This report examines the socioeconomic conditions of refugees and host communities in Ethiopia, with the aim of generating high-quality evidence to inform policy and guide development interventions toward greater self-reliance and economic integration. Ethiopia hosts more than 922,000 registered refugees and asylum seekers as of end-2023, primarily from South Sudan (420,000), Somalia (280,000), Eritrea (170,000), and Sudan (49,000). Most refugees (92 percent) live in roughly 30 camps located in mostly rural, remote border regions, including Afar, Amhara, Benishangul-Gumuz, Gambella, Somali, and Tigray, while some 70,000 live in Addis Ababa under an Out-of-Camp Policy (OCP). The displacement situation is overwhelmingly protracted: refugees have been in Ethiopia for an average of about 14 years.
While Ethiopia has adopted progressive refugee policies, including the Comprehensive Refugee Response Framework (CRRF), challenges persist in translating these policies into tangible socioeconomic outcomes for refugees. The country’s new Refugee Proclamation grants refugees the right to basic services, work, and freedom of movement, but implementation delays hinder their realization.
The analysis draws on the Socio-Economic Survey of Refugees in Ethiopia (SESRE), conducted between November 2022 and January 2023 through a collaborative effort among the World Bank, the Ethiopian Statistical Service (ESS), the Refugees and Returnees Service (RRS), and UNHCR. The survey employed a stratified, two-stage cluster sample design covering three distinct population groups: in-camp refugees (stratified by nationality: Eritrean, South Sudanese, and Somali), out-of-camp refugees in Addis Ababa (stratified by arrival time), and host community members defined as non-displaced Ethiopian households living within approximately 5 kilometers of refugee camps. A total of 3,452 households were sampled.
Main findings:
- Refugee households are larger, more likely to be female-headed, and have markedly lower educational attainment than host households. Average household size among in-camp refugees is 6.2 members, compared to 5.2 for hosts. Female-headed households account for 73 percent of in-camp refugee households overall and reach 84 percent among South Sudanese refugees. Educational attainment is low among both hosts and refugees: 73 percent of in-camp adult refugees and 59 percent of host adults have either no education or did not complete primary school. While primary school enrollment rates are broadly similar between refugees and hosts, secondary enrollment for refugees stands at 23 percent compared to 41 percent for hosts.
- Child malnutrition rates among in-camp refugees are high, while refugees in camps have better access to health institutions for child delivery than hosts. Child stunting is a major child health problem for both hosts and refugees, but stunting rates are highest for refugee children: stunting affects 52 percent of Eritrean refugee children and 47 percent of Somali refugee children. At the same time, 87 percent of refugee mothers in camps deliver in health institutions, a higher share than among host mothers (75 percent), reflecting the concentration of primary health services within camps. A significant share of both in-camp refugees and hosts have no birth registration for their children.
- In-camp refugees live in significantly worse housing conditions, though they often have better access to safe drinking water than host communities. Both in camps and in Addis Ababa, refugee households live in more overcrowded conditions compared to hosts. Most in-camp refugees and their respective host households live in homes with low-quality walls, with only 2 percent of in-camp refugee households living in dwellings with an improved wall. Refugees have better access to drinking water and similar access to improved sanitation facilities compared to hosts. Both refugee and host households have low access to electricity, except for hosts of Eritrean refugees.
- Labor force participation among in-camp refugees is substantially lower than among hosts, and those who work earn significantly less per hour. In-camp refugee labor force participation stands at 31 percent, compared to 52 percent for hosts, and only 25 percent of in-camp refugees performed paid work in the week before the survey, compared to 48 percent for hosts. Camp-based refugees who do work earn approximately 57 percent less per hour than their host counterparts, a gap that is not explained by differences in education, demographic characteristics, occupation or sector. Among OCP refugees in Addis Ababa, only 17 percent are employed, with many relying on remittances; those who do work frequently experience occupational downgrading, taking on jobs below their skill level. Refugee inactivity and unemployment are high among youth (aged 15-24) in Addis Ababa.
- A substantial share of refugees holds aspirations of resettlement to Western countries that are inconsistent with actual resettlement rates. Approximately one-third of refugees reports expecting to be resettled to a Western country within three years, despite global resettlement rates of less than 2 percent. Refugee aspirations and expectations for resettlement may be important determinants of how much they will invest in their skills and socio-economic integration.
- In-camp refugees experience deep monetary and multidimensional poverty, while OCP refugees in Addis Ababa have lower monetary poverty rates but face distinct vulnerabilities. The poverty headcount among in-camp refugees is 84 percent, compared to 32 percent for adjacent host communities; across all refugee groups, 75 percent live below the international poverty line, compared to 25 percent among hosts. OCP refugees in Addis Ababa have a monetary poverty rate of 7 percent—lower than their hosts at 18 percent. Fifty percent of refugees are multidimensionally poor, with deprivation driven primarily by low living standards (cooking fuel, housing quality, and assets) and low educational attainment. In-camp refugees also face severe food insecurity, scoring 8.1 out of 10 on the food insecurity scale compared to 4.0 for hosts, and only 49 percent of in-camp refugees have acceptable food consumption status compared to 74 percent of hosts.
- Among in-camp refugees, employment, household size, and proximity to market hubs are among the strongest correlates of welfare, while returns to education are substantially muted compared to host households. Consumption per capita falls sharply as household size and dependency ratios increase, with the poorest households having sizes more than double those of the wealthiest. The share of employed household members is one of the most significant predictors of higher expenditure, and proximity to a woreda capital matters as well—a 1 percent increase in distance to the nearest woreda capital is associated with a 0.09 percent reduction in consumption per capita, and living in a high-market-accessibility area is associated with a 0.24 percent increase compared to low-accessibility areas. Although education is positively associated with welfare, the consumption returns to additional schooling are considerably smaller for in-camp refugees than for OCP refugees or host community members, a pattern the report attributes to the constrained opportunities to apply skills within camp settings.
- Most refugees depend on humanitarian transfers as their primary source of support. While 83 percent of the in-camp refugees depend on transfers and gifts to cover their food consumption needs, more than two-thirds of the host community households depend on market purchases for their food consumption. Similarly, most refugees depend on transfers or gifts for non-food consumption, while their hosts depend on market purchases.
- Proximity to urban centers and local market connectivity are the dominant local determinants of refugee employment, with distance to the nearest zone capital sharply reducing the probability of finding work. Refugees living within 10 kilometers of a zone capital have a 59 percent employment probability (male) and 47 percent (female), compared to 34 percent and 23 percent respectively at 100 kilometers. Poor market accessibility compounds this effect: employment probability falls below 25 percent in woredas more than 1.5 standard deviations below average market access. As remoteness increases, refugees shift toward agricultural work, while those near urban hubs concentrate in trade and services. High local unemployment among hosts further reduces refugee employment prospects, while areas with greater host labor force participation are associated with better outcomes for refugees.
- Host community members display broadly positive attitudes toward refugees, though actual social integration between the two groups remains limited. Sixty-five percent of host community respondents agree that refugees are “good people,” and most report low levels of perceived conflict. However, few refugees report having Ethiopian friends, suggesting that formal tolerance has not translated into meaningful social interaction or integration.
The report concludes that the current camp-based humanitarian model limits refugee self-reliance and generates substantial long-term costs for both the humanitarian system and Ethiopia’s development trajectory. The authors highlight what they term an “economic-inclusion dividend”: transitioning from a model of full aid dependency to one of full economic inclusion could reduce the annual cost of meeting basic needs per refugee from US$378 to US$78.
The report highlights several policy implications. First, enabling refugee mobility and easing work permit restrictions—including through automation of permit issuance—are identified as critical steps toward improving labor market access, given that the wage gap between refugees and hosts is not explained by observable human capital differences. Second, the report calls for integrating refugees into national systems for education, health (including the Community-Based Health Insurance scheme), and social protection, rather than maintaining parallel humanitarian service delivery. Third, the report recommends place-based investment in refugee-hosting woredas to benefit both refugees and host communities simultaneously, reducing resource competition and addressing the spatial disparities in opportunity documented in the analysis. Finally, the report highlights the need to redesign the Out-of-Camp Policy so that mobility is guided by economic opportunity rather than individual sponsorship arrangements, and to harmonize national and sub-national laws to support full implementation of Ethiopia’s 2019 Refugee Proclamation.