary return. The socioeconomic and human capital outcomes of IDPs are often comparable or slightly better than those of rural residents, highlighting the development deficits confronting rural populations in Somali regions. Improving rural access to services and livelihoods is thus an important component of helping IDPs who wish to return or relocate to rural areas. Options to be considered include start-up assistance and support to help restore rural livelihoods, as well as investing in inputs for agricultural production and restocking livestock. Other investments may include cash transfers for basic consumption, skills development to bolster human capital development, and land restoration and housing repair to recover losses from displacement. South Sudan 452.  Conflict in South Sudan sparked off in December 2013 and displaced 4 million people. The conflict began as a power struggle within the ruling party, the SPLM. President Salva Kiir, an ethnic Dinka, fired Vice President Riek Machar, an ethnic Nuer, suspecting a plot to overthrow the government. Fighting broke out in Juba between forces loyal to the two men, igniting a civil war that quickly spread across the country. Areas initially affected by violence were Greater Upper Nile as well as some areas of Greater Bahr el Ghazal. Eventually the war also trickled into Greater Equa- toria, spreading local violence. About 2.1 million became refugees in neighboring countries, including Uganda, Sudan, Ethiopia, Kenya, and DRC. Another 1.9 million IDPs fled within South Sudan and are concentrated in the Greater Upper Nile Region. 453.  The survey represents four of the largest IDP camps, and draws comparisons with urban populations of the country. The CRS 2017 represents IDPs in the four largest PoC camps with defined boundaries. These camps are Bentiu PoC, Bor PoC, Juba PoC, and Wau PoC. A majority of South Sudan’s IDPs are believed to be outside the camps. Volume B: Country Case Studies  | 213 However, the detailed micro-data on the PoCs fills important information and knowledge gaps for IDP-focused pro- gramming. The survey was designed to be precisely comparable to data representing urban South Sudanese popula- tions, from the High Frequency Survey South Sudan (HFS) 2017. This survey studied urban areas of 7 of the 10 pre-war states of South Sudan. As the PoC camps covered in the CRS are in urban areas, HFS 2017 allows for comparisons in the outcomes of IDPs and the residents of the areas where they are now located. However, HFS 2017 does not cover two of the pre-war states in which CRS camps are located (Jonglei and Unity). Thus, comparisons are drawn at the overall urban and IDP levels rather than for specific camps or pre-war states. Further, as both the CRS and HFS were conducted within a few months, several indicators have likely captured seasonal effects, for instance on employment and food security, which might vary at other times of the year. The results that follow should be interpreted with this caveat. 454.  IDPs in South Sudan are younger than urban residents, driving high dependency ratios. About 45 percent of IDPs are under 15 years old, as opposed to 32 percent of urban residents. Consequently, fewer IDPs are of the working age, which is between 15 and 64 years old. This translates to higher dependency ratios among the displaced popula- tions. In IDP households, each working-age member must look after one non-working–age dependent. 455.  IDPs have faced a drastic deterioration in living standards and live in worse conditions than urban resi- dents today. In a country with staggering poverty prevalence, IDPs are a particularly marginalized group. More than 9 in 10 IDP households are living in poverty. Before the December 2013 conflict, more than 4 in 10 IDP households had improved housing, and 9 in 10 owned their dwelling. Today, almost all IDPs and refugees live in overcrowded tents or temporary shelters provided by camps. Severe overcrowding in dwellings and toilets reduces sufficient access to hygienic facilities, flares the spread of communicable diseases, and increases the threat of GBV and harassment. 456.  Along with living conditions, the displaced have also lost income-generating assets, and now rely over- whelmingly on aid. Agricultural land access for the average IDP household has gone from 0.8 acres pre-December 2013 conflict to about 0.2 acres today. IDP households also lost almost all their livestock holdings over the course of displacement, from 42 livestock units pre-conflict to 2 units today. These losses further limit the ability of displaced populations to create employment opportunities for themselves, hampering self-reliance. More than three in four IDP households now rely on humanitarian aid as their chief source of livelihood. 457.  Security is the predominant factor driving future settlement intentions. About 6 in 10 IDP households want to stay in their current location, while 1 in 3 want to return to their origin. Households that want to stay in their cur- rent location are motivated by better security, services, and assistance in the camps. Households that want to resettle outside the camps are also motivated by better security and access to health and education. In addition, about 4 in 10 households that want to relocate also report access to home, land, livestock, or employment as a motivating factor. 458.  IDPs in South Sudan have two distinct typologies. The two groups, Group 1 and Group 2, are of a roughly equal size and represent 40 percent and 60 percent of the IDPs, respectively. Before displacement, Group 2 had more agricultural livelihoods, worse housing, and was more likely to be displaced by armed conflict. Contrary to this, Group 1 had wage- and business-based livelihoods. A majority of Group 1 was also driven by armed conflict, but to a lesser degree than Group 2. Differences in the current conditions of the two groups are possibly driven from their different pre-displacement situations. Group 2 has larger households, higher dependency ratios, higher poverty and aid depen- dence, and feels less safe. Group 2 is more likely to be confident of a moving timeline, but also more likely to seek 214  |  Informing Durable Solutions for Internal Displacement information to decide on a move, while Group 1 is more optimistic about the future. Group 1 is largely in Juba PoC and Bor PoC, and Group 2 is concentrated in Bentiu and Bor PoCs. 459.  Security, services, and humanitarian assistance influence durable solutions. While security is the main rea- son that IDPs left their places of origin and settled in the camps, access to services and humanitarian assistance are also important. Should security conditions improve, the provision of services and humanitarian assistance in places of origin could encourage IDPs to return. About one in three IDPs want to return home now. The main reasons they want to move include access to security, services, assets, and employment, as well as livelihood opportunities. Where security conditions have already improved, the provision of services and humanitarian assistance in the immediate run, combined with property and asset restoration mechanisms in the medium run, can help IDPs make sustainable returns. 460.  Looking ahead, policies and programs need to shift from encampment to sustainable development for IDPs, returnees, and their hosts. For IDPs who want to stay, in the short term, the focus needs to be on (a) maintaining and building their human capital including food security, health, and education; and (b) improving their living condi- tions such as housing and sanitation. In the medium term, the focus needs to gradually shift to (a) providing opportu- nities for the socioeconomic integration of IDPs outside camps, for example, social cohesion, skills development, jobs, and access to land and livestock; (b) continuing to build their human capital including nutrition, health, and education; (c) reducing humanitarian assistance where possible; and (d) supporting host communities to absorb IDPs and improve their own living conditions, for example, services and economic opportunities. For IDPs who want to return, the focus needs to be on (a) providing opportunities for the socioeconomic reintegration of returnees (social cohesion, skills development, jobs, and property and asset restoration); (b) continuing to build their human capital such as in nutrition, health, and education; (c) providing humanitarian assistance where necessary; and (d) supporting host communities to absorb the returnees and improve their own living conditions, for example, services and economic opportunities. Sudan 461.  Most IDPs in Sudan’s surveyed camps were displaced at the height of the Darfur conflict in 2003–04. In camps around Al Fashir, IDPs were driven at the peak of the conflict, and today, around half of them want to stay. About 80 percent of camp residents were displaced shortly after the beginning of the conflict and have since lived there. Only 30 percent have returned to their previous places of residence. Instead, most working-age IDPs have income- generating activities in or around the camps. Half of the IDPs do no longer entertain plans to return to their place of origin or to move elsewhere. Security is a key consideration, yet IDPs also appreciate access to health and education services that camps offer. 462.  The IDP profiling survey covered two camps near the city of Al Fashir, and the host community residing in the city. In the Sudan IDP Profiling Survey 2018, IDP households from the two camps of Abu Shouk and El Salam as well as households from the immediately neighboring city of Al Fashir were interviewed. Since both camps are in the immediate neighborhood of the city, this setup allows for good comparability between IDPs and the host popu- lation. At the same time this setup limits the representativeness of the data to IDPs in camps around Al Fashir. As the survey was conducted over the course of a few weeks, several indicators likely captured seasonal effects, for instance on employment and food security, which might vary at other times of the year. The results that follow should be inter- preted with this caveat. Volume B: Country Case Studies  | 215 463.  IDPs and hosts are staggeringly poor, requiring urgent employment and livelihood opportunities. About 80 percent of IDPs live under the international poverty line of US$1.90 PPP (2011) per day per person. A recent spike in food prices following the removal of subsidies, as well as seasonal shortages, may have played a role and would sug- gest that these extreme levels of poverty may be temporary. But high levels of severe food insecurity (64 percent) and the general lack of access to productive assets further underline the severity of the situation. Lack of employment and livelihood opportunities in the camps are also the most important reasons IDPs cite when they consider relocating. 464.  Most IDPs live in permanent structures provided by the humanitarian community, but do not depend on aid for food or income. Virtually all IDPs live in traditional structures of mud, clay, and wood that are largely provided by NGOs or the UN. By contrast, most IDPs are self-dependent for their income; only 6 percent of households live mainly off aid and only 20 percent receive it at all. This independence is generally positive, yet it is also difficult to access aid for IDPs. Given the high poverty rates, the opportunities for income generation are not sufficient, and more aid may be needed in the short term. 465.  While IDPs enjoy access to many services they value, food availability and access to electricity are often lacking. IDPs have adequate access to housing. Overcrowding is rarely an issue. They also often have access to improved sources of drinking water and improved sanitation, as well as health centers, schools, and markets. However, 60 percent of IDPs exhibit high levels of food insecurity and only 9 percent of IDP households have electricity in their homes. 466.  Both IDPs and hosts describe their relations as good, and they generally feel safe. About 9 in 10 hosts and 4 in 5 IPDs describe the relations between their two groups as good. Most of the IDPs are from within the district, so there are likely only minor cultural differences. Despite a considerable risk of robberies, the camps are perceived as rel- atively safe. These are important preconditions for achieving durable solutions. 467.  Neither IDPs nor members of the host community exhibit high levels of civic engagement. About 9 in 10 households have never attended a public meeting in which community affairs were discussed, and about equally as many have never interacted with a community leader. This is true for both IDPs and hosts. It is therefore difficult to assess whether a higher rate of participation is something IDPs aspire to or to what extent it is important for achieving durable solutions for them. 468.  IDPs in Sudan have two different typology profiles. Before displacement, the two groups had different sources of income, and households from Group 1 were more likely to be displaced in 2003 and 2004. Households from Group 1 are also more likely to live in shelters provided in the camp, and thus are closer to services and more likely to have improved water sources. There are also differences in the current conditions of the two groups. Group 1 has a higher and deeper poverty incidence, is more likely to face food insecurity, and to rely on assistance from development partners or NGOs. As for the future, most households in Group 2 prefer to stay in the camp guided by security reasons, while a majority of IDPs in Group 1 want to relocate based on employment conditions and other considerations. The timeline for moving is clearer for IDPs in Group 2, and they are more likely to have all the information they need to inform their decision. 469.  Improving security and economic opportunities in camps and return areas, as well as an upgrading of camp infrastructure, would be key elements in bringing about durable solutions. Most of the IDPs living in Abu 216  |  Informing Durable Solutions for Internal Displacement Shouk and El Salam have no intention today of leaving the camps. If they are to stay, investment in infrastructure, especially access to grid electricity as well as a broadening of economic opportunities, will be keys in bringing about a durable solution to their displacement. Potential returnees often cite security concerns outside of the camps that would need to be addressed before a durable solution based on return becomes viable. As they were typically engaged in agriculture before their displacement, conditions for them to re-engage successfully in the sector should be put in place. Ethiopia 470.  A dramatic increase of refugees in Ethiopia has led to a shift in the country’s approach to displacement. Almost 1 million refugees settled in Ethiopia over the last decade, with nearly a half million South Sudanese refugees coming in over the last five years alone. Ethiopia faces compounding humanitarian crises and protracted displacement that put a strain on domestic actors’ coping capacities and question the sustainability of the current reception system. As a result, the GoE unveiled a strategic plan (‘nine pledges’) to address the development aspects of the refugee situ- ations in the medium to long term. Currently under implementation, the multi-sectoral initiative aims to improve the material condition of refugees, expand their rights, transition them from the country’s camp-reliant approach to a more progressive refugee policy, and support synergies and local integration mechanisms with host communities. 471.  In Ethiopia, the survey represents refugees living in camps and their host populations. In the SPS 2017, household surveys were conducted with refugees from South Sudan, Somalia, Eritrea, and Sudan living in camps in Ethiopia, and with host community members within a 5 km radius of a camp. The survey is only representative of ref- ugees living in camps—who, nonetheless, are the majority. The sample frame was the list of all refugee camps in the four main regions that host refugees: Tigray and Afar (hosting mostly Eritrean), Gambella (South Sudanese), Benishangul Gumuz (Sudanese), and Somali (Somali). As the survey was conducted over the course of a few weeks, several indicators may have captured seasonal effects, for instance on employment and consumption, which might vary at other times of the year. The results that follow should be interpreted with this caveat. 472.  Refugees from Eritrea, Somalia, South Sudan, and Sudan experience different displacement dynamics while in Ethiopia. Except Eritreans, all groups reported that armed violence and a general sense of insecurity were the chief drivers of displacement, and that improved security was the leading reason for settling in the current location. Specific drivers also include environmental degradation leading to a strain in resources for Somalis, and political per- secution, which is the leading driver for Eritreans. The timeline of displacement is quite different from group to group. While South Sudanese are the newest group, the other three groups had repeated patterns of displacement in the last three decades. Geographically, refugees are compartmentalized according to nationality. Each group settled in the border area in proximity of the respective country of origin. 473.  Poverty levels vary across refugees’ nationalities, but refugees are much poorer than host communi- ties. Every two in three refugees live below the international poverty line of US$1.90 PPP (2011) per day per person compared with around one in four host community members. Poverty rates among refugees range from 38 percent (Eritreans) to 72 percent (Sudanese). Eritreans are the ones that enjoy more rights compared to others, and, as a result, display higher standards of living and much lower poverty rates. In turn, less than 30 percent Eritreans experience high food insecurity, while the other three groups have rates ranging between 60 percent and over 80 percent. Similarly, Volume B: Country Case Studies  | 217 trends in housing conditions and overcrowding highlight Eritreans’ better situation with respect to other groups. South Sudanese are the poorest group by many indicators. 474.  Being in camps, refugees have comparable or better access to services than the host community. This sit- uation underlines the relative high quality of camp management compared to domestic service delivery mechanisms. Both refugees and host community members have nearly complete access to improved water sources. Refugees have far better access to improved sanitation facilities (69 percent) compared to host community members (37 percent) and have net primary enrollment rates for refugee children (79 percent) comparable to host communities. All refugee groups also have easier access to basic services compared to their pre-displacement situation. 475.  Refugees have very low labor force participation, depending massively on aid for livelihoods. Only 22 percent of working-age refugees (15–64) are currently employed, compared to 66 percent of the host community working-age population who are employed. Indicators on livelihood and access to services highlight the nearly com- plete dependence of refugees on aid. Over 90 percent of Eritrean, South Sudanese, and Sudanese households rely on aid as their greatest source of livelihood, while Somalis are the least aid dependent and have the lowest rate of food consumption derived from aid. Among the four refugee groups, Somalis are the most integrated group with host com- munities: Somalis display higher rates of labor force participation, slightly higher rates of home ownership, stronger feelings of safety and security, and better host populations’ perceptions. 476.  Future intentions of refugees mirror causes of displacement. As the main driver of displacement is personal security, all refugee groups regard security as the utmost priority when making considerations about future intentions and durable solutions. Overall, one in two refugees prefer to stay in the current location, although there is great varia- tion across groups. South Sudanese refugees, who are the poorest group among the four, have the highest percentage of respondents who do not want to move; Eritreans and Somalis have large percentages of refugees who want to move to a new country, while over one in two Sudanese want to return to Sudan. 477.  Typology analysis for Ethiopia identifies two groups of refugees with different profiles. The displacement situation of each group is different according to the conflict context in each country of origin. Group 1 is mainly com- posed of households displaced by armed conflict and coming from South Sudan, Somalia, and Sudan, while refugees from Group 2 come mostly from Eritrea and Somalia and were displaced for various reasons. Before displacement, Group 1 was more inclined to an agricultural or agro-pastoralist livelihood and Group 2 to depend on aid and remit- tances. Currently, Group 1 have larger households, are less likely to have Ethiopian relatives, and are usually headed by women without education. Group 2 are more likely to have better access to services and to rely on aid and remittances, yet Group 1 is poorer, poverty deeper, and more likely to face food insecurity. As for the future, most households in Group 1 prefer to stay in the camp for security reasons, while most refugees in Group 2 intend to move to another country guided by access to land, services, and employment. 478.  Refugee conditions underline the need for a shift from a humanitarian approach to a more sustainable medium-term development perspective. Specifically, policy resources should be devoted to (a) gradually provide out-of-camp opportunities for the socioeconomic integration of refugees (for example, freedom of movement, work permits, skills development, jobs, and integration into national systems); (b) improve living standards for host com- munities in peripheral areas (economic opportunities, services); (c) build human capital among refugees (including 218  |  Informing Durable Solutions for Internal Displacement education, health, and nutrition) and align them with host communities; and (d) transition away from humanitarian assistance where relevant. 479.  Given marked regional differentiation, focus areas should be tailored to specific contexts. For South Suda- nese, there may be limited prospects for economic integration in the short to medium term, but a major effort is needed to support host communities and build human capital. For Somalis, the situation varies across subgroups: some refugees need support to access out-of-camp opportunities; others require help to build human capital. Host communities need significant support to absorb refugees and boost their own potential. Finally, for Eritreans who do not immediately engage in onward movements, major progress could be achieved through an effective out-of-camp approach that combines freedom of movement, work permits, skills development, and job opportunities. 480.  Further research on a needs assessment of displacement-affected regions will be useful. Analytically, the study underscored the unique specificities of the four refugee groups and their respective host communities. More analysis is needed to understand the ad hoc development needs and durable solution prospects of the four refugee groups against host communities. Refugees have distinct displacement drivers, which somewhat determine their dura- ble solutions’ prospects. In turn, host populations largely live in peripheral regions of Ethiopia that have remarkable differences and are far away from one another. There is scope for further qualitative analysis to identify development needs of subregions and respective refugee populations. Such analysis should include cross-border dynamics. 481.  Needs assessment can be complemented by further analysis from the SPS. The SPS collected data on refu- gees and host communities in Tigray and Afar, Somali, and Benishangul Gumuz. However, the case study in this report analyzed and compared refugee groups among them, and against an average host community. There is scope to zoom in on each specific host community to make more meaningful comparisons between each refugee group and its respective host community. Summary of Findings from Cross-Country Analysis 482.  The timing of the surveys in the different countries can have a bearing on the results. The surveys for the five countries were each conducted over the course of a few weeks to a few months. Further, they were conducted at different times of the year, in different countries. Thus, seasonal effects were likely captured in several indicators. Indict- ors such as employment rates might vary in other seasons. This is a caveat for the analysis that follows. 483.  Agricultural IDPs displaced into urban centers face a starkly different labor market environment and higher poverty. Large numbers of displaced populations with agricultural backgrounds are believed to flee from rural areas to urban centers. The report examines whether IDPs with an agricultural background face greater challenges in urban centers. Results show that IDPs and refugees relied more heavily on agriculture before displacement than their hosts do now. Agricultural IDPs have adjusted to current labor markets in different ways, depending on country context, but are often poorer and keener to return to the original residence than nonagricultural IDPs. Their stronger preference for a return might reflect a desire to restore agricultural livelihoods. 484.  IDPs based in camps are poorer, face lower service access, and are more aid dependent than hosts and IDPs outside camps. IDPs and refugees in camps are often believed to receive better services through international Volume B: Country Case Studies  | 219 humanitarian actors than services available locally through country systems. The report uses micro-data to study how the socioeconomic outcomes of IDPs inside and outside camps differ. Contrary to notions of better service provision in camps, service access is worse for camp-based IDPs. While labor market outcomes depend heavily on country context, camp-based displaced are more aid dependent across the board. 485.  Inequality in host communities is linked to worse perceptions of IDPs and signals the poor socioeco- nomic conditions that many members of host communities face. An IDP presence can alter the socioeconomic fabric of host communities. The report examines how inequality and heterogeneity affect hosts’ perceptions toward IDPs. More unequal host communities believe strongly that the arrival of the displaced has worsened job prospects. More prosperous host communities have better relations within the community and more favorable perceptions of the displaced population. Heterogeneity along characteristics other than income also affects the community’s perceptions. Thus, heterogeneity in host communities drives nuanced dynamics with IDPs and warrants greater developmental investment in host community households. 486.  IDPs displaced further from the original residence are more nonagricultural, have been displaced longer, and prefer to return. IDPs who have been displaced for longer periods of time can have specific needs and opportu- nities. Geographical proximity to the original residence can also influence IDP outcomes and return intentions. Results show that IDPs who have been displaced for a longer time have traveled farther from the origin. Longer-displaced IDPs report better social relations with hosts, but this does not translate to higher standards of living. IDPs with agricultural backgrounds are often displaced close to the original residence, while IDPs who live farther from their origin have a higher preference for return. Synthesizing Comparative Findings from Country Case Studies Demographic and Displacement Profiles 487.  Demographic and displacement profiles complement the poverty assessment of displacement situa- tions. While the analysis of the standard of living and livelihoods represents the core information collected by the surveys, the demographic and displacement data contextualize the socioeconomic data. First, demographic disaggre- gation provides an evidence base for targeted intervention and is functional to programming. Second, by profiling the displacement situations through displaced people’s perceptions, the analysis integrates poverty data and displacement data to provide policy directions for development-oriented durable solutions. In fact, a comprehensive and integrated analysis of both poverty and displacement dynamics provides more explanatory power to the policy quest for durable solutions. 488.  Age disaggregation among displaced and non-displaced people is similar across the countries, but dependency ratios are higher among the displaced. Displacement situations in Somalia, Nigeria, South Sudan, and Sudan feature a large presence of children, and so do the countries’ societies and refugees in Ethiopia. Individuals under 15 years of age range from 43 percent among IDPs in Sudan to 46 percent among IDPs in South Sudan, 51 percent among IDPs in Somalia, 57 percent among IDPs in Northeast Nigeria, and 59 percent among the four refugee groups in Ethiopia. Host communities experience similar proportions between children and adults. Approximately 50 percent of community members are under 15 years of age in Nigeria, Somalia, and Ethiopia, whereas this percentage is 40 percent for residents of the city of Al Fashir in Sudan and 32 percent for non-displaced urban residents in South Sudan. High 220  |  Informing Durable Solutions for Internal Displacement percentages of children reveal high dependency rates and large household sizes among the displaced. Dependency ratios265 range from 1.1 for IDPs in Sudan, 1.2 among South Sudanese IDPs to 1.4 among Somali IDPs to 1.8 among Nigerian IDPs to 1.9 among refugees in Ethiopia, while for nondisplaced communities in these countries, they are 1, 0.7, 1.2, 1.4, and 1.1, respectively. Displaced groups have larger household sizes in some countries (Somalia, South Sudan, and refugees in Ethiopia) but smaller households in others (Nigeria and Sudan). 489.  Sex disaggregation is uneven across the countries, but women-headed households are more vulnerable than men-headed ones. In Sudan, 50 percent of the IDP households are headed by women, as opposed to 30 percent of households in the host community. Similarly, 65 percent of households among refugees in Ethiopia are headed by women, while trends run the opposite way among host communities: the percentage of men-headed households in Ethiopian host communities is 65 percent. On the other hand, in South Sudan and Somalia, men- and women-headed households have similar proportions, with a slightly greater presence of men-headed households. In Nigeria too, women-headed households account for nearly 40 percent of both IDP and host community households. Women- headed households in South Sudan are smaller in size compared to men, but women have more dependents than men. This means that women who are household heads do not have a spouse—only 17 percent of them have spouses, compared to 78 percent of men who are household heads and have a spouse. This wide difference underpins the socioeconomic vulnerability of women-headed households. 490.  Security and conflict are the most important push and pull factors of displacement, and security is the most important priority for displaced people. Across most displacement situations, IDPs and refugees report that personal security is the main driver for fleeing—except for 40 percent of Somali IDPs who report environmental con- ditions as the main driver for leaving. Reasons for displacement that are associated with armed conflict, violence, and insecurity range from 40 percent to over 90 percent, depending on the displaced group and the specific demographics. In northeast Nigeria, 75 percent of households were displaced due to armed conflict, violence or communal clashes. Nearly 80 percent of South Sudanese IDPs report having been threatened with a weapon after the start of the conflict in late 2013. In Sudan, more than 90 percent of IDPs cite armed conflict in their own village as the primary driver of displacement. While South Sudanese IDPs and refugees in Ethiopia, and Sudanese refugees overwhelmingly perceive security as the main driver, Somali IDPs and refugees in Ethiopia, and Eritrean refugees link their fleeing also to environ- mental conditions and persecution, respectively, in addition to conflict. The case of Somalis is important as it testifies the blurred boundaries between conflict- and environmental conditions–induced IDPs: 40 percent of Somali IDPs and 25 percent of Somali refugees in Ethiopia report being displaced due to drought, famine or flood. Similarly, reasons for settling in a specific location in displacement mirror the drivers of displacement: the data reveal what the main priorities for displaced people currently are, that is, during displacement. Personal security is the greatest priority when displaced people decide where to settle, followed by some group-specific reasons including access to humanitarian aid and employment opportunities. 491.  Security should be central to the adoption of durable solutions, including local integration. These specular trends in push factors (that is, causes for fleeing) and pull factors (that is, incentives for settling in) are revealing as causes connect to needs and to solutions, and ‘before’ displacement dynamics connect to ‘during’ displacement situations and future intentions. Cause- and need-based lenses are at the core of the study’s approach. A focus on security is therefore 265. Number of non-working age dependents (children, disabled, elderly) per working age household member. Volume B: Country Case Studies  | 221 central to durable solutions. Obviously, a security-focused approach is applied to potential return situations. However, more important for development actors, security concerns should be central to local integration options, which—for a host of reasons—are mostly pursued today. Household-level data highlight that people will decide to move again and again until they find safety. Thus, local integration solutions should focus on displaced host community ties, in addition to improved standards of living and expanded economic opportunities. Prevention approaches, including inclusion practices, shared resources, and voice and participation, may play important roles too. 492.  Most IDPs in Nigeria, Somalia, South Sudan, and Sudan are not displaced far from home, which has impli- cations for policies. As security is the lead priority, people flee their home and settle in areas they perceive as secure. In both South Sudan and Somalia, most IDPs (approximately 70 percent) report being displaced in the same district where they originally lived and did not have to travel far from home to find safety, especially in the presence of avail- able humanitarian aid. In Nigeria, 95 percent of IDPs are displaced within the same state regardless of whether they are in camps or living among host communities. Similarly, in Sudan, most IDPs did not travel very far—97 percent lived in the same state before displacement as they do now, North Darfur. The practical implication associated with the short distance between the place of habitual residence and the current displacement location represents an important dimension for durable solution policies. While fleeing a limited distance does not equal experiencing a smaller trauma from displacement, not having traveled far from home to find safety implies, that those IDPs may experience a relatively easier adaptation in socioeconomic terms. IDPs that are closer to home may also increase the chances for return or fam- ily reunification. In fact, family separation due to displacement is often high, ranging from 25 percent among refugees in Ethiopia to 37 percent South Sudanese IDPs, though in Nigeria it is reported at 12 percent. Thus, durable solution policies for IDPs may want to take into account the distance that IDPs travel. 493.  Most IDPs and refugees in the five case studies prefer to stay in the current location, an answer that hints to their feeling of being stuck. Displaced people were asked about their present and future intentions in line with durable solution options. They were asked if they preferred to (a) stay in the current location and not move, (b) return to country or area of origin, or (c) resettle to a new area or country. On local integration, they were not specifically asked whether they wanted to integrate locally, but only their intention of either moving or staying. Nearly 50 percent of the IDPs in Sudan, 58 percent IDPs in Nigeria, 58 percent South Sudanese IDPs, and 70 percent Somali IDPs reported wanting to stay in their current location. This trend is mirrored in Ethiopia—half the refugees in Ethiopia want to stay where they are, in camps. Percentages of those wanting to return range between 16 percent of refugees in Ethiopia to 23 percent Somali IDPs to 33 percent of South Sudanese IDPs and 25 percent of Nigerian IDPs, and 25 percent of Suda- nese IDPs. The preference to return among Nigerian IDPs is driven by those in camps—76 percent of whom want to return compared to 28 percent of those who live among host communities. These findings point to the consideration that IDPs and refugees, apart from the hosted IDPs in Nigeria, feel stuck in their current situation and are unable to plan for the future. In Sudan, the protracted nature of the displacement, with IDPs having stayed in the camps for around 15 years, has led to the camps obtaining the features of more permanent settlements. In the less protracted displace- ment situations, the answer that most displaced people give (‘Do not want to move/Want to stay in current location’) hint to a personal state of idleness. While durable solution policies should actively consider displaced people’s prefer- ences, there is a need to address the profound lack of agency that displaced people exhibit. Economic opportunities that strengthen self-reliance are key, among others. 222  |  Informing Durable Solutions for Internal Displacement 494.  In line with the drivers of displacement and priorities for settling in a location, IDPs perceive security as being the most important factor in any future decision. Future intentions represent the third piece of the puzzle in the link between causes, needs, and solutions—an approach that corresponds to before, during, and after displace- ment. While IDPs in the four countries and refugees in Ethiopia have different views on which would be their preferred durable solution, all answers for most groups converge on security as the main factor leading their decision to either stay, return, or resettle to a new area or country—even when displacement driven by reasons other than conflict, as in the case of drought-displaced Somali IDPs. 495.  Displacement timing of IDPs is strongly associated with conflict dynamics. South Sudan, Nigeria, Sudan, and Somalia have different experiences when it comes to the timing of displacement. South Sudan, Nigeria, and Sudan experience displacement events that run in parallel to conflict outbreaks: escalation of violence was clearly followed by dramatic peaks in displacement events. On the other hand, displacement spikes in Somalia do not correlate clearly with either conflict or climate events. These differing experiences suggest how displacement patterns reflect the spe- cific conflict dynamics in the countries. In South Sudan, there is a civil war with escalation and pauses that are in line with classic conflict cycle dynamics: forced displacement is an extreme manifestation of these shock events. In Nigeria, violence from the Boko Haram insurgency peaked between mid-2013 and 2015 when Boko Haram captured and occu- pied thousands of square kilometers of towns and villages. More than half of Nigeria’s IDPs fled their homes during this time. In Sudan, most of the IDPs surveyed were displaced in the initial phase of the Darfur conflict in 2003–04, although violence outbreaks have spiked sharply in recent years also. In contrast, forced displacement in Somali regions runs more independently from conflict dynamics because of the very nature of conflict and instability in the country. Soma- lia has underlying protracted conflict dynamics characterized by an overlap between power fragmentation, resource depletion and environmental degradation, fragility of institutions, and decades-long armed violence. As a result, forced displacement is also multicausal and continuous. Standard of Living 496.  IDPs live in staggering levels of poverty, largely more so than hosts. About 74 percent of IDPs in Somalia (compared with 66 percent of the national population), 82 percent IDPs in Sudan (compared to 60 percent hosts), 87 percent of Nigerian IDPs (compared with 83 percent hosts), and 91 percent of IDPs in South Sudan (compared with 75 percent of the urban population) are living below the international poverty line of US$1.90 PPP (2011) per day per capita.266 South Sudanese IDPs have the highest poverty rates. Similarly, 65 percent of refugees in Ethiopia (compared with 27 percent of the host community) are living in poverty, with South Sudanese refugees facing a higher incidence of poverty than other refugee populations in the country. This may be due to the more recent and devastating nature of the crisis in South Sudan. Refugees in Ethiopia, more generally, have the lowest poverty rates, but the gap between them and their hosts is the largest (37 percent). In Nigeria, and to an extent Somalia, South Sudan, and Sudan, resident populations have similar poverty rates to the IDPs because they have suffered from the same conflicts and droughts, albeit to a lesser extent. 266. Refugees and IDPs also have a larger poverty gap, which measures how much less the poor consume relative to the poverty line. The poverty gap is 28 percent for refugees in Ethiopia (compared to only 6 percent for the host community), 35 percent for IDPs in Somalia (compared to 27 percent for the resident population), and 54 percent for IDPs in South Sudan (compared to 40 percent for the urban residents). Volume B: Country Case Studies  | 223 497.  IDPs are highly food insecure, often more so than hosts. In Somalia, IDPs are more likely to be poor and highly food insecure (41 percent compared to 23 percent for national residents).267 In Nigeria, though IDPs and hosts have similar poverty rates, IDPs are more likely to be highly food insecure (62 percent compared to 48 percent of hosts). In Sudan, IDPs are twice as likely as hosts to be highly food insecure (64 percent and 32 percent, respectively). Refugees in Ethiopia are also more likely to be poor and highly food insecure (68 percent compared to 24 percent for local residents). In South Sudan, however, IDPs have higher poverty rates but marginally lower hunger rates (58 percent compared to 65 percent for urban residents). This reflects the effectiveness of the food distribution system in South Sudanese IDP camps. Surprisingly, refugees in Ethiopia have the highest hunger rates, despite having the lowest pov- erty rates, and almost half of their food comes from humanitarian assistance. Moreover, the gap between them and the local residents is once again large (44 percent). 498.  Apart from in Sudan, few IDPs have access to improved housing, and they are less likely to have access than their hosts.268 About 1 in 4 IDPs in Somalia (compared with 3 in 5 national residents), 3 in 4 IDPs in Nigeria (com- pared to 9 in 10 host community residents), and only 1 in 100 IDPs in South Sudan (compared with 1 in 5 urban resi- dents) have access to improved housing. Refugees in Ethiopia face similar conditions, with only 1 in 5 (compared with 3 in 5 hosts) occupying improved dwelling. In Sudan, too, hosts are more likely than IDPs to live in improved housing. Across the five displacement situations, South Sudanese IDPs have the worst housing, with 2 in 3 living in tents. Before displacement, however, 2 in 5 had access to improved housing (compared to 1 in 4 refugees in Ethiopia and IDPs in Somalia). This indicates a significant deterioration in housing for South Sudanese IDPs. In Sudan, though IDP dwellings are often unimproved, they are permanent structures and similar to IDPs’ houses before the conflict. This underlines the permanent nature of the Sudanese camps where less than 1 percent of households live in tents. 499.  IDPs have comparable or better proximity to basic services than the non-displaced. Basic services refer to the nearest water point, health facility, school, and market. Access to services refers to the time it takes to reach these amenities. In Somalia, where there are camp and non-camp IDPs, there is little difference in access to services between the two groups. Similarly, in Nigeria, the distance to basic amenities is comparable for IDPs and host community house- holds. Sudan mirrors these trends: for both IDPs and host community households, distances from homes to important services are typically short and considerably shorter than what IDPs faced before their displacement. However, in South Sudan, IDPs and urban residents had similar access before the conflict but now IDPs have better access. In Ethiopia, access has improved for refugees since their displacement and is now similar to the host community. The displaced in Ethiopia and South Sudan have comparable or better access to services because they are all in camps where they are close to service providers. 500.  Most IDPs have access to improved drinking water, and their access is similar (or better) to the non- displaced, but this does not factor in overcrowding.269 Somali IDPs and nationals have comparable access to improved drinking water (78 percent and 77 percent, respectively). Nigerian IDPs’ access to improved water sources is slightly better than host community households (90 percent, compared to 83 percent for host community house- holds). Refugees and local residents in Ethiopia also have similar access (98 percent and 91 percent, respectively). South 267. Food insecurity is calculated using the rCSI, which is a weighted index that combines information on the frequency and severity of coping strategies used in a single score for household food security. 268. Improved housing is a structure made of wood, brick or concrete, and intended for habitation. 269. Improved drinking water sources include household connection, public standpipe, borehole, protected dug well, protected spring, and rainwater collection. 224  |  Informing Durable Solutions for Internal Displacement Sudanese IDPs, however, have better access to improved drinking water than urban residents (99 percent compared with 82 percent, respectively). Similarly in Sudan, about 80 percent of IDP households have access to an improved source of drinking water, a standard that was met by 56 percent at their places of origin and which is only the case for about 60 percent of host households. Across the displacement situations, refugees in Ethiopia and IDPs in South Sudan and Nigeria have better access to drinking water than IDPs in Somalia and Sudan. 501.  While sizeable proportions of IDPs have access to improved sanitation facilities, overcrowding often ren- ders toilets less fit than those used by hosts. In Somalia, IDPs are less likely to have access to improved sanitation (one in two compared to two in three national residents). Likewise, in South Sudan, IDPs are less likely to have access (only one in one hundred compared to one in three urban residents). In Nigeria, IDPs are more likely to share toilets with other households, which increases the health risks associated with insufficient sanitation. In Ethiopia, however, refugees have better access to improved sanitation (two in five compared to one in five local residents). In Sudan, for both hosts and surveyed IDP populations, nine in ten households have access to an improved sanitation facility. However, 55 per- cent of IDP households share these sanitation facilities with other households, compared to 17 percent among hosts. Latrines that are shared by too many individuals pose health risks and discourage their use.270 Trends indicate that the facilities in camps may be improved, but severe overcrowding limits access and increases health risks. 502.  Many IDPs have access to health care, though less than their hosts. About 54 percent of Nigerian IDPs (com- pared to 76 percent of host communities) and 63 percent of IDPs in Somalia (compared with 82 percent of urban and 30 percent of rural residents) have given birth in a maternity clinic or hospital. In Ethiopia, 90 percent of refugees (compared with 88 percent of local residents) gave birth in such facilities. Likewise, about 55 percent of Nigerian IDPs (compared to 77 percent of hosts) and 54 percent of IDPs in Somalia (compared with 82 percent of urban and 31 per- cent of rural residents), and 90 percent of refugees in Ethiopia (compared with 89 percent of local residents), have had their births attended by a midwife, nurse, or doctor. In Nigeria, IDPs hosted by the community had better child delivery practices than IDPs in camps. 503.  IDP children are less likely to be enrolled in school than resident populations, though trends vary at primary and secondary education levels. In South Sudan, IDPs and urban residents have similar primary school attendance rates (72 percent and 76 percent, respectively), but urban residents are more likely to attend secondary school (22 percent compared with 8 percent for IDPs). IDPs in Nigeria have lower primary enrollment than the host community, but higher secondary enrollment rates, though girls have worse educational outcomes in both groups. In Sudan, both primary and secondary school enrollment rates are substantially lower for IDPs than hosts, while there are no discernible sex differences. In Ethiopia, refugees and local residents have comparable primary school enrollment rates (79 percent and 74 percent, respectively), but local residents are more likely to enroll in secondary school (35 per- cent compared with 13 percent for refugees). The sex gaps in Ethiopia and South Sudan, while present, are lower, while Somalia has a negligible sex gap. Many displaced and non-displaced children are not in secondary school because they are still in primary school. 504.  About half of IDPs are literate and their literacy rates are comparable to their hosts. About 53 percent of IDPs in South Sudan (compared with 62 percent of urban and 33 percent of rural residents), 52 percent of IDPs in 270. World Health Organization and UNICEF. 2006. “Core Questions on Drinking Water and Sanitation for Household Survey.” Volume B: Country Case Studies  | 225 Somalia (compared with 73 percent of urban and 46 percent of rural residents), 70 percent of IDPs in Sudan (compared to a similar proportion of hosts), and 55 percent of refugees in Ethiopia (compared with 55 percent of local residents) are literate. The literacy rates of these displaced populations are remarkably similar to their hosts and to each other. This suggests comparable levels of human development. The sex gaps (difference between men and women) are stark, ranging from 16 percent in Sudan to 35 percent in South Sudan. 505.  Overall, the displaced have higher poverty incidence, comparable proximity to services, and more seri- ous overcrowding than resident populations. The displaced are poorer and generally have worse access to housing. In most cases, they are also more likely to be hungry (Nigeria, Somalia, Sudan, and refugees in Ethiopia). However, they have similar or better proximity to many services (water, primary education, and health), while they have worse access to other services (primary or secondary education). In most countries, they also have worse access to sanitation largely due to overcrowding. Yet, the literacy rates of the displaced are similar to the non-displaced. This suggests similar levels of human capital and that the more difficult living conditions are due to displacement. Policy and programming inter- ventions are urgently needed to improve living conditions by investing in food security, housing, and sanitation, and education of school age children. Employment and Livelihood 506.  The IDPs of South Sudan are highly aid dependent, while in Nigeria, Somalia, and Sudan they are more likely to have a source of income. The displaced in South Sudan mostly rely on humanitarian assistance as the source of household livelihood, a trend mirrored in the refugees in Ethiopia. About 83 percent and 76 percent, respectively, of the livelihoods of refugees in Ethiopia and IDPs in South Sudan come from various types of assistance. However, IDPs in Somalia today largely rely on salaries and family businesses, and in Nigeria and Sudan they get income from agriculture. Somali refugees in Ethiopia are also more self-reliant than other refugee groups in the country. In Sudan, agricultural livelihoods are now lower in number, with many households engaging in salaried work. In contrast, IDPs in Nigeria have shifted their livelihoods toward rather than away from agriculture. Almost 70 percent of IDPs in Nigeria today get their livelihood primarily from agriculture, compared to 50 percent before displacement. IDP livelihoods have shifted to agriculture, leading the IDP livelihood structure to become similar to that of host communities. 507.  Employment and labor force participation vary considerably across the displacement contexts. The IDPs in South Sudan and refugees in Ethiopia are less likely to be employed, while in Somalia and Nigeria IDP employment rates are similar to the non-displaced, and in Sudan, IDP youth are more likely to be employed. In Ethiopia, refugees are less likely to be employed than the host community (22 percent compared with 61 percent, respectively). Likewise, in South Sudan, IDP youth and adults are less likely to be employed (32 percent and 48 percent, respectively) than urban youth and adults (61 percent and 85 percent, respectively). However, in both countries, the displaced are more likely to be enrolled in education than the non-displaced, indicating that some of those not employed are gaining skills. In Somalia, IDPs, urban residents, and rural residents have similar employment rates (45 percent, 46 percent, and 44 per- cent, respectively). In Nigeria, IDPs and hosts have similar rates of employment at around 66 percent and 60 percent, respectively. In Sudan, employment levels are similar for adults (25 years and above) and IDPs and hosts, at 69 percent and 63 percent, respectively. Compared to youth (15 to 24 years old) in the host population, IDP youth are more likely to be working (44 percent compared to 25 percent of hosts) and less likely to be in education. Overall, women are more likely to be not working, looking for work, or enrolled in education. Refugees in Ethiopia have the lowest employment 226  |  Informing Durable Solutions for Internal Displacement rates and the highest dependency rates. This is not surprising as there are legal challenges around being able to work as a refugee. 508.  In most cases (except Nigeria), IDPs have shifted from spending time in agriculture and salaried jobs to family businesses. In South Sudan, IDPs have shifted from salaries to helping businesses. They are now more involved in running and helping businesses and less involved in agriculture than urban residents. In Ethiopia, refugees have shifted from agriculture to running and helping businesses. They are now more involved in helping businesses and less involved in agriculture than the local community. In Somalia, however, the employment structure of IDPs has not changed much since displacement, with 67 percent of them involved in the same employment activities. They are now similar to urban residents but more involved in running businesses and less involved in agriculture than rural residents. Similarly, in Sudan, most working IDPs were own-account agricultural workers before their displacement, but today a majority are salaried employees. These changes in employment patterns have important implications for durable solu- tions. For example, those who want to return, depending on the length of their displacement, may no longer have the skills or experience necessary to grow crops or herd animals; and those who want to stay and integrate may need to develop new business skills to be able to compete in the local markets. 509.  IDPs are less likely to have access to productive assets,271 livestock,272 and land than they did before displacement. South Sudanese IDPs had worse access to land, similar access to productive assets, and better access to livestock before displacement, indicating that they may have been comparable to urban residents. They now also have considerably worse access, having lost more assets during the conflict. Sudan mirrors this trend—before displacement, IDPs had much more agricultural land, livestock, or other productive assets than they do now. Refugees in Ethiopia had better access to assets before displacement, suggesting that they may have been better off than local residents. They now have significantly worse access. Massive declines in access to productive assets, livestock, and land limit the ability of displaced populations to create employment opportunities and become self-reliant. In Nigeria, IDPs have lost their agricultural land, but renting land from host communities could be allowing them to maintain agricultural livelihoods.273 510.  Reliance on aid or having sustainable income sources reflects on the local opportunities. The IDPs in South Sudan and refugees in Ethiopia are more likely to rely on humanitarian assistance because they are less likely to be employed or have access to land, livestock, and productive assets. Not having access to assets limits their ability to develop livelihoods and explains the movement away from agriculture and toward helping and running businesses. In Nigeria, though IDPs lost the land they owned, they now rent agricultural land from their hosts, which could explain why many of them have switched to agricultural jobs. In Somalia, too, the employment rates of the IDPs are similar to their hosts. Their livelihood and employment structure has also not changed much since their displacement. In Sudan, IDPs have shifted from own-account agriculture to salaried work and businesses. Policy and programming inter- ventions need to reduce dependency and increase self-reliance among the displaced by (a) supporting freedom of movement and the right to work (for refugees in Ethiopia); (b) creating employment opportunities; (c) investing in skills development; (d) improving access to assets, livestock, and land; and (e) eventually targeting humanitarian assistance. 271. Productive assets include moefer and kember, axe/sickle, plow, weaving, buildering and welding equipment, carpentry tools, woodcutting and block production equipment, refrigerator, private car, and Bajaj. 272. Livestock include cattle, horses, donkeys/mules, pigs, sheep, goats, poultry, and camels. In Ethiopia, beehives are also included. 273. World Bank Group. 2016. “North-East Nigeria. Recovery and Peace Building Assessment.” Volume B: Country Case Studies  | 227 Social Cohesion, Safety, and Access to Social Remedies and Justice 511.  Indicators assessing safety, quality of social relations, and trust give more depth to the poverty profile of displacement. To complement the poverty profile of displaced groups, the study collected perception data on a num- ber of indicators related to safety and security, trust, displaced-host relations, and civic participation. These measures can be loosely combined to assess the level of social cohesion that displaced and non-displaced groups experience. While social cohesion has had multiple applications in development practice, its relationship with forced displacement is novel.274 As forced displacement represents a demographic shock to receiving countries, it has detrimental impacts on social relations.275 There is consensus in the literature over the fact that social relations between the displaced and non-displaced further worsen in the presence of real or perceived disparities over access to services and opportunities, among others. The ensuing competition when access is contested also fuels bad social relations.276 The study finds remarkable differences among nationalities, suggesting that the quality of social relations can be explained through the different length and displacement dynamics. 512.  Trends in safety and security are different across nationalities, and women-headed households tend to be less safe. Somali refugees and IDPs have high perceptions of safety and security compared to low perceptions among South Sudanese refugees and IDPs. Nearly 80 percent of Somali IDPs and over 90 percent of host community members feel either safe or very safe, while nearly 65 percent of IDPs and 80 percent of hosts have confidence in the police to address crime and violence. In Sudan, IDPs feel considerably safe in the day (more than 90 percent) but less so at night (60 percent). Similarly, over 90 percent of Somali refugees in Ethiopia and host community members in that country feel either safe or very safe. While Eritreans and Sudanese in Ethiopia also have high perceptions of safety, only 25 percent of South Sudanese refugees and 20 percent of South Sudanese IDPs feel safe. Women-headed refugee households are less likely to feel safe than men-headed ones in Ethiopia: only 43 percent of women feel safe vs. 80 per- cent of men. Among IDPs in South Sudan and Somalia, women- and men-headed households experience similar levels of safety, with IDP women feeling slightly less secure than men. 513.  Relations between IDPs and resident communities mirror one another and are generally good, but with some remarkable differences across groups. Overwhelmingly, IDPs in Somalia have good ties with the non- displaced. Nigeria reflects a similar trend, where hosts and IDPs enjoy good bilateral relations, and hosts are more likely than IDPs themselves to feel that IDPs do not get sufficient aid. Sudan mirrors the trend too, where 88 percent of the host population and 79 percent of the IDPs evaluate the relationship to be good or very good. With the exception of South Sudanese refugees, over 85 percent of the other refugee groups in Ethiopia report having good relations with host communities. In keeping with this trend, 45 percent of IDPs in South Sudan and 37 percent South Sudanese ref- ugees in Ethiopia report having either bad or very bad relations with host communities. Similarly, host communities’ perceptions about relations with the displaced mirror displaced people’s perceptions about the host: when the latter is positive, the former is positive too, and vice versa. 514.  Differences in displacement dynamics, household income, and in overall level of integration explain dif- ferences in IDPs’ social integration with resident communities. Arguably, differences between the cases are due 274. World Bank. 2018. “Social Cohesion and Forced Displacement: A Desk Review to Inform Programming and Project Design.” GSURR and FCV, World Bank. 275. World Bank. 2017. “Forcibly Displaced: Toward a Development Approach Supporting Refugees, the Internally Displaced, and their Hosts.” 276. World Bank. 2018. “Social Cohesion and Forced Displacement: A Desk Review to Inform Programming and Project Design.” Washington, DC. 228  |  Informing Durable Solutions for Internal Displacement to the length of displacement, income levels, and the level of integration with host communities.277 For the cases of IDPs in South Sudan and all refugees in Ethiopia, a higher income level of the displaced respondent corresponds to a more positive assessment of relations with the non-displaced. Percentages of either good or very good relations range between 67 percent among the poorest quintile of refugees in Ethiopia to 96 percent for the richest quintile. In South Sudan, approximately 25 percent of IDPs among the poorest quintile reported good or very good relations as opposed to nearly 60 percent of the richest quintile. In Nigeria, host communities are sympathetic despite having very high levels of poverty themselves, and the positive relations seem to have translated into some economic cooperation with hosts renting agricultural land to IDPs. 515.  However, a higher level of integration does not correspond to more frequent civic participation. Differences in civic participation across displaced groups do not mirror differences in displaced-host relations. Civic participation represents an additional dimension of social cohesion as it strengthens trust in institutions; however, it runs somewhat counterintuitively. Surprisingly, the degree of civic participation does not follow the degree of integration nor the level of relations with the host community. Over 70 percent of Somali IDPs never attend public meetings nor interact with community leaders. In Nigeria, public participation of IDPs is lower than of hosts. About 70 percent of IDPs in Nigeria have not participated in public meetings, and 60 percent of have not met a community leader in the last year. In Sudan, IDPs and hosts have similar levels of public participation, but both are extremely low (about 10 percent of households in each group have attended a public meeting in the last year, or interacted with a community leader). Among refugee groups in Ethiopia, 75 percent of Somalis never attend public meetings nor interact with community leaders, which is striking since Somalis are the most integrated group by several measures. Percentages of host community’s civic participation are much higher in Ethiopia (70–80 percent) than urban host communities in Somalia (approximately 20 percent), host communities in Nigeria (57 percent), and non-displaced urban residents in South Sudan and Somalia, respectively around 40 percent and 30 percent. 516.  IDPs in South Sudan and Somalia have different experiences of trust in the government and the secu- rity sector. Over 60 percent of Somalis have confidence in the justice system to deliver justice and punish offend- ers, should one be victim of a crime. In addition, nearly 75 percent of Somali IDPs received compensation for lost property or assets due to conflict, which is a percentage similar to what non-displaced Somali residents experi- enced. On the other hand, South Sudanese IDPs largely report government ineffectiveness in addressing the most important issues, including increasing living standards, delivering services, and tackling corruption. Still, South Sudanese express confidence in the future. Although South Sudanese IDPs largely report to having lost housing due to destruction during conflict and abandonment, 60 percent of them do have confidence that they will be able to regain lost housing once violence is over. Access to documentation is also a concern: over 60 percent of South Sudanese IDPs report having lost documents, and only 10 percent have access to document restoration mecha- nisms. Similarly, in Sudan, IDPs are less likely to have legal identification documents than hosts, mostly because obtaining them is too expensive.278 277. The quality of displaced vs. non-displaced relations and other socioeconomic and structural factors, including the length of displacement, are overlooked dimensions in the literature. World Bank. 2018. “Social Cohesion and Forced Displacement: A Desk Review to Inform Programming and Project Design.” Washington, DC. 278. Legal identification documents include birth certificates, nationality certificates, and passports. Volume B: Country Case Studies  | 229 Policy Implications 517.  Policies and programs need to focus on sustainable solutions for permanent settlement of IDPs and ref- ugees. Differential policy approaches must be embraced for populations who want to stay and for those who want to return. For those who want to stay in situ, the focus needs to be on: (a) maintaining and building human capital by providing adequate nutrition and access to health services and education, and (b) improving living conditions in terms of housing and sanitation—mainly in camps. On a longer term basis, the focus needs to gradually shift to: (a) providing opportunities for socioeconomic integration outside of camps, including social cohesion programs, skills development, job opportunities, access to land, and livestock; (b) continuing to build human capital by ensuring adequate health and education; (c) reducing humanitarian assistance where possible; and (d) supporting host communities to absorb the displaced by improving their living conditions, making services available and enhancing economic opportunities. For those who want to return, the focus needs to be on: (a) providing opportunities for socioeconomic reintegration through skills development, safety net programs, and asset restoration; (b) continuing to build human capital; (c) pro- viding humanitarian assistance where necessary; and (d) improving host communities living conditions to support the reintegration of returnees. 518.  In Nigeria, differences between IDPs in camps and those in host communities have implications for pol- icy. IDPs living in camps tend to have worse housing, more overcrowding, and lower school enrollment than those living in host communities. The data also indicate economic interaction among hosts and hosted IDPs, with the former renting agricultural land to the latter, providing a means of income. A majority of camp-based IDPs report that they plan to return home, while most IDPs living among host communities intend to stay. This implies a durable solution that involves integration with the host community, at least for the hosted IDPs. However, the needs of the host community, which is itself struggling with poverty and low living standards, must be considered in building a durable solution. A durable solution for camp IDPs will be influenced by the capacities of the host community to further absorb people, and the feasibility of a return for the camp IDPs in the face of uncertainty in the Boko Haram conflict. 519.  In Somalia, a durable solution would involve strengthening the areas where IDPs are located. About 70 percent of IDPs express a desire to stay in their current locations, which are mostly in urban areas. Strengthening the viability of urban and peri-urban areas, and enabling IDPs to better integrate into them, is key. The survey findings high- light that socioeconomic and human development indicators of IDPs are often comparable or even better than those of rural residents, highlighting the vulnerabilities and development deficits confronting rural populations in Somalia. Improving rural access to basic services and investing in socioeconomic infrastructure will be critical in supporting IDPs who wish to return. In the process, the needs of IDPs from different backgrounds will need to be considered. The typology profiles as well as the descriptive analysis highlight the differences between climate-displaced and conflict- displaced IDPs. Uncertainty around security dominates return intentions for both groups, outweighing their potentially different tendencies. It is possible that in a post-conflict stage, differences among the two groups might become more pronounced. Group 1 had more agricultural livelihoods and continues to have more access to agricultural land today, which could create scope for a move (or a return) to agricultural locations. Group 2, in contrast, depended more on salaries and businesses, which might allow for integration in an urban setting, whether at the origin or in a new location. 520.  In South Sudan, the data highlight the urgency of maintaining human capital in the face of sharp losses in physical capital, which is precarious given the unpredictability of conflict eruption. While the country is still facing uncertainty driven by waves of conflict, ensuring that the IDPs do not lose their human capital is crucial. Ensuring 230  |  Informing Durable Solutions for Internal Displacement basic and timely care for young children, particularly infants and those under five years of age, is critical to maintain and enhance long-term human capital which will have a bearing on the future of the nation’s productivity, poverty, and workforce. This is especially pertinent because physical capital in the form of housing and assets have largely been lost, and replacing them in a post-conflict stage will need to be accompanied by strong human capital, which can start being nurtured and built in the current stage through health, education, and skill-based programming. Any durable solutions will depend on the nature of peace, the improvement of the security situation, and the provisions of services after the peace agreement. Mostly built in UNMISS camps, PoCs were not intended as durable. 521.  In Sudan, displacement is extremely protracted with at least half the households intending to stay in the settlement-like camps. The IDP camps are permanent settlements and perceived as such, where relations to the host community are good. Housing in the camps is largely in permanent traditional houses where overcrowding is low and IDPs are mostly working for their main sources of income. A clear majority describe relations with the host community as good, a view that is also shared by the latter. Therefore, the camps do have good potential to become a permanent residence for many. However, the challenges associated with such an integration must be incorporated in a plan for a durable solution. Given the sustained conflict in the region, the situation of the urban Al Fashir population is also dire, which undermines their capabilities to host the displaced. The urban environment contributes to the drastic change in living conditions for most IDPs, which entails a loss of human capital. Before their displacement, IDPs were mostly own-account agricultural workers who, in search of better security, found themselves in an urban environment in the camps. Half of the IDPs want to stay in the camps, while most of the others want to return to their places of origin, which might also be problematic given their long time of absence and the conflict situation. 522.  In Ethiopia, the micro-data results underline differences driven by refugees’ different nationalities, high- lighting the need for responses customized based on the country of origin. The living conditions and aspirations of the Eritrean, Somali, South Sudanese, and Sudanese refugees vary widely, culminating in differing return intentions. Most South Sudanese refugees want to stay in the current location. About 40 percent of Somali and Sudanese refugees also want to stay in the current location. Of the remaining 60 percent, most Somalis want to go to a new country while most Sudanese want to return. Most Eritreans want to move onto a new country. The GoE should continue with the implementation of their strategic approach to improve the rights and expand services to benefit households that want to stay in the current location and require better conditions to ultimately bring a durable solution to their displace- ment. Specific household characteristics of the different nationalities also point at customizable durable solutions. For instance, South Sudanese refugee households are almost unanimously headed by women. Policy efforts for this group should consider gender-based vulnerabilities related to domestic work and caring labor, in addition to GBV and discrim- ination against women—common in forced displacement contexts. Somali refugees report highly positive relations with hosts, which can be a foundation for economic interaction with hosts for the aid-dependent refugees. Evidence to Inform Policy 523.  Important data gaps are present in understanding forced displacement, and especially internal dis- placement, to inform policy solutions.279 Data on forced displacement suffer from numerous flaws related to data 279. World Bank. 2017. “Forcibly Displaced: Toward a Development Approach Supporting Refugees, the Internally Displaced, and Their Hosts” Washington, DC. See also ECOSOC. 2015. “Report of Statistics Norway and the Office of the United Nations High Commissioner for Refugees on Statistics on Refugees and Internally Displaced Persons.” Statistical Commission, Forty-Sixth Session E/CN.3/2015/9. New York: ECOSOC; Kriebaum. 2016. “Their Suffering, Our Burden? How Congolese Refugees Affect the Ugandan Population.” World Development 78: 262–287; Ruiz, I., and C. Vargas-Silva. 2013. “The Economics of Forced Migration.” The Journal of Development Studies 49 (6): 772–784; Zetter, R., and C. Vargas-Silva. 2011. Assessing the Impacts and Costs of Forced Displacement. Volume I. Oxford, U.K., and Washington, DC: Refugee Studies Centre and World Bank. Volume B: Country Case Studies  | 231 methodology and clear definitions, reliability, and comparability across countries, and logistical concerns that make collection of primary data challenging, among others. There is a lack of micro-level data and understanding on the socioeconomic conditions of displaced populations, which hampers effective and targeted policy intervention. Data on IDPs are particularly problematic: the boundaries between an IDP and a migrant are not consistent across countries, as IDP numbers and presence are often politicized. Finally, host communities living in the vicinity of forcibly displaced groups are an overlooked dimension of the research on forced displacement. The policy and data communities are currently addressing these flaws, in addition to making efforts to standardize the criteria for when internal displacement ends. The current policies focus on more integrated approaches between the displaced and non-displaced should be accompanied by more research in this direction. 524.  This study helps fill data gaps on the dynamics of internal displacement, both within the displaced pop- ulations and in comparison to the non-displaced. The study on micro-level data on four IDP situations helps fill the current data gap and aims to inspire similar studies in other forced displacement contexts. The results on refugees in Ethiopia complement the IDP analysis for Somalia, South Sudan, and Sudan. The study demonstrates that the collection of household-level data on displaced and non-displaced people enable a comprehensive analysis on causes, needs, and solutions, while allowing cross-country comparisons. In sum, the report’s innovative approach includes (a) collec- tion and analysis of micro-level quantitative data, (b) systematic analysis of both displaced people and host commu- nities, (c) cross-country comparison, and (d) integrated analysis of causes and needs through socioeconomic data and perceptions, which feeds into the search for durable solutions policies. 525.  Four internal displacement situations in Sub-Saharan Africa are analyzed and compared. Through the analysis of household-level surveys conducted with displaced and non-displaced populations, the report sheds light on four internal displacement situations in Sub-Saharan Africa: IDPs in northeast Nigeria, across Somalia, in four camps in South Sudan, and in two camps in Sudan. Additionally, refugees in camps in Ethiopia are also surveyed, from four ori- gin countries including Somalia, South Sudan, and Sudan. The study profiles IDPs according to demographics and fac- tors leading to displacement. It assesses a series of socioeconomic indicators associated with poverty levels, standards of living, livelihood means, and social cohesion. Household-level data were collected and indicators were developed on each of these areas. To contribute to durable solution policies, the report gives a snapshot on forced displaced people’s drivers of displacement, their present material conditions, and their future intentions, following a causes-, needs-, and solutions-based approach. The resulting analysis constitutes a robust cross-country evidence base to be used to inform policies on durable solutions. 526.  The study makes a positive contribution in the areas of micro-data collection methods, understanding of micro-level internal displacement dynamics, and policy directions on durable solutions. The report’s value stems from its uniqueness of comparatively and systematically treating internal displacement at the micro level across four cases, complemented by a study of refugees in Ethiopia. Crucially, the study opens the way for an evidence-based understanding of poverty dynamics of internal displacement (including both displaced and non-displaced groups). On the policy side, the report aims to influence decision makers on the policy options to be devised and implemented, based on the study’s findings. In particular, the study makes a contribution in the following areas: (a) An innovative evidence-based approach that links causes, needs, and solutions. To provide a comprehensive analysis, the study adopts a three-pronged approach around causes of displacement, needs of displaced people 232  |  Informing Durable Solutions for Internal Displacement and host communities, and potential policy solutions. Causes, needs, and solutions translate into the three phases of ‘before’, ‘during’, and ‘after’ displacement, with the aim of devising policy directions that are based on compre- hensive evidence. Thus, the study connects the causes of displacement (‘before’) to the present socioeconomic conditions and needs at the time the surveys were conducted (‘during’) to the future prospects of displaced groups and the possible solutions that they envisage for themselves (‘after’). The micro-level analysis is able to connect IDPs’ preferences and priorities to specific development needs. In turn, this knowledge informs durable solutions policies. (b) A comparative methodology across countries. The study is unique as it adopts a comparative method on two different levels. First is its cross-country comparison. Using the same methodology (that is, country-specific household-level surveys that were designed in a coordinated manner), the study has included four stand-alone IDP case studies and a comparative refugee study. All findings are compared across countries to determine relative poverty levels of the different displaced groups. Comparisons are limited to the household-level data, bearing in mind the heterogeneity of the displacement situations. (c) A comparative methodology between displaced and non-displaced. Second, the study puts special emphasis on the relationship between displaced populations and non-displaced populations, which in most cases (Nigeria, Somalia, Sudan, and Ethiopia) include host communities. Whether in camps or not, displaced people necessarily interface with and have an impact on their social and economic surrounding. Host communities can reap increas- ing economic benefits from expanded local markets (including labor, real estate, and goods and services). The study compares the displaced and non-displaced on access to and quality of services, and on labor force partic- ipation, among others. It also assesses the quality of social ties between the two groups, which is critical for local integration. (d) The comparison between IDPs and non-displaced residents provides an evidence base to link displace- ment factors to specific vulnerabilities. Analytically, the study emphasizes specific vulnerabilities that are directly linked to displacement factors. For IDPs, vulnerabilities concern both the protection and assistance needs that are attributable to their condition of the displaced. For non-displaced host communities, displacement-related vul- nerabilities concern the negative impacts from the influx of displaced people, including strains in service delivery and distortions of local markets. The comparison between the displaced and non-displaced is therefore functional to assess to what extent those vulnerabilities are displacement specific. In a systematic and comparative way, the study demonstrates that forced displacement factors either drive or enhance the socioeconomic vulnerability of IDPs. (e) Using perception data, a displacement profile identifies causes of displacement, priorities and needs, and future intentions of displaced people, all of which feed into policy directions for durable solutions. The household surveys that were conducted in the four IDP case studies and the refugee study in Ethiopia collected critical data on (i) displaced people’s perceptions about the immediate reasons for fleeing to and settling in a cer- tain location; (ii) feelings about displaced people’s priorities while in displacement; and (iii) perceptions related to future intentions. In line with the cause-, need-, and solution-based approach, this household-level data enable the analysis of (i) household-level causes and drivers of displacement; and (ii) what the incentives and immediate pri- orities are for displaced groups. Together with poverty data, the understanding of displaced people’s perspectives about the reason for fleeing, their present priorities, and their future intentions feed into policy directions for dura- ble solutions. As it was demonstrated, there are important connections between displaced people’s perceptions and priorities ‘before’, ‘during’, and ‘after’ displacement. Appendices Appendix A. Nigeria IDP Survey Sampling The Nigeria IDP survey is a household survey that was administered in July 2018. It aimed to survey 1,500 IDP households and 1,500 host community households in Northeast Nigeria. This sample is sufficient to produce accurate estimates for Northeast Nigeria. The sample was drawn based on the estimated IDP populations using the IOM’s Displacement Tracking Matrix (DTM) 2017280 as the sampling frame. IDP populations from six states were used to employ a multistage clustered sample design. The states are Adamawa, Bauchi, Borno, Gombe, Taraba, and Yobe. Each state was divided into EAs of 150 x 150 m using geographic information system (GIS) technology. The number of EAs to be selected from each state was obtained proportional to the IDP population sizes in the state. Thus, the first- stage probability of selection for the EA is the IDP population of the EA divided by the total number of IDPs in the strata multiplied by the number of selected EAs. A total of 111 EAs were selected across the six states. These EAs were either in camp settings where IDPs were located inside camps, or in host community settings where IDPs stayed among the non-displaced communities. Thus, 48 EAs were camp settings and 77 were host community settings. In the 77 EAs which were host community settings, host community households were also interviewed. Since the intended sample for host communities was 1,500 households, several host community EAs were oversampled for host households. The first-stage probability of selection of the EA is the same for both IDP and host community households. Thus, the survey is representative for IDPs and host communities where host communities are defined as the non-displaced population that lives in the areas with IDP populations in non-camp settings. All the households in the selected EAs were first listed, and 12 IDP households and 12 host community households (where applicable) were randomly selected and interviewed.281 For IDPs, the second stage probability of selection is the number of selected IDP households in the EA divided by the number of listed IDP households in the EA. For host com- munity households, the second stage selection probability was calculated as the number of selected host community households in the EA divided by the number of listed host community households in the EA. Certain challenges were faced during the listing and survey period. To draw the sample, the listing data were restricted to IDP households and host community households. Households that had duplicate names and contact numbers and that were outside the cluster boundaries were dropped. Additionally, clusters with less than 12 households listed were also dropped. This resulted in further oversampling from existing EAs. Additionally, due to security concerns, bad 280. IOM. 2008. “Nigeria Displacement Tracking Matrix, (DTM) Round 23, June 2018.” 281. In many cases 24, 36, or 48 households were surveyed when oversampling was required to ensure equal numbers of host communities and IDP households across states. 233 234  |  Informing Durable Solutions for Internal Displacement weather, and non-consent by many selected households, the fieldwork was prolonged and only 2,898 out of the 3,000 intended interviews have been used in this analysis. Weights Sampling weights are used to make survey observations representative for the sample. Weights for IDP surveys are constructed to be representative of IDP populations in the different regions and of the overall IDP population living in northeast Nigeria. Similarly, weights for host populations are constructed to be representative of the host households living in host community type settlements. The sampling weight is the inverse probability of selection. The selection probability P for a household can be decomposed into the selection probability P1 of the EA and the selection proba- bility P2 of the household within the EA: (1) P = P1P2 As IDP population in the different regions lived in different camps and host communities, the selection probability P1 of an EA k is calculated as the number of households within the EA divided by the number of households within the region multiplied by the number of selected EAs in the region. ˆk ∗ K n (2) P1 = N where ˆ nk denotes the number of households in EA k (obtained by multiplying the number of households in an EA), K is the number of EAs selected in the corresponding region and N is the total number of households in the region. The selection probability P2 for a household within an EA k is constant across households and can be expressed as H (3) P2 = nk where |H| is the number of households selected in the EA and nk denoting the number of listed households in EA k. Usually, the number of households per EA is 12 while a few exceptions exist due to invalid interviews. Appendix B. Somali HFS Sampling Wave 2 of the SHFS employed a multistage stratified random sample with households as the unit of observation. Strat- ification ensured that each subpopulation of interest was sufficiently represented in the sample. Strata were defined across Somali regions based on two main criteria: administrative location (prewar regions and emerging states) and population type (urban areas, rural settlements, IDP settlements, and rural nomadic). Based on this stratification strat- egy, the sample consists of 57 strata, of which 16 are exclusively urban strata, 15 exclusively rural with 2 strata being combined urban-rural.282 In addition, there are 7 IDP and 6 nomadic strata (Table B1). The design had to take into consideration the security situation, ensuring that it was safe for enumerator teams to visit households. The security 282. Due to missing IDP boundaries in Galgaduud, the corresponding IDP stratum was not generated. The sample size in some strata was too small, thus urban and rural areas were merged into one single stratum; this was the case for Sool and Sanaag. Volume B: Country Case Studies  | 235 assessment was conducted by producing an access map depicting which areas are deemed secure and which are deemed insecure.283 No interviews were conducted in insecure areas. The security assessment led to the complete exclusion of strata Middle Juba urban and Middle Juba rural. The resulting sample was thus representative of the entire Somali population within secure areas. Several subpopulations were of particular interest and were therefore deliberately oversampled. This concerned, on the one hand, the urban centers of Mogadishu, Baidoa, and Kismaayo; and, on the other hand, fishery livelihood zones in coastal areas. In addition, a sample of 500 households from IDP host communities was added to the survey immedi- ately before fieldwork. The sampling frame consisted all urban EAs adjacent to IDP settlements across all regions. The resulting list was stratified implicitly by the prewar region. About 42 EAs were selected in the first stage with probability proportional to size. Second and final stage sample selection followed the protocol employed in urban and rural areas. There were 500 expected interviews in IDP settlements. Additionally, urban and rural resident populations had a chance to identify themselves as IDPs. The total planned sample size was 6,384 interviews. The sample size was determined to allow for high-precision con- sumption estimates, with less than 1 percent relative standard errors, both overall and for key subpopulations as speci- fied in the preceding paragraph. The sample was allocated across strata using optimal allocation (Neyman) techniques, with the aim of minimizing the global sampling error of the key estimate—overall consumption.284 Hence, the number of households to be interviewed per stratum was mainly determined by the variability of consumption within the stratum. The variability of consumption was derived from the results of the SHFS Wave 1. The population size is also a parameter in optimal allocation, but it only matters for very small strata (below 10,000 households). In the absence of a recent population census, the population of each stratum was derived from UNFPA’s 2014 PESS, which contains detailed estimates for each population type and administrative unit of interest.   TABLE B1    Sample overview of Somali HFS 2017–18 Administrative Population Total Total Strata ID unit type interviews Total EAs Emerging state interviews 1 Central Regions IDP 36 3   Central Regions 684 2 Galmudug IDP 0 0   Galmudug 576 3 Jubaland IDP 84 7   Jubaland 1,248 4 Mogadishu IDP 108 9   Banadir 984 5 North East IDP 192 16   North East 840 6 North West IDP 24 2   North West 732 7 South West IDP 24 2   South West 1,296 8 Central Regions Nomadic 60       (continued) 283. The access map is based on information from key informant interviews with security experts and regional coordinators based in the field. The final assessment was supported by publicly available information and incident reports provided by a local security company. Information in the access map was then triangulated with security analysts from a security NGO and private security company. 284. Optimal allocation is given by nh = n × (Nh × σh)/[Σ(Ni × σi)] where nh is the sample size for stratum h, n is total sample size, Nh is the population size for stratum h, and σh is the standard deviation of stratum h. 236  |  Informing Durable Solutions for Internal Displacement   TABLE B1    Continued Administrative Population Total Total Strata ID unit type interviews Total EAs Emerging state interviews 9 Galmudug Nomadic 36     Prewar region 10 Jubaland Nomadic 84     Hiraan 264 12 North East Nomadic 96     Middle Shabelle 420 13 North West Nomadic 144     Galgaduud 576 13 South West Nomadic 84     Gedo 228 25 Hiraan Rural 144 12   Lower Juba 996 26 Hiraan Urban 48 4   Middle Juba 24 27 Middle Shabelle Rural 264 22   Bari 420 28 Middle Shabelle Urban 48 4   Mudug 324 29 Galgaduud Rural 144 12   Nugaal 96 30 Galgaduud Urban 396 33   Awdal 84 31 Lower Juba Urban 804 67   Sanaag 108 32 Gedo Rural 108 9   Sool 48 33 Gedo Urban 48 4   Toghdeer 192 34 Lower Juba Rural 108 9   Woqooyi Galbeed 300 35 Middle Juba Rural 0 0   Bakool 84 36 Middle Juba Urban 0 0   Bay 900 37 Banadir Urban 792 66   Lower Shabelle 312 38 Bari Rural 48 4   Banadir 984 39 Bari Urban 264 22       40 Mudug Rural 24 2   Urban/rural/IDP/nomad 41 Mudug Urban 96 8   urban 3,936 42 Nugaal Rural 12 1   rural 1,356 43 Nugaal Urban 36 3   IDP 468 44 Awdal Rural 24 2   nomad 504 45 Awdal Urban 36 3       46 Sanaag Urban+Rural 72 6   Oversampled populations 47 Sool Urban+Rural 24 2   Fisheries 324 48 Toghdeer Rural 12 1   Baidoa 540 49 Toghdeer Urban 108 9   Kismaayo 612 50 Woqooyi Galbeed Rural 36 3   Mogadishu 900 51 Woqooyi Galbeed Urban 156 13   Host communities 504 52 Bay Urban 540 45     53 Bakool Rural 48 4     54 Bakool Urban 12 1     55 Bay Rural 180 15     56 Lower Shabelle Rural 204 17     57 Lower Shabelle Urban 48 4     Not applicable Host community Urban (IDP sample adjacent) 504 42     Total 6,384     Volume B: Country Case Studies  | 237 Weights The sampling weight of each household is the inverse of its probability of selection. Its probability of selection is the combination of selection probabilities at each stage of sample selection. A household’s probability of selection is the probability of selection of the primary sampling unit in which it is located, multiplied by the probability of selection of the secondary sampling unit in which it is located, and so on. Urban (non-host communities) and rural households In urban and rural households, the EA was the primary sampling unit and the enumeration block (EB) was the second- ary sampling unit. Enumerators followed a micro-listing protocol on the ground in which they first listed all the struc- tures in the EB, selected a structure, and then, listed all the households in the selected structure. Thus, the probability of selection for urban and rural households is the following: EAj Hi BSi SSk HSm = Phij PP= PP 1 2 3 4 , H j Bi Sk Hm where EA H P1: Probability of selecting the EA, given by j i . Hj BS P2: Probability of selecting the enumeration block, given by i . Bi SS P3: Probability of selecting the structure, given by k . Sk P4: Probability of selecting the household, given by HSm . Hm EAj: Number of EAs selected in stratum j. Hi: Number of households in the sample frame for the original EA i. Hj: Number of households in the sample frame in stratum j. BSi: Number of blocks selected in EA i. Bi: Total number of blocks in EA i. SSk: Number of selected structures in block k. Sk: Total structures in block k. HSm: Number of households selected in structure m. HLm: Total number of households in structure m. 238  |  Informing Durable Solutions for Internal Displacement Urban and host communities Since the host community sample was drawn from a subset of urban EAs, urban households selected in the host com- munity’s sample were part of two separate sampling processes. They thus had two positive probabilities to be selected into the final sample. To reflect this, the probability of selection for these groups is the following: BSi SSk HSm = Phij PP= PP 1 2 3 4 , Bi Sk Hm where P1: Probability of selecting the EA given by two successive sampling processes P1 = (P1a + P1b – P1a * P1b), such that, EAj Hi EAhost Hhost =P1a = and P1b . Hj Hhost EAj: Number of EAs selected in urban stratum j. Hi: Number of households in the sample frame for the original urban EA i. Hj: Number of households in the sample frame in urban stratum j. EAhost: Number of EAs selected in the host community sample. Hhost: Number of households in the sample frame for the original host community EA. Hhost: Number of households estimated in the host community sample. BSi P2: Probability of selecting the enumeration block, given by . Bi SSk P3: Probability of selecting the structure, given by . Sk P4: Probability of selecting the household, given by HSm . Hm BSi: Number of blocks selected in EA i. Bi: Total number of blocks in EA i. SSk: Number of selected structures in block k. Sk: Total structures in block k. HSm: Number of households selected in structure m. HLm: Total number of households in structure m. Volume B: Country Case Studies  | 239 IDP households For IDP settlements, wave 2 of the SHFS employed a slightly different sampling strategy. IDP settlements were first sampled with probability proportional to size to determine the number of EAs to be selected in each settlement. Then, the probability of selection follows the same schema as for urban and rural households, multiplying the first-stage probability of selection with the probability of selecting EA, which is selected with probability proportional to size, with the size given by the number of equal size EBs in the EA. This is then multiplied by the probability of selection of the EB, the structure, and the household—all of which are selected with equal probability. Thus, the probability of selection of IDP households is given by: C j Hc EAc Bi BSi SSk HSm =Phijc PP= PP P 1 2 3 4 5 , H j Bc Bi Sk Hm with C j Hc P1: Probability of selecting the IDP settlement, given by . Hj EA B P2: Probability of selecting the EA, given by c i . Bc BS P3: Probability of selecting the enumeration block, given by i . Bi SS P4: Probability of selecting the structure, given by k . Sk P5: Probability of selecting the household, given by HSm . Hm CAj: Number of camps selected in stratum j. Hc: Number of households in the sample frame for camp c. Hj: Number of households in the sample frame in stratum j. EAc: Number of EAs selected in camp c. Bi: Number of blocks in the original EA i. Bj: Number of blocks in camp c. BSi: Number of blocks selected in the EA i. Bi: Total number of blocks in EA i. SSk: Number of selected structures in block k. Sk: Total structures in block k. HSm: Number of households selected in structure m. HLm: Total number of households in structure m. 240  |  Informing Durable Solutions for Internal Displacement Appendix C. Crisis Recovery and High Frequency Survey South Sudan Sampling: CRS Forced displacement in South Sudan is studied using surveys on IDPs, refugees, and urban residents. Displacement profiles for South Sudan are drawn using data from the CRS 2017 for IDPs, and the SPS 2017 for refugees in Ethiopia. Dis- placed populations are also compared to urban resident populations in 7 of the 10 prewar states of South Sudan, using the HFS 2017. Comparisons are drawn among IDPs, refugees, and urban residents to draw the differences between the displaced and non-displaced. Heterogeneity among IDPs is analyzed using subgroups based on the sex of the household head, the camp where the IDPs are located, and the consumption quintiles, for household-level outcomes (Table C1). Sex and age cuts are made for analysis at the individual level.   TABLE C1    Heterogeneity among IDP households Comparison group Percentage of CRS sample Man-headed household 54 Woman-headed household 46 Bentiu PoC 40 Bor PoC 2 Juba PoC 37 Wau PoC 21 Poorest quintile 20 Quintile 2 20 Quintile 3 20 Quintile 4 20 Richest quintile 20 Source: Authors’ calculations using CRS 2017. South Sudan is a fragile country with several security constraints for field work. The sampling methodology was adapted to the context by excluding several inaccessible areas. The CRS was designed to be representative at the camp level for the PoC camps. Databases and registries for IDP camps are often outdated given the widespread and continuing displacement. Satellite imagery of the camps was therefore used as the sampling frame. Four of the largest PoC camps with defined boundaries were selected. Visible camp boundaries were essential to identify how many households were in each camp. The four camps are Bentiu PoC, Bor PoC, Juba PoC, and Wau PoC, located in the prewar states of Upper Nile, Jonglei, Central Equatoria and Western Bahr-el-Ghazal, respectively. All four camps are in urban areas. The sample follows a stratified two-stage clustered design. Within each stratum, the primary sampling unit is the EA and within each EA, 12 households were selected as the unit of observation. Each camp was selected as a stratum, with a target of 600 interviews per stratum. Satellite imagery of the camps was used to determine the number of structures in the camp. The strata were divided into EAs and from each stratum, 50 EAs were selected proportional to size. Each EA was further divided into 12 blocks. One structure was selected per block, and if the structure had more than one household, one household was selected from the structure for interview. Thus, 12 households were interviewed per EA. Volume B: Country Case Studies  | 241 The exception was Bor PoC, where a census was conducted as there were only 611 households and 8 EAs in the camp (Table C2).   TABLE C2    Sample characteristics: CRS South Sudan Overall Bentiu PoC Bor PoC Juba PoC Wau PoC Sample size (households)   2,396    597   611    589    599 Covered households  31,093 12,414   611 11,463  6,605 Sample size (individuals)  12,571  3,832 2,474  2,479  3,786 Covered individuals 173,339 80,321 2,474 48,895 41,649 Number of EAs     158     50     8     50     50 Source: Authors’ calculations using CRS 2017. Along with satellite imagery, a micro-listing approach was employed to maximize accuracy of the sampling frame. After identifying strata, EAs, and blocks using satellite imagery, enumerators conducted a micro-listing of the block, counting and listing the number of structures in the block. One of the structures was then randomly selected (using the CAPI software, Computer Assisted Personal Interviewing Software). If the structure had only one household, the household was interviewed. If the structure had more than one household, the enumerator listed the number of households in the structure, and one household was randomly selected (by the CAPI software) for interviewing. Weights: Crisis Recovery Survey Sampling weights are used to make sampled observations representative of the entire survey population. Observations from all camps were weighted with the exception of the Bor PoC, the smallest camp, where a census was conducted. The sampling weight is the inverse probability of selection. The selection probability P for a household can be decom- posed into the selection probability P1 of the EA and the selection probability P2 of the household within the EA: P = P1P2 The selection probability P1 of an EA k is calculated as the number of households within the EA divided by the number of households within the stratum multiplied by the number of selected EAs in the stratum: ˆk Kn P1 = ∑ k ∈K ˆk n where ˆnk denotes the number of households in EA k estimated using satellite imagery data and K is the set of EAs selected in the corresponding stratum. The selection probability P2 for a household within an EA k is constant across households and can be expressed as: H P2 = nk where |H| is the number of households selected in the EA, and nk denotes the number of listed households in EA k based on the micro-listing. For each EA, 12 households were interviewed. 242  |  Informing Durable Solutions for Internal Displacement Sampling weights were scaled to equal the number of households per strata using the satellite imagery data. Thus, the sampling weight W can be written as: W= c with c = ∑ k ∈K ˆk n P ∑ k ∈K nk Sampling: High Frequency Survey The HFSSS W4 2017 represents the urban population of seven states of South Sudan. The HFSSS W4 was conducted at the same time as the CRS and is representative of the urban areas in 7 of the 10 prewar states. For security reasons, three states in South Sudan (Jonglei, Unity, and Upper Nile) were excluded from the sample design. Like the CRS, the HFSSS W4 employs a stratified two-stage cluster design with EAs as primary sampling units and households as the units of observation. Each of the seven prewar states is sampled as a stratum. Within each stratum, EAs are drawn proportional to size using a sampling frame derived from the Fifth Sudan Population and Housing Census, 2008. For each of the selected EAs, enumerators conducted a listing and 12 households were drawn randomly for interview (Table C3).   TABLE C3    Sample characteristics: high frequency survey South Sudan Northern Western Bahr el Bahr el Western Central Eastern Overall Warrap Ghazal Ghazal Lakes Equatoria Equatoria Equatoria Sample size 944 144 126 137 133 156 95 153 (households) Covered 960,714 172,359 141,818 84,623 98,299 131,616 165,076 166,923 households Sample size 4,903 1,044 724 546 973 620 412 584 (individuals) Covered 5,745,252 109,9556 943,003 519,326 75,6410 716,247 903,411 807,300 individuals Number of EAs 101 14 15 14 15 14 14 15 Source: Authors’ calculations using HFS 2017. Weights: High Frequency Survey The selection probability P for a household can be decomposed into the selection probability P1 of the EA and the selection probability P2 of the household within the EA: P = P1P2 The selection probability P1 of an EA k is calculated as the number of households within the EA divided by the number of households within the stratum multiplied by the number of selected EAs in the stratum. ˆk Kn P1 = ∑ k ′∈K ˆk ′ n Volume B: Country Case Studies  | 243 where ˆnk denotes the number of households in EA k estimated using the Census 2008 data and K is the set of EAs selected in the corresponding stratum. The selection probability P2 for a household within an EA k is constant across households and can be expressed as H P2 = nk where |H| is the number of households selected in the EA and nk denoting the number of listed households in EA k. Usually, the number of households per EA is 12 while a few exceptions exist due to invalid interviews. Sampling weights were scaled to equal the number of households per strata using the Census 2008 data. Thus, the sampling weight W can be written as: W= c with c = ∑ k ∈K ˆk n P ∑ k ∈K nk Questionnaire The CRS and HFS were designed to be comparable. The questionnaire design is identical. Modules on food, nonfood, and durable goods are used to compute consumption-based poverty statistics. Modules on household and individual characteristics are used to chart demographic characteristics like age, sex, and dependency ratios, as well as education and labor outcomes. Modules on welfare include living standards based on access to services as well as respondents’ sense of well-being and opinions. In addition, households identified as displaced have a section on displacement, guided by the IASC Framework to understand reasons for displacement, return intentions, sense of security, relations with the surrounding community, and a variety of pre-displacement outcomes in the standard of living, education, and labor.285 Appendix D. Sudan IDP Profiling Survey Sampling The Sudan IDP profiling survey employed a stratified cluster sampling approach. The sample is divided into four strata, two for IDPs (Abu Shouk and El Salam camps) and two for host communities (the neighboring and the non-neighboring parts of Al Fashir). The planned sample sizes were 996 households each for the two IDP strata and 504 households each for the two host strata, adding up to 3,000 households. The four strata were divided into clusters (or EAs) of similar population, based on a grid developed on a map of the areas. Clusters were selected from each stratum with a uniform probability of selection. This is because there was no reliable population data available which would have made sampling probabilities to size possible. All the households in the selected EAs were to be listed, while 12 were to be selected for interviewing in each cluster in simple random draws. The listing exercise resulted in a significantly lower number of clusters than was planned. To draw the sample, the listing data was restricted to IDP households in IDP clusters and host community households in host community clus- ters. Households that had duplicate names and contact numbers and that were outside the cluster boundaries were 285. The Brookings Institution—University of Bern Project on Internal Displacement. April 2010. IASC. 2010. “Framework on Durable Solutions for Internally Displaced Persons.” 244  |  Informing Durable Solutions for Internal Displacement dropped. Additionally, clusters with less than 12 households listed were dropped too. Moreover, some clusters were found to be inaccessible during listing, largely in the El Salam stratum. The following table lays out the planned vs. listed clusters and the households to be interviewed per stratum (Table D1).   TABLE D1    Planned and listed interviews Strata Households Number of Number of Population type ID Strata selected clusters (listed) clusters (planned) IDPs 1 Abu Shouk camp 996 82 84 2 El Salam camp 996 50 84 Host community 3 Neighboring Al Fashir 504 40 42 4 Non-neighboring Al 504 41 42 Fashir Total 3,000 213 252 To compensate for the lower number of clusters, some of the listed clusters were selected for over-sampling. Selecting 12 households per cluster for interviews would have led to the sample size falling short of the planned 3,000. There- fore, in each stratum the missing clusters were replaced with randomly selected clusters from among those listed. The selected clusters were to have 24 interviews in them. Only the clusters with more than 23 listed households were eligi- ble to be selected for this over-sampling. Weights Sampling weights were calculated with regard to these planned and non-planned characteristics of the sample. There- fore, they were calculated differently for EAs with more than and EAs with less than 24 households. EAs that had less than 24 listed households were not eligible to be over-sampled. There were thus certainly only 12 households selected in these EAs, one of which had the following probability of selection: EAS j 12 P= hij PP= 1 2 EAj Hi where P1: Probability of selecting the EA P2: Probability of selecting the household EASj: Number of selected EAs in stratum j EAj: Number of EAs in the sample frame in stratum j Hi: Number of listed households in EA i Volume B: Country Case Studies  | 245 EAs that had more than 23 listed households were eligible to be over-sampled. The following tree diagram illus- trates the resulting selection process, with the corresponding probabilities in brackets. >23 households Cluster non-selection Cluster selection Cluster (1-P1.1) (P1.1) 0 intv No over-sample Over-sample (1-P1.2) (P1.2) 12 intv 24 intv Household HH non-selection HH selection HH non-selection HH selection (1-P2a) (P2a) (1-P2b) (P2b) In particular, P1.1: Probability of selecting the EA from the sample frame P1.2: Probability of selecting the EA for the over-sample, from those with >23 listed households P2a: Probability of selecting the household, given the EA was not selected for the over sample P2b: Probability of selecting the household, given the EA was selected for the over sample. To calculate the selection probabilities, the assumption is made that households in EAs with more than 23 listed house- holds do not differ systematically from the others in the sample frame. Then the cluster and the household selection processes can be looked at independently. The selection probability of a cluster as the sum of all separate possibilities that it could have been selected is then simply, P1 = P1.1P1.2 + P1.1(1 – P1.2) = P1.1 246  |  Informing Durable Solutions for Internal Displacement Consequently, the selection probability differs between households from oversampled clusters and households from clusters that were not selected again. It is in the first case EAS j 24 = Phij P1= P .1 2 b EAj Hi and in the case without over-sample EAS j 12 = Phij P1= P .1 2 a EAj Hi where again EASj: Number of selected EAs in stratum j EAj: Number of EAs in the sample frame in stratum j Hi: Number of listed households in EA i All else equal, the weight in the first case will be half the weight of the second case, because due to the 24 interviews, it represents half as many households in the EA. Household non-response was accounted for by adjusting the respective probability of household selection. In par- ticular, the probabilities P2, P2a or P2b in the above formulas were changed to reflect the reduced number of actually conducted interviews. This assumes that household non-response occurred randomly, which is difficult to investigate within the scope of this exercise. The resulting weights could not be tested against population data due to the lack of reliable information. There was no population data available for the surveyed for all the surveyed strata, so the final weights could not be scaled to match population totals. However, for the camps, population sizes suggested by the weights are close to the latest population estimates by the IOM. Appendix E. Skills Profile Survey Ethiopia Sampling The chosen survey methodology aimed to survey refugees of four main nationalities—South Sudanese, Somalis, Eri- treans, and Sudanese—living in camps in Ethiopia.286 The list of refugee camps, sites, and locations provided by the UNHCR-Ethiopia as of January 2017 was used as the sample frame. Of all refugee households287 in Ethiopia, 33 percent live outside of camps. An overwhelming majority (85 percent) of these out-of-camp refugees are Eritrean refugees who are either spontaneously settled in different locations of Ethiopia, or who benefitted from the 2010 ‘Out-of-Camp’ policy and as a result, moved to Addis Ababa or other locations outside camps. The refugee households living out of camp were excluded from the sample frame because tracing these households for the survey exercise was not feasible. 286. The case study looks specifically at refugees and does not include IDPs in the country. 287. Household here is defined as all people living in the same dwelling and sharing all meals and finance. Volume B: Country Case Studies  | 247 The survey is therefore only representative of refugees living in camps in Ethiopia. Table E3 lists all the refugee camps in the sample frame. The sample frame was then divided into four strata based on four regions (main refugee group in parenthesis): Tigray Afar (Eritreans), Gambella (South Sudanese), Benishangul Gumuz (Sudanese and South Sudanese), and Somali (Somalis). Each region hosts a predominant majority of one refugee nationality leading to an implicit strat- ification based on nationality. The sample design uses a multistage stratified random sample. Camps in each stratum were divided into EAs of 150 x 150 meters using GIS technology. The number of EAs to be selected from each camp was obtained proportional to the size of the camp. In this way, all the camps in the sample frame were selected in the sample and were surveyed. Within camps, EAs were selected using equal probability to make up the required number of EAs for that camp. In total, 82 EAs were selected from each stratum. All the households in the selected EAs were listed, and 12 households were randomly selected and surveyed per EA, making up to a total of 900 refugee households per stratum (Table E1).   TABLE E1    Number of refugee and host community households interviewed by stratum Stratum Tigray Afar Gambella Benishangul Gumuz Somali Total Refugees 894 439 1,423 871 3,627 Host community 412 0 975 303 1,690 For the host community sample, all households within a 5 km radius of a camp were classified as host community households. Areas within a 5 km radius of camps were divided into EAs of 300 by 300 meters using GIS technology. Of these, EAs marked as residential by Open Street Maps were included in the sample frame. EAs within a stratum were then selected using proportional probability sampling with the probability of selection of an EA equal to the area of the EA outside the camp. In total, 42 EAs were selected for each stratum. Like EAs within camps, all the households in the EAs selected for host community sampling were listed, and 12 households were selected randomly and surveyed per EA, making up to a total of 500 host community households per stratum. Due to security concerns, major revisions were made to the sample during fieldwork. Enumerators in the Gambella region faced repeated security threats and could survey only 439 of the intended 900 refugee households in the region. Because the survey team was withdrawn from the Gambella region, the host community was not surveyed at all. The remaining interviews with refugees in the Gambella region were substituted by oversampling EAs in Benishangule, as 25 percent of the refugee population in this region is South Sudanese. Similarly, in early September 2017, violent conflict in Oromia and Somali regions escalated, rendering some of the camps in the Somali stratum inaccessible. The EAs of Jijiga subregion were replaced by EAs in nonviolent areas of the Somali stratum. Also, as most refugee camps are in remote areas with a sparse host population, the final number of host households surveyed fell short of the original intended sample of 500 host households per stratum. However, despite the changes in the sample, the survey captured a roughly similar number of refugee households of the four main ref- ugee nationalities (Table E2). 248  |  Informing Durable Solutions for Internal Displacement   TABLE E2    Sampled population by country of nationality Number of households Percentage of households in surveyed Country surveyed population (%) South Sudanese 837 16 Somalis 871 16 Eritreans 893 17 Sudanese 1,016 19 Ethiopians—host community 1,690 32 Other country 10 0 Total 5,317 100   TABLE E3    Refugee camps in sample frame Region Nationality of refugees Camp Tigray Eritrean Mai-Aini Adi Harush Shimelba Hitsats Afar Aysaita Barahle Gambella South Sudanese Pugnido Kule Jewi Okugo Tierkidi Pugnido II NGUENYYIEL camp (new) Benshangul Gumuz 75% Sudanese, 25% South Sudanese Sherkole Bambasi Tongo Tsore Somali Somali Ken-Borena Kebribeyah Aw-barre Sheder Bokolmanyo Melkadida Kobe Hilaweyn Buramino Source: UNHCR. Volume B: Country Case Studies  | 249 Weights Sampling weights are used to make survey observations representative for sample. As the sample was divided by stra- tum, weights for refugee surveys are constructed to be representative of refugee populations in the different regions and of the overall refugee population living in camps in Ethiopia. Similarly, weights for host populations are constructed to be representative of the host households living within a 5 km radius of refugee camps. The sampling weight is the inverse probability of selection. The selection probability P for a household can be decomposed into the selection probability P1 of the EA and the selection probability P2 of the household within the EA: (4) P = P1P2 As refugee population in the different strata lived in different camps, the selection probability P1 of an EA k is calculated as the number of households within the EA divided by the number of households within the stratum multiplied by the number of selected EAs in the stratum: ˆk ∗ K n (5) P1 = N where ˆ nk denotes the number of households in EA k (obtained by multiplying the percentage of camp area covered by the EA with the number of households in the camp as information on number of households in an EA was not available prior to listing), K is the number of EAs selected in the corresponding stratum and N is the total number of households in the stratum. For host community sampling, as information on the number of host households living within 5 km of camps in a stratum was not available, the selection probability of an EA for host sampling is calculated as the number of EAs selected divided by the total number of EAs in the stratum. K (6) P1 = T where K is the number of EAs selected in a stratum and T is the total number of EAs in the corresponding stratum. Replacement EAs were assigned the sampling weight of the EA that they were replacing. Due to changes in sample during fieldwork, the number of EAs surveyed in each stratum differed from the original sample. The weights were therefore scaled at the end to correct for the change in the value of K. The selection probability P2 for a household within an EA k is constant across households and can be expressed as H (7) P2 = nk where |H| is the number of households selected in the EA and nk denoting the number of listed households in EA k. Usually, the number of households per EA is 12 while a few exceptions exist due to invalid interviews. Sampling weights were scaled to equal the number of households per stratum using the information for number of households provided by the UNHCR. There was no source of information on number of host households living within 5 km distance of the camps. The weights for host community surveys were therefore not scaled. 250  |  Informing Durable Solutions for Internal Displacement Appendix F. Typologies Ethiopia   FIGURE F1    Clusters of households in 2D and 3D for Ethiopia   Source: Authors’ calculation using SPS 2017.   TABLE F1    Size of each group of refugees in Ethiopia % of the population No. of observations in the sample Group 1 78 2,669 Group 2 22 958 Source: Authors’ calculation using SPS 2017.   FIGURE F2    Housing conditions pre-displacement for Ethiopia 70 50 % of households 30 10 0 Tukul/gottiya Mud/wood Concrete/brick Overcrowded Group 1 Group 2 Source: Authors’ calculation using SPS 2017. Volume B: Country Case Studies  | 251   FIGURE F3    Source of livelihood pre-displacement for Ethiopia 70 60 50 % of households 40 30 20 10 0 Agriculture Wages, salaries, Aid, remittances, & own business & other Group 1 Group 2 Source: Authors’ calculation using SPS 2017.   TABLE F2    Main source of income of groups in Ethiopia Dependent variable Independent variables (1) (2) (3) Pre-conflict: wages, salaries, and own business −0.413*** −0.412*** −0.308** Pre-conflict: aid, remittances, and other −0.828*** −0.830*** −0.743*** Wages, salaries, and own business — −1.188*** −1.072*** Aid, remittances, and other — −1.149*** −1.067*** Household size — — 0.030 Share of children in HH — — −0.013 Household headed by a woman — — −0.215** Literate household head — — −0.682*** Household head employed — — 0.095 Household head unemployed — — 0.119 Region fixed effect Yes Yes Yes Observations 3,627 3,627 3,626 Source: Authors’ calculation. Note: The coefficients were estimated from a logistic regression. Significance level: 1 percent (***), 5 percent (**), and 10 percent (*). 252  |  Informing Durable Solutions for Internal Displacement   TABLE F3    Current household characteristics for Ethiopia   Group 1 Group 2 Household size 5.9 5.0 Dependency ratio 2.1 1.2 Share of children in the household 17 16 Share of elderly in the household 2 2 Share of households with Ethiopian relatives 8 37 Share of households headed by women 71 50 Share of household heads without education 65 38 Share of employed household heads 28 34 Source: Authors’ calculations using SPS 2017.   FIGURE F4    Current access to key services for Ethiopia Electricity Shared toilet Improved sanitation 0 10 20 30 40 50 60 70 80 90 % of households Group 1 Group 2 Source: Authors’ calculation using SPS 2017.   TABLE F4    Poverty and household characteristics of groups in Ethiopia Dependent variable: household group (Group 2 corresponds to the reference group) Independent variables (1) (2) (3) Poor household 0.685*** 0.669*** 0.455 Household headed by a woman — 0.147 −0.099 Literate household head — — −0.773*** Region fixed effect Yes Yes Yes Observations 2,364 2,364 2,364 Source: Authors’ calculation. Note: The coefficients were estimated from a logistic regression. Significance level: 1 percent (***), 5 percent (**), and 10 percent (*). Volume B: Country Case Studies  | 253  FIGURE F5    Information required to decide whether to stay or move for Ethiopia Source: family, friends, or leaders Source: radio, TV, Internet, or written media Have all the information 0 20 40 60 80 100 % of households Group 1 Group 2 Source: Authors’ calculation.   TABLE F5    Living conditions of groups in Ethiopia Dependent variable: household group (Group 2 corresponds to the reference group) Independent variables (1) (2) (3) (4) (5) Household size 0.055* 0.044 0.072*** 0.080*** 0.088*** Overcrowded dwelling 0.016 −0.033 −0.207 −0.265** −0.320** Share of children in HH 0.163 0.018 0.494 0.448 0.415 Household headed by a woman — −0.238** −0.177* −0.171* −0.169 Literate household head — −0.734*** −0.648*** −0.670*** −0.646*** Improved water source pre-conflict — — 0.194* 0.050 0.044 Improved water source — — 0.388 0.500 0.717 Improved sanitation pre-conflict — — — 0.368** 0.449** Improved sanitation — — — 0.277* 0.253 Electricity pre-conflict — — — — 0.367** Electricity — — — — −1.100*** Employment status of head No Yes Yes Yes Yes Region fixed effect Yes Yes Yes Yes Yes Observations 3,627 3,626 3,511 3,487 3,487 Source: Authors’ calculation. Note: The coefficients were estimated from a logistic regression. Significance level: 1 percent (***), 5 percent (**), and 10 percent (*). 254  |  Informing Durable Solutions for Internal Displacement Nigeria   FIGURE F6    Clusters of households in 2D and 3D for Nigeria   Source: Authors’ calculation.  TABLE F6    Size of each group of IDPs in Nigeria % of the population No. of observations in the sample Group 1 74 496 Group 2 26 941 Source: Authors’ calculation.   FIGURE F7    Timing of moving for Nigeria 90 80 70 % of households 60 50 40 30 20 10 0 In less than 12 months In more than 12 months Don't know Group 1 Group 2 Source: Authors’ calculation. Volume B: Country Case Studies  | 255   TABLE F7    Main source of income of groups in Nigeria Dependent variable: household group (Group 2 corresponds to the reference group) Independent variables (1) (2) (3) Pre-conflict: wages, salaries, and own business 0.762*** 0.794*** 0.861*** Pre-conflict: aid, remittances, and other 0.258 0.173 0.243 Wages, salaries, and own business — −0.053 −0.089 Aid, remittances, and other — 0.123 0.006 Household size — — −0.059*** Share of children in HH — — −1.215* Household headed by a woman — — 0.339*** Literate household head — — −0.114 Region fixed effect Yes Yes Yes Observations 1,417 1,417 1,409 Source: Authors’ calculation. Note: The coefficients were estimated from a logistic regression. Significance level: 1 percent (***), 5 percent (**), and 10 percent (*).   TABLE F8    Poverty and household characteristics of groups in Nigeria Dependent variable: household group (Group 2 corresponds to the reference group) Independent variables (1) (2) (3) Poor household −0.505 −0.397 −0.399 Household headed by a woman — 0.441*** 0.464*** Literate household head — 0.168 Region fixed effect Yes Yes Yes Observations 1,388 1,388 1,380 256  |  Informing Durable Solutions for Internal Displacement   TABLE F9    Living conditions of groups in Nigeria Dependent variable: household group (Group 2 corresponds to the reference group) Independent variables (1) (2) (3) (4) (5) Household size −0.042** −0.031 −0.035 −0.031 −0.038 Overcrowded dwelling −0.272*** −0.291** −0.257** −0.262** −0.179 Share of children in HH −0.949 −0.931 −1.051 −1.059 −0.761 Household headed by a woman — 0.405*** 0.398** 0.397** 0.216* Literate household head — 0.027 0.014 −0.035 −0.057 Improved water source pre-conflict — — −0.296* −0.203 −0.218 Improved water source — — 0.648*** 0.639*** 0.637*** Improved sanitation pre-conflict — — — −0.021 −0.032 Improved sanitation — — — −0.265 −0.214 Electricity pre-conflict — — — — −0.254* Electricity — — — — 0.696*** Other controls No No No No Yes Region fixed effect Yes Yes Yes Yes Yes Observations 1,417 1,409 1,389 1,380 1,380 Source: Authors’ calculation. Note: The coefficients were estimated from a logistic regression. Significance level: 1 percent (***), 5 percent (**), and 10 percent (*). Somalia   FIGURE F8    Clusters of households in 2D and 3D for Somalia   Source: Authors’ calculation.   TABLE F10    Size of each group of IDPs in Somalia % of the population No. of observations in the sample Group 1 40 405 Group 2 60 623 Volume B: Country Case Studies  | 257 South Sudan   FIGURE F9    Clusters of households in 2D and 3D for South Sudan   Source: Authors’ calculation.   TABLE F11    Size of each group of IDPs in South Sudan % of the population No. of observations in the sample Group 1 40 1,213 Group 2 60 1,183 Source: Authors’ calculation.   FIGURE F10    Camp for South Sudan   FIGURE F11    Reasons for displacement from origin in South Sudan 100 90 80 80 % of households 70 % of households 60 60 50 40 40 30 20 20 0 10 0 C C C oC Po Po Po rP lla ct illa t in nf ce n er iu tio a au vi fli th nt co en b Bo ge ge t r v flic in con lic a Ju O W Be ot iol in he on im tn v ed c bu sed cr ed Group 1 Group 2 m is Ar D m ea ot Ar cr In Group 1 Group 2 Source: Authors’ calculation. 258  |  Informing Durable Solutions for Internal Displacement   TABLE F12    Main source of income of groups in South Sudan Dependent variable: household group (Group 2 corresponds to the reference group) Independent variables (1) (2) (3) Pre-conflict: crop farming −0.318** −0.289* −0.605*** Pre-conflict: wages and salaries 0.418*** 0.512*** 0.749*** Pre-conflict: owned business enterprise 0.273* 0.421** 0.603*** Pre-conflict: aid 0.853*** 0.995*** 0.970*** Crop farming — 0.049 −0.421 Wages and salaries — −0.813*** −0.871*** Owned business enterprise — −0.711*** −1.282*** Aid — −0.453*** −0.371** Household size — — −0.155*** Overcrowded dwelling — — −0.932*** Share of children in HH — — −1.186*** Household headed by a woman — — −0.092 Literate household head — — −0.119 Region fixed effect Yes Yes Yes Observations 2,395 2,387 2,373 Source: Authors’ calculation. Note: The coefficients were estimated from a logistic regression. Significance level: 1 percent (***), 5 percent (**), and 10 percent (*).   FIGURE F12    Assets and housing pre-displacement in South Sudan Improved housing Ownership of productive assets Ownership of livestock Access to agricultural land 0 20 40 60 80 100 % of households Group 1 Group 2 Source: Authors’ calculation. Volume B: Country Case Studies  | 259  TABLE F13    Poverty and household characteristics of groups in South Sudan Dependent variable: household group (Group 2 corresponds to the reference group) Independent variables (1) (2) (3) Poor household −0.424*** −0.417*** −0.413*** Household headed by a woman — −0.103 −0.075 Literate household head — — 0.054 Region fixed effect Yes Yes Yes Observations 2,382 2,382 2,382 Source: Authors’ calculation. Note: The coefficients were estimated from a logistic regression. Significance level: 1 percent (***), 5 percent (**), and 10 percent (*).   FIGURE F13    Changes in water and sanitation conditions in South Sudan 60 50 % of households 40 30 20 10 0 Better water sources Worse sanitation Group 1 Group 2 Source: Authors’ calculation.   TABLE F14    Living conditions of groups in South Sudan Dependent variable: household group (Group 2 corresponds to the reference group) Independent variables (1) (2) (3) (4) (5) Household size −0.110*** −0.111*** −0.115*** −0.126*** −0.127*** Overcrowded dwelling −0.758*** −0.759*** −0.752*** −0.775*** −0.787*** Share of children in HH −0.869*** −0.851*** −0.878*** −0.882*** −0.904*** Household headed by a woman — −0.036 −0.024 −0.028 −0.014 Literate household head — 0.107 0.076 0.057 0.052 Improved water source pre-conflict — — −0.368*** −0.388*** −0.375*** Improved water source — — −0.512* −0.490 −0.500 Improved sanitation pre-conflict — — — 0.195 0.187 Improved sanitation — — — 0.513*** 0.540*** Improved housing pre-conflict — — — — 0.042 Improved housing — — — — −1.253* Region fixed effect Yes Yes Yes Yes Yes Observations 2,382 2,382 2,375 2,362 2,361 Source: Authors’ calculation. Note: The coefficients were estimated from a logistic regression. Significance level: 1 percent (***), 5 percent (**), and 10 percent (*). 260  |  Informing Durable Solutions for Internal Displacement Sudan   FIGURE F14    Clusters of households in 2D and 3D for Sudan   Source: Authors’ calculation.   TABLE F15    Size of each group of IDPs in Sudan % of the population No. of observations in the sample Group 1 39 701 Group 2 61 1,273 Source: Authors’ calculation.  FIGURE F15    Household characteristics pre-displacement in Sudan Agricultural livelihood Access to agricultural land Improved sanitation Dwelling of Tukul/gottiya 0 10 20 30 40 50 60 70 80 90 100 % of households Group 1 Group 2 Source: Authors’ calculation. Volume B: Country Case Studies  | 261   TABLE F16    Current household composition in Sudan Group 1 Group 2 Household size 6.2 6.0 Proportion of children in the household 38 39 Dependency ratio 1.13 1.14 Share of households headed by women 42 56 Share of literate household heads 65 55 Share of inactive household heads 14 25 Source: Authors’ calculation.   FIGURE F16    Current household characteristics in Sudan Far from main services Without lighting Improved water sources Shelter provided Dwelling of Tukul/gottiya 0 20 40 60 80 100 % of households Group 1 Group 2 Source: Authors’ calculation.   TABLE F17    Main source of income of groups in Sudan Dependent variable: household group (Group 2 corresponds to the reference group) Independent variables (1) (2) (3) Pre-conflict: crop farming 0.342** 0.400** 0.447*** Crop farming — −0.586*** −0.622*** Wages, salaries, and owned business enterprise — −0.725*** −0.766*** Aid, remittances, and other — −0.324* −0.217 Household size — — 0.012 Share of children in HH — — −0.002 Household headed by a woman — — −0.324*** Literate household head — — 0.185* Region fixed effect Yes Yes Yes Observations 1,974 1,974 1,974 Source: Authors’ calculation. Note: The coefficients were estimated from a logistic regression. Significance level: 1 percent (***), 5 percent (**), and 10 percent (*). 262  |  Informing Durable Solutions for Internal Displacement   TABLE F18    Poverty and household characteristics of groups in Sudan Dependent variable: household group (Group 2 corresponds to the reference group) Independent variables (1) (2) (3) Poor household 0.350** 0.332** 0.345** Household headed by a woman — −0.312*** −0.234** Literate household head — — 0.202** Region fixed effect Yes Yes Yes Observations 1,974 1,974 1,974 Source: Authors’ calculation. Note: The coefficients were estimated from a logistic regression. Significance level: 1 percent (***), 5 percent (**), and 10 percent (*).  TABLE F19    Living conditions of groups in Sudan Dependent variable: household group (Group 2 corresponds to the reference group) Independent variables (1) (2) (3) (4) (5) Household size 0.023 0.015 0.017 0.019 0.017 Overcrowded dwelling −0.342** −0.293* −0.310** −0.304* −0.326** Share of children in HH −0.001 −0.001 −0.001 −0.001 −0.001 Household headed by a woman — −0.237** −0.280*** — −0.275*** Literate household head — 0.161* 0.172* 0.181** 0.195** Improved water source pre-conflict — — 0.141 0.051 0.034 Improved water source — — −0.723*** −0.683*** −0.659*** Improved sanitation pre-conflict — — — 0.326** 0.281** Improved sanitation — — — 0.130 0.157 Other controls No No No No Yes Region fixed effect Yes Yes Yes Yes Yes Observations 1,974 1,974 1,974 1,974 1,974 Source: Authors’ calculation. Note: The coefficients were estimated from a logistic regression. Significance level: 1 percent (***), 5 percent (**), and 10 percent (*).   FIGURE F17    Information required to decide whether to stay or move in Sudan 70 60 % of households 50 40 30 20 10 0 Have all the info Source of info: radio/TV/Internet Group 1 Group 2 Source: Authors’ calculation. Volume B: Country Case Studies  | 263 Appendix G. Targeting: Household Classification Definitions Effective policy efforts should consider the different needs among displaced households. A precondition for evidence-based planning are data to understand the links between socioeconomic dimensions and the displacement situation of populations to address vulnerabilities and support their economic self-reliance. Identifying groups with different needs allows targeting policy responses that reflect their specific circumstances. Targeting analysis classifies households into three groups based on their poverty status and their ability to participate in income-generating activities. The poverty status reflects the welfare conditions of households, yet additional circumstances might prevent them from participating in income-generating activities. The classification of vulnerable households considers chronic illnesses and disability of working-age members and household heads, alongside the poverty status of the household. For the poverty status, the consumption aggregate used excluded free food or aid received since the aim is to assess the vulnerability of the household before receiving assistance from devel- opment partners, NGOs, or the government. Support-dependent households are the most vulnerable population. The first and most vulnerable group of house- holds identified corresponds to those that are at a disadvantage from participating in income-generating opportuni- ties. This group includes households that either have no working-age adults without disabilities, or are women-headed households with only the household head being a working-age adult without disabilities. Support-dependent house- holds face more challenges to raise their living standards and are more likely to require assistance. Programs whose main component is providing aid should target this group of households. In addition, this group would benefit from gender-responsive programs that address vulnerabilities related to domestic work and caring labor, in addition to GBV and discrimination. Productive but poor households are those that can participate in income-generating activities, yet they con- sume below the standard international poverty line. Skills and human capital are relevant due to the loss of physical capital during displacement and since economic inactivity makes it harder for household members to find employment. Households classified in this second group are those who are poor but are not support-dependent. That is, they have working-age members that can work, but they consume below the international poverty line. Policy efforts should aim to upgrade the skills of members from these households, as well as support their participation in income-generating activities to raise their living standards. Self-reliant households can participate in income-generating activities and are not poor. The last group iden- tifies households that have working-age members who can participate in income-generating activities, and whose consumption level is above the poverty line. Self-reliant households require other interventions to sustain their living conditions, build resilience, and avoid falling into poverty should they experience an unexpected decrease in con- sumption levels. 264  |  Informing Durable Solutions for Internal Displacement Appendix H. Glossary of Key Analysis Indicators Poverty: Consuming less than US$1.90 PPP (2011) per day per person. Poverty gap: The shortfall of actual consumption compared to the poverty threshold of US$1.90 PPP (2011) per day per person. This statistic is calculated for populations that are poor (living below the poverty line). Push factors: Factors about the current location that inform IDPs’ return intention. Pull factors: Characteristics about the potential different location (this can be the origin, or a new location, but not the current location) that informs IDPs’ return intention. Overcrowding: Overcrowding is defined as having four or more persons per room.288 Improved housing: Improved housing is defined as a structure that is made of wood, concrete, or block and is intended for habitation. Improved sanitation: Facilities that are not shared and that hygienically separate human excreta from human contact are con- sidered ‘improved’ while others are ‘unimproved’.289 Improved water source: The type of water source or technology specified by the household used as an indicator for whether the drinking water is of suitable quality. The water sources likely to be of suitable quality are ‘improved’. ‘Improved’ water sources include a piped water supply into the dwelling, piped water to a yard/plot, a public tap/standpipe, a tube well/borehole, a pro- tected dug well, a protected spring, and rainwater. Other sources are generally ‘unimproved’. Literacy: Literacy is the ability to read and write a simple sentence about everyday life. Livestock units: Livestock units are used for aggregating the numbers of different categories of livestock for regional and global comparisons and are obtained by converting the body weight into the metabolic weight. The livestock unit coefficients used are those corresponding to the regions of Near East North Africa: cattle—0.70, buffalo—0.70, sheep—0.10, goats—0.10, pigs—0.20, asses—0.50, horses—0.40, mules—0.60, camels—0.75, chickens—0.01.290 288. UN-Habitat. 2017. “Sustainable Cities and Communities. SDG Goal 11. Monitoring Framework. A Guide to Assist National and Local Governments to Monitor and Report on SDG Goal 11 Indicators.” 289. World Health Organization and UNICEF. 2006. “Core Questions on Drinking Water and Sanitation for Household Survey.” 290. Chilonda and Otte. 2006. “Indicators to Monitor Trends in Livestock Production at National, Regional and International Levels.” Volume B: Country Case Studies  | 265 Labor force activity: Labor force status comprises three mutually exclusive and exhaustive categories for the population in working age. All persons ages 15–64 are defined as being of working age. The three types of labor force status are defined as follows: Employment: A person is employed if he/she is of working age and has engaged, over the previous seven days, in one of the 1.  following work activities: a.  Working as an apprentice b.  Working on the household’s farm, raising livestock, hunting, or fishing c.  Conducting paid or commissioned work d.  Running a business of any size for oneself or for the household e.  Helping in a household business of any size The definition further includes persons who are temporarily absent from their work due to training or working time arrange- ments such as overtime leave and paid interns. Note that the definition excludes household work. 2.  Unemployment: A person is unemployed if he/she is of working age, not in employment during the short reference period, and has been seeking employment within the past four weeks. 3. Outside the labor force or inactivity: A person is outside the labor force (or ‘inactive’) if he/she is of working age and neither employed nor unemployed, according to the preceding definitions. An inactive person is not necessarily idle, especially in the context of a developing economy. The data break this group down into those who are inactive because they do household work, those who are enrolled in education, those who are discouraged, and so on. Labor force: Refers to the sum of persons in employment and in unemployment. It is the counterpart of the group of inactive persons, that is, the labor force plus the inactive sum up to the entire working-age population.291 Working-age population (15–64 years old) Labor force/active Outside of the labor force/inactive In employment In unemployment Enrolled in education Not enrolled in education REFERENCE 291. Bourmpoula, V., S. Kapsos, and J. M. Pasteels. 2013. ILO Estimates and Projections of the Economically Active Population: 1990–2030 (2013 edition). Geneva: ILO (International Labour Organisation). References ACLED (Armed Conflict Location and Event Data). 2016. “ACLED-Africa, v.7,” http://www.acleddata.com. Africa Program at the Council on Foreign Relations. n.d. “Nigeria Security Tracker.” Agbiboa, Daniel Egiegba. 2013. “Why Boko Haram Exists: The Relative Deprivation Perspective.” African Conflict and Peace Building Review 3, no. 1: 144–57. Ali, Osman Mohamed Osman, and Ali Mohamed Mahmoud. 2016. “From a Temporary Emergency Shelter to an Urban- ized Neighborhood: The Abu Shoak IDP Camp in North Dārfūr.” Sudan Working Papers. Bergen: Chr. Michelsen Institute. Alix-Garcia, Jennifer, and David Saah. 2010. “The Effect of Refugee Inflows on Host Communities: Evidence from Tanza- nia.” World Bank Economic Review 24, no. 1: 148–70. Arensen, Michael. 2015. “Historical Grievances and Fragile Agreements: An Analysis of Local Conflict Dynamics in Akobo.” The South Sudan Humanitarian Project. https://publications.iom.int/system/files/pdf/if_we_leave_0.pdf. . 2016. “If We Leave We Are Killed—Lessons Learned from South Sudan Protection of Civilian Sites (2013–2016).” IOM—South Sudan. Armed Conflict Location & Event Data Project. 2017. “Conflict Summaries of Hotspots of Political Violence.” https://www .crisis.acleddata.com/category/somalia/. Bohnet, Heidrun. 2016. “Back to Turmoil: Refugee and IDP Return to and within South Sudan.” BICC Working Papers 7: 38. Buchanan-Smith, Margie, and Abduljabbar Abdulla Fadul. 2008. “Adaptation and Devastation: The Impact of the Conflict on Trade And Markets in Darfur.” Medford, MA: Feinstein International Center. Carter, B., and B. Rohwerder. 2016. “Rapid Fragility and Migration Assessment for Ethiopia.” Central Bureau of Statistics (CBS), and UNICEF Sudan. 2014. “Multiple Indicator Cluster Survey 2014 of Sudan, Final Report.” Khartoum, Sudan: UNICEF and Central Bureau of Statistics (CBS), December. Centre for Public Impact. 2017. “Universal Basic Education in Nigeria.” Chilonda, P., and J. Otte. 2006. “Indicators to Monitor Trends in Livestock Production at National, Regional and Interna- tional Levels.” Livestock Research for Rural Development 18, no. 8. http://www.lrrd.org/lrrd18/8/chil18117.htm. “Constitution of the Federal Republic of Nigeria.” 1999. www.nigeria-law.org/ConstitutionOfTheFederalRepublicOf Nigeria.htm. Council, U.E.a.S. 1998. “Guiding Principles on Internal Displacement.” 266 Volume B: Country Case Studies  | 267 De Vries, Lotje, and Mareike Schomerus. 2017. “South Sudan’s Civil War Will Not End with a Peace Deal.” Peace Review 29, no. 3: 333–40. https://doi.org/10.1080/10402659.2017.1344533. de Waal, Alex. 2005. “Who Are the Darfurians? Arab and African Identities, Violence and External Engagement.” African Affairs 104, no. 415: 181–205. DFID and UN-Habitat. 2009. “Darfur: Profile of El Fasher Town and Abu Shouck IPD Camp.” Dupuy, Kendra, Gates Scott, Mokleiv Nygård Håvard, Rudolfsen Ida, Aas Rustad Siri, Strand Håvard, and Urdal Henrik. 2017. “Trends in Armed Conflict, 1946–2016.” Conflict Trends, 2. Oslo: PRIO. Famine Early Warning Systems Network. July 2018. “SUDAN Price Bulletin July 2018,” https://reliefweb.int/report/sudan/ sudan-price-bulletin-july-2018. FAO. 2017. “South Sudan: Situation Report,” December 4. http://www.fao.org/fileadmin/user_upload/emergencies/ docs/FAO_SouthSudan_SitRep4December2017.pdf. Federal Government of Somalia, World Bank, United Nations, and European Union. 2018. “Somalia Drought Impact and Needs Assessment.” Ferris, Elizabeth. n.d. “Internal Displacement in Africa: An Overview of Trends and Opportunities.” Ethiopian Community Development Council. Feyissa, D. 2014. “The Spillover Effect of South Sudan in Gambella.” http://life-peace.org/hab/the-spillover- effect-of-the-south-sudan-conflict-in-gambella/. Flint, Julie, and Alex de Waal. 2008. Darfur: A New History of a Long War. African Arguments. London: Zed Books. Gemechu, Eyob Asfaw. 2016. “The Recent Attack in Gambela and Its Implications for Humanitarian Operations.” Horn of Africa Bulletin 28, no. 2. Girma, E. 2016. “Assessing the Impact of South Sudanese Refugees on the Host Communities of Itang Woreda.” Addis Ababa University. Global Digital Portal. “Comprehensive Refugee Response Framework (CRRF).” Accessed May 2, 2018. http://www.global crrf.org/. GoS. 2013. “Developing Darfur: A Recovery and Reconstruction Strategy.” Hawkins, Robert L., and Katherine Maurer. 2010. “Bonding, Bridging and Linking: How Social Capital Operated in New Orleans Following Hurricane Katrina.” The British Journal of Social Work 40, no. 6 (September 1): 1777–93. https://doi .org/10.1093/bjsw/bcp087. Human Rights Watch. 2015. “‘Men with No Mercy’ Rapid Support Forces Attacks against Civilians in Darfur, Sudan.” New York: Human Rights Watch. IASC (Inter-Agency Standing Committee). 2010. “IASC Framework on Durable Solutions for Internally Displaced Per- sons.” The Brookings Institution-University of Bern Project on Internal Displacement. http://www.unocha.org/sites/ unocha/files/dms/Documents/IASC%20Framework%20DS%20for%20IDPs.pdf. 268  |  Informing Durable Solutions for Internal Displacement ICRC. 2017. “Translating the Kampala Convention into Practice: A Stocktaking Exercise.” International Review of the Red Cross, Reports and Documents, 99, no. 1: 365–420. IDMC. 2017. “Africa Report on Internal Displacement.” Geneva: Internal Displacement Monitoring Centre (IDMC). . 2018. “Global Report on Internal Displacement.” IDMC and NRC. 2018. “Global Report on Internal Displacement 2018.” Geneva: Internal Displacement Monitoring Centre (IDMC) and Norwegian Refugee Council (NRC). Internal Displacement Monitoring Center. 2018. “UnSettlement: Urban Displacement in the 21st Century; City of Flight,” November. “Internal Displacement Profiling in Mogadishu.” 2016. International Crisis Group. 2013. “Ethiopia: Prospects for Peace in Ogaden.” Africa Report. Nairobi/Brussels. International Organization for Migration. 2016. “Displacement Tracking Matrix Sudan: Dashboard January–December 2016.” . 2019. “Displacement Tracking Matrix, South Sudan Mobility Tracking Round 4,” March. IOM (International Organization for Migration). 2014. “IOM Sudan: Annual Report 2014.” IOM. . 2015a. “North Darfur—Al Salam IDP Camp Profile 2015.” Khartoum. https://displacement.iom.int/reports/ sudan-%E2%80%94-site-profile-%E2%80%94-al-salam-idp-camp-july-2016. . 2015b. “North Darfur State—Abu Shouk IDP Camp.” . 2016a. “Abu Shouk IDP Camp Celebrates the Official Opening of Its First Women’s Centre.” Abu Shouk IDP camp celebrates the official opening of its first Women’s Centre | IOM Sudan Mission. https://sudan.iom.int/news/ abu-shouk-idp-camp-celebrates-official-opening-its-first-women%E2%80%99s-centre. . 2016b. “Sudan Livelihoods Factsheet.” IOM, August. . 2017. “Nigeria Displacement Tracking Matrix, (DTM) Round XX, December 2017.” . 2018. “Nigeria Displacement Tracking Matrix, (DTM) Round 23, June 2018.” IOM (International Organization for Migration)/Camp Coordination and Camp Management. 2015. “South Sudan’s Crisis Response Displacement Tracking Matrix,” February. https://www.humanitarianresponse.info/sites/www .humanitarianresponse.info/files/documents/files/DTM%20Report%20February%202015.pdf. JIPS (Joint IDP Profiling Service). 2018. “Durable Solutions Analysis Guide: A Tool To Measure Progress Toward Durable Solutions to Internal Displacement.” Johnson, D. H. 2014. “Briefing: The Crisis in South Sudan.” African Affairs 113, no. 451: 300–309. . 2003. The Root Causes of Sudan’s Civil Wars. Vol. 5. Justin, Peter Hakim, and Lotje De Vries. 2017. “Governing Unclear Lines: Local Boundaries as a (Re)Source of Conflict in South Sudan.” Journal of Borderlands Studies, March 9, 1–16. https://doi.org/10.1080/08865655.2017.1294497. Volume B: Country Case Studies  | 269 Khan, Aslam, and Lawan Cheri. 2016. “An Examination of Poverty as the Foundation of Crisis in Northern Nigeria.” Insight on Africa 8, no. 1: 59–71. Kindersley, Nicki. 2018. “Politics, Power and Chiefship in Famine and War: A Study of the Former Northern Bahr El-Ghazal State, South Sudan.” South Sudan Customary Authorities. Nairobi: Rift Valley Institute. Mallet, R., J. Hagen-Zanker, N. Majidi, and C. Cummings. 2017. “Journeys on Hold: How Policy Influences the Migration Decisions of Eritreans in Ethiopia.” ODI, no. 506. Accessed August 23, 2018. https://www.odi.org/ publications/10728-journeys-hold-how-policy-influences-migration-decisions-eritreans-ethiopia. Matfess, Hilary. 2017. Women and the War on Boko Haram: Wives, Weapons, Witnesses. Zed Books. Norwegian Refugee Council. 2017. “Protection of Civilian Sites: Lessons from South Sudan for Future Operations.” OCHA (United Nations Office for Coordination of Humanitarian Affairs). 2015. “Nigeria: Northeast Crisis Situation.” . 2017a. “South Sudan: Humanitarian Needs Overview 2018,” November. https://reliefweb.int/sites/reliefweb. int/files/resources/South_Sudan_2018_Humanitarian_Needs_Overview.pdf. . 2017b. “South Sudan: Humanitarian Snapshot,” November. https://reliefweb.int/sites/reliefweb.int/files/ resources/20171213_SS_Humanitarian_Snapshot_November_final.pdf . 2018. “Northeast Nigeria Humanitarian Situation Update, July 2018 Edition.” Oxfam. 2017. “South Sudan Gender Analysis.” REACH. 2015. “South Sudan Displacement Trends Analysis.” . 2016. “South Sudan Intentions Study,” March. http://www.reachresourcecentre.info/system/files/resource-doc- uments/reach_ssd_report_idp_intentions_march_2016.pdf. Richards, S., and G. Bekele. 2011. “Conflict in the Somali Region of Ethiopia: Can Education Promote Peace-Building?” Feinstein International Center, Tufts University. Sarzin, Zara. 2017a. “Background Paper for the Ethiopia Forced Displacement Strategy Note.” World Bank. . 2017b. “Stocktaking of Global Forced Displacement Data.” Policy Research Working Paper. Washington, DC: World Bank. Sefa-Nyarko, Clement. 2016. “Civil War in South Sudan: Is It a Reflection of Historical Secessionist and Natural Resource Wars in ‘Greater Sudan’?” African Security 9, no. 3 (July 2): 188–210. https://doi.org/10.1080/19392206.2016.1208476. Sharkey, Heather J. 2008. “Arab Identity and Ideology in Sudan: The Politics of Language, Ethnicity, and Race.” African Affairs 107, no. 426: 21–43. https://doi.org/10.1093/afraf/adm068. South Sudan Law Society, and United Nations Development Program. 2016. “Search for a New Beginning: Perceptions of Truth, Justice, Reconciliation and Healing in South Sudan.” http://www.ss.undp.org/content/dam/southsudan/ library/Rule%20of%20Law/Perception%20Survey%20Report%20Transitional%20Justice%20Reconciliation%20 and%20Healing%20-.pdf. SUDO UK. 2012. “The Humanitarian Situation in Abu Shouk Camp, Al-Fasher, North Darfur.” London. 270  |  Informing Durable Solutions for Internal Displacement Tearfund. 2007. “Darfur: Water Supply in a Vulnerable Environment: Phase Two of Tearfund’s Darfur Environment Study.” London: Tearfund. http://www.tearfund.org/darfurwatervulnerability. Tesfaye, B. 2016. “The Point of No Return in Ethiopia.” Africa Is a Country. https://africasacountry.com/2016/10/ the-point-of-no-return-in-ethiopia. “The Impact of War on Somali Men.” 2015. http://www.logica-wb.org/PDFs/LOGICA_The_Impact_of_War_on_Somali_ Men.pdf. Udo-Udo, Jacob, Maire Abia-Bassey, Emily Nkanga, and Abdulahi Aliyu. 2016. “Narratives of Displacement: Conversa- tions with Boko Haram Displaced Persons in Northeast Nigeria.” Contemporary French and Francophone Studies 20, no. 2: 176–90. UN General Assembly. 2016. “New York Declaration for Refugees and Migrants.” Resolution A/RES/71/1, Adopted on September 19. New York: UN General Assembly. UN OCHA. 1998. “Guiding Principles on Internal Displacement.” https://drc.ngo/media/1217434/guiding- principles-on-internal-displacement.pdf. . 2016a. “Sudan: Humanitarian Bulletin.” Geneva: UN OCHA, June 20. . 2016b. “Sudan Humanitarian Needs Overview 2017,” December. https://reliefweb.int/sites/reliefweb.int/files/ resources/Sudan_2017_Humanitarian_Needs_Overview.pdf. UN-Habitat. 2016a. “Monitoring Framework, SDG Goal 11.” . 2016b. “Sustainable Cities and Communities. SDG Goal 11. Monitoring Framework. A Guide to Assist National and Local Governments to Monitor and Report on SDG Goal 11 Indicators.” https://unhabitat.org/wp-content/ uploads/2016/02/SDG-Goal%2011%20Monitoring%20Framework%2025-02-16.pdf. UNHCR. 2011. “Convention and Protocol Relating to the Status of Refugees.” http://www.unhcr.org/uk/3b66c2aa10. . 2016. “SUDAN: Refugees, Asylum Seekers, IDPs and Others of Concern to UNHCR by State.” Nairobi: UNHCR, April 30. http://www.refworld.org/country,,UNHCR,,SDN,,573ad3274,0.html. . 2017. “Statistical Yearbook 2016.” UNHCR. http://www.unhcr.org/uk/statistics/country/5a8ee0387/unhcr- statistical-yearbook-2016-16th-edition.html. . 2018a. “UNHCR Ethiopia Fact Sheet" Ethiopia. https://reliefweb.int/sites/reliefweb.int/files/resources/UNHCR_ Ethiopia_May_2018.pdf . 2018b. “UNHCR Global Trends: Forced Displacement in 2017.” Geneva: United Nations High Commissioner for Refugees. http://www.unhcr.org/statistics/unhcrstats/5b27be547/unhcr-global-trends-2017.html. . 2018c. “South Sudan Situation (31 December 2017)”. . 2019. United Nations High Commissioner for Refugees. “Figures at a Glance. Statistical Yearbooks”: https://www .unhcr.org/figures-at-a-glance.html . “South Sudan Situation: Refugee Population in Gambella Region,” n.d. United Nations. 2015. “World Population Prospectus: The 2015 Revision.” Volume B: Country Case Studies  | 271 United Nations, and World Bank. 2018. “Pathways for Peace: Inclusive Approaches to Preventing Violent Conflict.” Wash- ington, DC: World Bank. UNMISS. 2018. “UNMISS PoC Update.” UNOCHA. 2004. “Guiding Principles on Internal Displacement.” https://www.unocha.org/sites/dms/Documents/Guid- ingPrinciplesDispl.pdf. . 2017. “2018 Humanitarian Needs Overview: Somalia.” Verwimp, Philip, and Maystadt Jean-Francois. 2015. “Forced Displacement and Refugees in Sub-Saharan Africa: An Eco- nomic Inquiry.” Background Paper for “Poverty in a Rising Africa.” World Bank. WFP. 2016. “Darfur Food Security Monitoring November 2016.” Sudan: WFP, November. . 2017a. “Darfur Food Security Monitoring May 2017.” Sudan: WFP, May. . 2017b. “Darfur Food Security Monitoring November 2017.” Sudan: WFP, November. Williams, P. 2017. “Continuity and Change in War and Conflict in Africa.” PRISM 6, no. 4: 33–45. World Bank. 2015. “The Sustainable Transformation of Youth in Liberia (STYL) Program.” . 2017a. “Cities of Refuge in the Middle East. Bringing an Urban Lens to the Forced Displacement Challenge.” Washington, DC: World Bank. . 2017b. “South Sudan Poverty Assessment.” . 2017c. “Sudan IDP Profiling Preparation.” Washington, D.C.: The World Bank. . 2019. “Macro Poverty Outlook South Sudan,” April. World Bank Group. 2016. “North-East Nigeria. Recovery and Peace Building Assessment.” World Bank, and UNHCR. 2015. “Forced Displacement and Mixed Migration in the Horn of Africa.” Geneva and Washing- ton, DC: World Bank and UNHCR. World Data Lab. 2018. “World Poverty Clock.” World Food Programme. 2019. “WFP South Sudan Situation Report #242,” March. World Health Organization and UNICEF. 2006. “Core Questions on Drinking Water and Sanitation for Household Survey.” Ylönen, Aleksi. 2009. “On Sources of Political Violence in Africa: The Case of ‘Marginalizing State’ in Sudan.” Política y Cul- tura, no. 32: 37–59. Young, Helen, and Karen Jacobsen. 2013. “No Way Back? Adaptation and Urbanization of IDP Livelihoods in the Darfur Region of Sudan.” Development and Change 44, no. 1: 125–45. https://doi.org/10.1111/dech.12003. Zetter, R., and Raudel, H. 2016. KNOMAD Study. Refugees’ Right to Work and Access to Labor Markets—An Assess- ment. KNOMAD Global Knowledge Partnership on Migration Development. https://www.knomad.org/sites/default/ files/2017-12/KNOMAD%20Study%201-%20Part%20II-%20Refugees%20Right%20to%20Work%20-%20An%20 Assessment.pdf