Policy Research Working Paper 11108 Host Environments and Welfare of In-Camp Refugees Evidence from Ethiopia Takaaki Masaki Nitsuh Mengist Nega Christina Wieser Poverty and Equity Global Department April 2025 Policy Research Working Paper 11108 Abstract This paper examines welfare disparities between in-camp that more favorable socioeconomic environments benefit refugees and host com- munities in Ethiopia, using data refugees, even when labor market outcomes are constrained from the 2023 Socio-Economic Survey of Refugees in by legal and structural barriers. Furthermore, the paper Ethiopia. The analysis reveals significant welfare gaps explores how employment outside camps may be linked between in-camp refugees and hosts even after accounting to the welfare of in-camp refugees, as well as the reasons for various other baseline socioeconomic char- acteristics: behind their pursuit of such opportunities despite restric- in-camp refugees exhibit 60 percent lower consumption tive policies. It finds that refugees working outside camps per capita and poverty rates that are 40 percentage points often originate from poorer households but report a greater higher compared to host communities. These disparities sense of autonomy and control over their lives. These find- are also associated with limited employment opportunities ings highlight the importance of economic inclusion as a and dependence on humanitarian aid among in-camp ref- pathway to self-reliance for in-camp refugees and emphasize ugees. The study examines the relationship between host the need to remove barriers to formal employment oppor- community characteristics and refugee welfare, revealing tunities outside camps to improve their welfare. This paper is a product of the Poverty and Equity Global Department. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at cwieser@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Host Environments and Welfare of In-camp Refugees: Evidence from Ethiopia∗ Takaaki Masaki† Nitsuh Mengist Nega ‡ Christina Wieser § Keywords: Poverty, labor outcomes, refugees, forced displacement, Ethiopia JEL Classification Codes: D6, F22, N37 ∗ The authors wish to thank Rinku Murgai, Aissatou Dicko and Maja Lazi´ c for their comments and guidance. This work is funded by the World Bank - UNHCR Joint Data Center on Forced Displacement and its partners: the Government of Denmark represented by the Danish Ministry of Foreign Affairs; the European Union represented by the Directorate-General for International Partnerships (INTPA) at the European Commission; and the United States Government represented by the U.S. Bureau of Population, Refugees, and Migration (PRM). The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the World Bank and its affiliated organizations, nor do they represent the views of the Executive Directors of the World Bank and the governments they represent, or the views of UNHCR. † Asian Development Bank; tmasaki@adb.org ‡ World Bank; nnega@worldbank.org § World Bank; cwieser@worldbank.org 2 1 Introduction The protracted nature of refugee situations has prompted a paradigm shift in interna- tional assistance, moving from a purely humanitarian response to a more comprehensive development-oriented approach. This transition recognizes that long-term displacement cre- ates complex socio-economic challenges for refugees. One of the key metrics to measure progress towards the self-reliance of refugees includes welfare and poverty (GCR 2023).1 That said, refugees are often excluded from a national survey that collects data on welfare and poverty and this exclusion has hitherto limited our understanding of the socio-economic status of refugees (Denaro and Giuffr´e, 2021). This data landscape, however, has been chang- ing in recent years – albeit gradually – with more countries now willing to collect detailed socio-economic data on displaced populations (Masaki and Madson, 2023). Using the 2023 Socio-Economic Survey of Refugees in Ethiopia (SESRE), which offers extensive data on both refugees and host communities, this paper presents new evidence on the stark welfare disparities between in-camp refugees and hosts. One of the unique features of this survey is that SESRE has a survey strata for three groups: (i) refugees in camps; (ii) refugees outside camps in Addis Ababa; and (iii) host communities2 ; all of which require a distinct sampling procedure. The richness of this dataset allows for a detailed comparative analysis of welfare and poverty outcomes and other key socio-economic characteristics like labor outcomes across these groups. SESRE also allows us to delve into the disparities in welfare among in-camp refugees. Unlike other refugee surveys conducted elsewhere that sampled only from few refugee settlements in a given country, SESRE covered 24 refugee camps in Ethiopia. This provides us with rich spatial variation in understanding the interplay between various host environments and the welfare of refugees. Our findings indicate that refugees residing in camps are significantly poorer than host communities, even after controlling for other socio-economic differences and location/camp fixed effects. Welfare for in-camp refugees, measured by total consumption per capita, is 60 percent lower than that of hosts, while the poverty rate among in-camp refugees is roughly 40 percentage points higher. In-camp refugees largely depend on aid as their main source of income, and our analysis shows that this aid is insufficient to address the high levels of poverty 1 In 2018, under the IAEDG-SDGs initiative focused on data breakdown, the UNHCR and the Joint IDP Profiling Service (JIPS), as contributors to the Expert Group on Refugee and IDP Statistics (EGRIS), presented the International Recommendations on Refugee Statistics (2018). These recommendations aimed to standardize and enhance the quality of data on populations affected by forced displacement. Building on this, in December 2020, the UNHCR, JIPS, and STATS4SD pinpointed 12 key SDG indicators for which they recommended detailed analysis by displacement status, including the poverty status of refugees. 2 Host communities are defined as Ethiopian non-displaced households living in enumeration areas adja- cent to the refugee camps. 3 observed in camps. Poor labor outcomes further exacerbate poverty within camps, with in- camp refugees showing lower labor force participation and employment rates compared to their host communities. We also show that favorable host environments contribute to the welfare of in-camp refugees. When host environments have a higher level of welfare, low poverty, and/or better access to services, all these favorable conditions spill over to benefit the welfare of in-camp refugees. That said, favorable host environments have not yet been translated into improved labor outcomes for refugees due to restrictions on their rights to seek job opportunities outside camps. Despite not having formal permits to work outside camps at the time of SESRE imple- mentation, some still sought employment outside the camps, due to the need to improve their living standards and gain a greater sense of control or autonomy over their own fate. Prior studies have shown that a low sense of control over one’s own fate – which is prevalent among in-camp refugees who are highly dependent on international assistance – is associ- ated with poorer mental health or economic outcomes (Wieser et al., 2023; Hahn et al., 2019; Thum, 2014; Tsionis et al., 2024; Hussam et al., 2022). In Jordan, for instance, some refugees choose to live outside camps in search of greater autonomy and freedom (Hoogeveen and Obi, 2024). Our analysis shows that working outside camps is positively associated with in-camp refugees’ perceived control over their own lives and fate. We also show that the jobs in-camp refugees find outside the camps are not necessarily “better” than those available within the camps. In fact, in-camp refugees working outside the camps are less likely to secure off-farm or formal jobs compared to those working within the camps. Since formal jobs are often provided by humanitarian organizations operating inside the camps, the lack of such opportunities in the surrounding host communities further limits pathways to formal employment for in-camp refugees and perpetuates their dependence on humanitarian assistance. This paper makes three important contributions. First, there is a notable scarcity of studies that compare welfare outcomes between refugees and hosts. This is mainly driven by the fact that comparable welfare statistics for hosts and refugees are largely absent though there are a few exceptions such as Chad (Nguyen et al., 2021), Ethiopia (Pape et al., 2021; Wieser et al., 2023) and Uganda (World Bank, 2019). Our study draws from Wieser et al.’s comprehensive socio-economic analysis of refugees in Ethiopia (2024) and extends it by delving deeper into the sources of welfare disparities between in-camp refugees and hosts through an econometric analysis. Identifying factors that drive such disparities is crucial for designing effective policies that help enhance the self-reliance of refugees. Second, our study reveals a connection between the characteristics of host communities 4 and the welfare of refugees, showing that favorable host environments (e.g., socio-economic conditions, host attitudes) can enhance refugee welfare. The existing literature has dispro- portionately focused on the impact of refugees on various socio-economic outcomes of host communities, including poverty (Ayenew, 2021), jobs (Walelign et al., 2022), commercial activity and trade (Vemuru et al., 2020), inflation (Alam et al., 2022), labor force participa- tion (Ceritoglu et al., 2017) and host community attitudes (Betts et al., 2023).3 Much less attention has been paid to the impact of host communities on the welfare of refugees except for a few studies that looked at how host environments may play a role in shaping “refugee economies” (Alloush et al., 2017; Betts et al., 2024). This study showcases how the welfare of refugees is interlinked with that of host com- munities where favorable host environments contribute to the welfare of in-camp refugees. Understanding the linkage between the welfare of refugees and their host environment is crucial for designing effective policies that foster both refugee integration and sustainable development in host communities. Refugees do not live in isolation; their well-being is deeply intertwined with the economic, social, and political conditions of their surroundings. A thriving host environment provides refugees with better access to employment, education, healthcare, and social networks, which in turn enhances their self-reliance and reduces de- pendency on aid. Conversely, when host communities struggle with economic stagnation or resource constraints, tensions may arise, exacerbating vulnerabilities for both refugees and local populations. By recognizing these interdependencies, policy makers, humanitarian organizations, and development agencies can implement strategies that not only improve refugee welfare, but also contribute to the resilience and prosperity of host societies. Third, it seeks to shed light on the question of why some refugees opt to seek employ- ment opportunities outside camps despite restrictive regulatory environments. It is well documented both qualitatively and quantitatively that even under restrictive host environ- ments where no work right for refugees is granted, refugees still seek “informal pathways” to find employment outside camps (Zetter and Ruaudel, 2018; Neikirk and Nickson, 2023). Understanding the factors that provide incentives for refugees to seek employment oppor- tunities outside camps is crucial for several reasons. It helps policymakers design more effective strategies to support the economic integration of refugees, allowing them to become self-reliant and less dependent on humanitarian aid. Employment outside camps can also foster social cohesion by enabling refugees to contribute to local economies and interact more with host communities (Zetter and Ruaudel, 2016; Crawford et al., 2022), reducing tensions and promoting mutual understanding (Jacobsen, 2002). 3 See Verme and Schuettler (2021) for a thorough review on the impact of forced displacement on host communities. 5 2 Background Conflict, political unrest, environmental disruption, and economic instability have led to the forced displacement of millions around the world (Ferris, 2010; Black, 2001). The count of forcibly displaced individuals has risen steadily over the past decade. As of mid-2023, there were 36.4 million refugees globally (UNHCR, 2023b). Although significant progress has been made in making a dent in global poverty, extreme poverty is increasingly found among vulnerable groups, including refugees (World Bank, 2017). The persistently poor socio-economic conditions of the forcibly displaced presents significant hurdles to global efforts aimed at eradicating extreme poverty. Ethiopia has a long history of hosting refugees and remains the host of one of the largest refugee populations in Africa. The issue of forced displacement in the country is a conse- quence of conflict, drought, floods, economic instability, and political unrest in neighboring nations (Martin, 2010; UNHCR, 2020; IPCC, 2019). By the end of 2023, Ethiopia was host- ing over 922,000 refugees and asylum seekers, mainly from South Sudan (420,000), Somalia (280,000), Eritrea (170,000), and Sudan (49,000). The majority of refugees (92 percent) are housed in around 30 camps and sites across the Afar, Amhara, Benishangul-Gumuz, Gambella, Somali, and Tigray regions, with a notable number residing in the capital, Addis Ababa (70,000) (UNHCR, 2023a). The refugee camps are varied in location, with ethnic and linguistic connections to the host communities, and are spread across different ecological zones. For instance, about 38 percent of refugees are in arid, lowland pastoralist areas, while 60 percent are in humid, agriculturally viable lowland regions (Wieser et al., 2023). Ethiopia has made considerable strides in developing a more advanced and inclusive approach to refugee response. The country has ratified the 1951 UN Convention on the Status of Refugees and its 1967 Protocol, thereby committing to the protection of refugees and asylum seekers. The Ethiopian government has notably reformed its refugee policies since 2016, moving away from a policy of encampment to one that emphasizes socio-economic integration. A series of national laws and policies have been implemented to safeguard refugees and uphold their rights. The cornerstone of these legal measures is the Refugee Proclamation, established in 2004, which outlines the principles for refugee protection and administration in the country. In 2017, Ethiopia distinguished itself by being the first nation to implement the Comprehensive Refugee Response Framework (CRRF), an international blueprint for addressing refugee crises. Following its support for the Global Compact on Refugees (GCR) in 2018, Ethiopia further reinforced its dedication to refugee welfare with the introduction of a revised and forward-thinking Refugee Proclamation in January 2019. 6 The Ethiopian government introduced the Out-of-Camp Policy (OCP) in 2010, allowing refugees to reside in Addis Ababa and other non-camp areas, expanding opportunities for self-reliance. In 2019, a Directive enabled refugees to apply for OCP residency permits after one month in a camp, provided they can support themselves or have a sponsor and receive a work permit. Special exemptions are also available for vulnerable groups. After the conflict in Tigray in November 2020, many Eritrean refugees relocated to Addis Ababa, with differences in demographics and health outcomes observed between those who arrived before and after the conflict. Most recently, Ethiopia introduced Directive No. 1019/2024,4 based on the 2019 Refugee Proclamation, to grant recognized refugees and asylum seekers the right to work in wage employment, commercial activities, agriculture, manufacturing, and services across urban and rural areas. The Directive aims to foster self-reliance and integration, helping refugees become productive members of Ethiopian society. Despite these progressive legal and policy measures, many refugees in Ethiopia continue to experience poverty and remain heavily reliant on humanitarian aid (Wieser et al., 2024). Ethiopia’s policy measures have not effectively materialized into concrete socio-economic benefits for the refugee population. Predominantly, refugees are confined to camps, with a minimal fraction gaining from the progressive policy environment. Formal employment opportunities are limited, and only a few refugees in camps manage to find stable work. For instance, South Sudanese refugees often engage in collecting grass for sale, brewing local alcohol, or selling goods from their rations (Vemuru et al., 2020). Somalia refugees, in addition to employment through humanitarian NGOs, international organizations, or Ethiopia’s Refugees and Returnees Service (RRS, formerly Agency for Refugee and Returnee Affairs), are typically involved in agriculture, livestock rearing, and small-scale commerce (Betts et al., 2019). Eritrean refugees commonly take up casual construction work or self- employment in small businesses, while Sudanese refugees participate in artisanal gold mining and daily farm labor (Vemuru et al., 2020). The sluggish enactment of these policies impedes refugees’ ability to move in search of improved economic opportunities or to secure access to land, financial services, or the necessary work permits to engage in employment outside the camp settings. 4 Ethiopia: Directive No. 1019/2024 to Implement Recognized Refugees’ and Asylum Seekers’ Right to Work, No. 1019/2024, 14 August 2024 (https://www.refworld.org/legal/decreees/natlegbod/2024/ en/148630). 7 3 Data Our analysis draws on data from SESRE, conducted between November 2022 and Jan- uary 2023. SESRE aimed to address two key issues: (i) a lack of data on the socio-economic conditions of refugees, and (ii) a shortage of analytical studies comparing the socio-economic outcomes of refugees and host communities. The absence of up-to-date information hin- ders effective policies for addressing various socio-economic challenges faced by both groups. SESRE helps assess social dynamics, socio-economic interactions, and social inclusion, pro- viding essential insights for governments and development partners to improve refugee inte- gration and the welfare of both refugees and their host communities. SESRE focuses on three distinct groups, each requiring a unique sampling strategy: refugees in camps, refugees out-of-camp, and host communities. • Refugees in Camps: Using UNHCR’s proGRES database, refugee camps were divided into three domains based on refugees’ country of origin (South Sudan, Somalia, Er- itrea). Camps were broken into enumeration areas (EAs) of 150-200 households each for sampling. • Refugees in Addis Ababa: Due to the challenges of locating refugees in urban areas, the sampling used UNHCR’s data, sorted by location. Pseudo-EAs were created based on Woreda and sub-city boundaries. Refugees were stratified based on whether they arrived before or after the Tigray conflict in November 2020. • Host Communities: Host communities were defined as those living within 5 km of a refugee camp. Sampling used Ethiopia’s 2018 cartographic database, with EAs in both rural and urban areas as the primary sampling unit. This approach ensured that data collection captured the diverse experiences of refugees and their host communities. The final sample consists of 1,296 in-camp refugee households, 431 out-of-camp refugee households and 1,725 host households. Since our main focus is on understanding the factors associated with the welfare of in-camp refugees vis-` a-vis hosts, out-of-camp refugees in Addis Ababa are excluded from the analysis. SESRE encompasses all major refugee camps, including those for Eritreans, South Su- danese, and Somalis, as well as out-of-camp refugees in Addis Ababa and the corresponding host communities. Eritrean refugees in Tigray were excluded due to the conflict between 2020 and 2022, but those in Afar and those who moved to the newly established Alemwach refugee site were included. The survey also sampled Eritrean refugees who moved to Alemwach after the conflict and those who arrived in Addis Ababa post-November 2020. 8 4 Models We first assess how welfare may vary depending on the displacement status of households and individuals. To do so, we estimate the following fixed-effects model: yic = αc + γ Refugeeic + β Xic + ϵic (1) where yic is the dependent variable representing welfare outcomes (e.g., consumption per capita (log), extreme poverty) for household i in camp c (or in the host community near camp c). αc represents the location-specific fixed effect, Refugeeic is a binary variable indicating whether the household is an in-camp refugee, Xic is a vector of household-level covariates (such as household size, age, literacy, and access to services), and ϵic is the error term. Accounting for camp fixed effects is crucial when comparing refugees and host communi- ties because it controls for location-specific characteristics that could influence both welfare outcomes and other underlying socio-economic characteristics of refugees and hosts. Each camp or host area might have unique factors, such as access to resources, infrastructure, or local policies, that affect both refugees and hosts differently. By incorporating fixed effects, the analysis isolates the impact of refugee status on welfare, ensuring that comparisons be- tween refugees and hosts are not skewed by variations in conditions across different camps or regions. Applying this same model, we also test how refugee status may be correlated with other key “intermediate” outcome variables that are also closely linked to welfare. Namely, we evaluate how refugee status may be associated with labor outcomes (e.g., labor force partic- ipation, employment) as well as main sources of income. In so doing, we seek to shed light on key factors that explain variation in welfare outcomes between refugees and hosts. We further investigate the factors that drive variations in welfare specifically among in- camp refugees. Previous research, both qualitative and quantitative, has highlighted how household characteristics – including household size, education level, access to services – may affect welfare disparities among in-camp refugees. In addition, host environments may also play a significant role. To pinpoint the key household and host environmental factors influencing welfare disparities among refugees, we estimate the following models: yic = α + δ Host Characteristicsc + β Xic + ϵic (2) where yic is the dependent variable for refugee household i in camp c, Host Characteristicsc represents the socio-economic characteristics of host communities in and around camp c. We consider three key host characteristics: average values of yic among nearby host communities, 9 access to markets, and host attitudes towards refugees. For example, when the dependent variable is household consumption, we compute the mean of that among host households living in or near camp c and include it as one of the predictors. To measure access to markets, we compute travel time from host households to the nearest urban agglomeration using data on urban agglomerations from GRID3 human settlement data5 and the friction map provided by Weiss (2018) and Weiss et al. (2018). Host attitudes towards refugees is measured based on the Attitudes Index – the average of 10 questions regarding beliefs about refugees’ character, the rights they should receive, and their impact on the host community, standardized to a mean of 0 and SD of 1, where positive indicates better attitudes. This model allows us to assess how the socio-economic environment and attitudes of the host community impact the welfare of refugees. Lastly, Xic represents a set of control variables at the individual or household level, such as age, literacy, household size, and years of arrival. Host community characteristics are expected to play a significant role in determining refugee welfare. For instance, refugees in communities with higher levels of economic devel- opment and more positive attitudes towards refugees may experience better welfare outcomes due to increased access to services and employment opportunities. Conversely, refugees in less favorable environments may struggle to achieve economic self-reliance and improve their welfare. We then estimate a model to assess the relationship between working outside the camp and the welfare of in-camp refugees: yic = αc + θWork Outside Campic + β Xic + ϵic (3) In this equation, yic represents the welfare outcome of refugee household or individual i in camp c. The term αc is a camp-specific fixed effect, capturing unobserved factors that vary across camps. The variable Work Outside Campic is a binary indicator, equal to 1 if the refugee household or individual works outside the camp and 0 otherwise. Xic includes household- or individual-specific characteristics as defined in Equation 2. This model is designed to help us understand how employment outside the camp influ- ences various welfare and labor outcomes as well as the individual’s sense of control over their life. Working outside the camp can provide refugees with access to a wider range of employment opportunities, potentially leading to higher wages and improved welfare. In ad- dition to the financial benefits, engaging in work outside the camp may enhance a refugee’s sense of autonomy and empowerment, both of which are important components of overall well-being. However, it is essential to recognize that working outside the camp may also 5 Available from https://data.grid3.org/datasets/GRID3::grid3-eth-settlement-extents-v3-0/about. 10 expose refugees to a range of risks. These risks can include exploitation in informal labor markets, discrimination based on their refugee status, and challenges in navigating legal and social systems that may not fully protect their rights. Furthermore, commuting to work outside the camp may involve transportation costs and safety concerns, which could offset some of the potential benefits. Lastly, we delve into the socio-economic profile of those refugees who choose to work outside camps. To this end, we estimate the following Linear Probability Model: yic = α + β Xic + δ Host Characteristicsc + ϵic (4) where yic takes on the value of one if individual i works outside camps, and zero other- wise. Xic represents a set of key individual characteristics (i.e., age, gender, literacy) as well as household-level features (i.e., per capita consumption (log), access to services as well as presence of other workers in the same family). Host Characteristicsc capture the character- istics of host environments – namely, access to market (measured by travel time to cities) and host attitudes towards refugees. By incorporating these considerations into the analysis, we aim to provide a comprehen- sive understanding of how working outside the camp may impact refugee welfare and what socio-economic factors drive refugees’ decisions to seek employment outside camps despite the regulatory restrictions. 5 Results 5.1 Baseline differences in Welfare: Hosts and in-camp refugees The analysis reveals significant differences in welfare between in-camp refugees and host communities. Even after controlling for baseline socio-economic characteristics and location- specific fixed effects, refugees in camps are found to be far poorer than their host counterparts (see Figure 1). Specifically, the welfare of in-camp refugees, as measured by total consump- tion per capita (log), is approximately 60 percent lower than that of hosts. Moreover, the poverty rate for in-camp refugees is roughly 40 percentage points higher than that of host communities. One major factor that drives this disparity between in-camp refugees and hosts is limited employment opportunities among in-camp refugees. Both labor force participation and em- ployment rate are significantly lower among in-camp refugees compared to hosts. Instead, in-camp refugees rely mostly on international aid to compensate for a lack of income coming from employment. 11 While the stark welfare gap between in-camp refugees and host communities is primarily driven by limited employment opportunities, it is important to consider the broader impli- cations of what could be described as “trapped poverty” (Malevolti and Romano, 2024). In-camp refugees are often placed in highly controlled environments with restricted access to labor markets, movement, and opportunities for economic advancement. Prolonged hu- manitarian aid, while essential for meeting refugees’ immediate needs, can also inadvertently trap them in poverty by fostering dependency and limiting opportunities for self-reliance. When aid is provided as a long-term substitute for employment rather than a complement to economic inclusion, refugees may struggle to develop skills, build professional networks, or access formal labor markets (Carpi, 2019). Restrictions on movement, work permits, and economic participation in many host countries further exacerbate this issue, forcing refugees to rely on diminishing aid rather than pursuing sustainable livelihoods. 5.2 Heterogeneity in welfare among in-camp refugees The analysis also reveals that the welfare of in-camp refugees is closely linked to the characteristics of the host communities where they are settled. No refugee camp, regardless of its location, is completely sealed off from the movement of goods, capital, and people from the outside (Betts et al., 2016; Werker, 2007). Our results show that, not surprisingly, refugees tend to experience better welfare outcomes when they are hosted in communities with more favorable socio-economic conditions and positive attitudes toward refugees. Our findings (Table 2) indicate a positive correlation between the welfare of host com- munities and that of refugees. The socio-economic environment of the host community plays a pivotal role in shaping refugee welfare. Overall, host communities with higher levels of welfare and better infrastructure (e.g., electricity, access to improved water and sanitation) are associated positively with better outcomes for refugees. These resources are vital not only for improving immediate welfare outcomes but also for supporting long-term integration and self-reliance. In contrast, refugees residing in economically disadvantaged communities may face significant barriers in accessing even the most basic services, which can perpetuate cycles of poverty and further marginalize them from the economic and social life of the host country (Pape et al., 2018). Social cohesion plays a crucial role in the economic integration of refugees, as the atti- tudes of host communities can significantly impact refugees’ opportunities for employment, entrepreneurship, and overall economic well-being. Our findings indeed show that positive attitudes of hosts towards refugees are associated positively with their welfare. When host communities have positive perceptions of refugees and view them as contributors rather than 12 competitors, refugees are more likely to access labor markets, establish business networks, and participate in the local economy (UNHCR, 2024; De Vroome and Van Tubergen, 2010). In contrast, when refugees are perceived as economic threats, they may face exclusionary practices, discrimination, or restrictions on accessing jobs, land, and essential social ser- vices (S¸ eker, 2023). Such negative perceptions not only hinder refugees’ ability to achieve self-reliance but also exacerbate social tensions, weakening trust and cooperation between refugees and hosts. A lack of social cohesion can disrupt commercial linkages, reduce market efficiencies, and ultimately create economic stagnation for both groups (Betts et al., 2023, 2016). Therefore, fostering inclusive policies and community engagement initiatives that promote social cohesion is essential for ensuring that refugees can contribute productively to the host economy while also strengthening economic resilience and social stability in host communities. Additionally, the overall resilience of the host community’s economy can influence how well both refugees and hosts cope with external shocks, such as economic downturns or environmental crises. In more economically stable areas, refugees may find it easier to establish livelihoods and contribute to the local economy, which can create a virtuous cycle of mutual benefit. Therefore, policies that aim to improve the welfare of refugees must also consider the broader development needs of host communities, recognizing that the well-being of both groups is often interdependent (Vemuru et al., 2020; Betts et al., 2022; Walelign et al., 2022). That said, it is worth highlighting that host environments do not appear to improve labor outcomes for refugees. Indeed, our analysis shows no strong correlation between labor outcomes in host communities and those for in-camp refugees. This finding applies to various labor outcomes, including labor force participation, employment rates, off-farm employment, and wages. However, it underscores the importance of easing restrictions on refugees’ rights to work outside camps – as was recently introduced in Ethiopia – because these restrictions may limit their ability to access better labor opportunities and integrate into local economies. Allowing refugees to work freely in host communities can not only improve their economic outcomes, but also contribute to local development, creating a mutually beneficial situation for both refugees and their hosts. On a related note, the relationship between refugees and host communities need not be unidirectional. Refugees, when given access to economic opportunities, can positively contribute to local economies by filling labor shortages, creating businesses, and generating demand for goods and services (Evans and Fitzgerald, 2017; Loschmann et al., 2019; Sanghi et al., 2016). In many settings, refugees have introduced new skills, entrepreneurial ventures, and labor force participation that strengthen local economies rather than burden them. For 13 example, studies have shown that refugee-led businesses and enterprises can create jobs not only for other refugees, but also for local residents, fostering economic spillovers that benefit the host population. Additionally, international aid directed toward refugees may also stimulate the local economy, as financial assistance and relief efforts increase the flow of resources, benefiting host communities indirectly through job creation and infrastructure development (Zhou et al., 2023). 5.3 Employment outside camps A significant proportion of in-camp refugees seek employment outside the camps despite not having work permits at the time. About 40 percent of in-camp refugees who worked during the last 12 months indicated that they worked outside the camps. This section delves into whether refugees seeking out-of-camp jobs would reap any welfare premiums compared to those working within camps. In-camp refugees with a job outside camps do not appear to reap any noticeable welfare premiums (Table 3). In fact, working outside camps is associated with lower household welfare. Furthermore, refugees working outside camps are less likely to work in off-farm or formal employment compared to those working inside camps, and agriculture and informal jobs are often associated with lower household welfare and higher poverty (Kaiser, 2006; Ruiz and Vargas-Silva, 2015). Given that international humanitarian NGOs provide many formal jobs within camps – accounting for 46 percent of formal jobs – limited formal employment opportunities in surrounding host communities present a challenge to those in-camp refugees who would need more productive jobs to enhance their living standards. That said, of those refugees working outside camps who reported wages earned (N = 78)– which account for a small fraction of the full sample of refugee workers (N = 1,028) – their reported monthly wages tend to be roughly 36 percent higher than those earned by refugees working inside camps. Despite not offering any clear welfare benefits, however, working outside camps is posi- tively associated with a stronger sense of autonomy and self-determination, which are crucial elements of well-being. This enhanced sense of control over their circumstances is a vital component of overall welfare. Prior research also shows that refugees in camps, who depend heavily on international aid, often experience a lower sense of control, which is linked to poorer mental health and economic outcomes (Wieser et al., 2023; Hahn et al., 2019; Thum, 2014; Tsionis et al., 2024; Hoogeveen and Obi, 2024; Hussam et al., 2022). However, working outside the camps can expose refugees to significant risks, including exploitation, discrim- ination, and potential legal consequences, especially when they do not have the necessary 14 documentation to work legally. In the absence of work permits, refugees are often susceptible to unfair labor practices, reduced wages, and poor working conditions, with limited access to legal protections. In addition to these legal and safety risks, refugees working outside the camps may also struggle to balance work with other responsibilities, such as caring for family members or managing obligations within the camp. The need to travel long distances for work or live separately from their families can introduce additional stress and logistical challenges, further complicating their efforts to maintain a stable life. Overall, refugees who choose to work outside camps tend to have lower socio-economic profiles. The regression results in Table 4 provide insights into the socio-economic factors influencing refugees’ decisions to work outside camps. The coefficients for gender, literacy, and total consumption are statistically significant across all four models, indicating a strong association with refugees’ decisions to work outside camps. Female refugees are less likely to engage in such employment, as reflected by consistently negative coefficients, particularly in Model 4. Literacy also shows a negative relationship with the probability of working outside camps. Similarly, lower levels of total consumption are linked to a higher likelihood of seeking employment beyond camp boundaries. In contrast, other household and location- level variables – such as access to services, the presence of other workers in the household, travel time, and host community attitudes – do not exhibit significant associations. 6 Conclusion In conclusion, this study offers a comprehensive analysis of welfare disparities between refugees and host communities in Ethiopia, drawing on data from SESRE. The findings highlight significant welfare gaps, with refugees – particularly those in camps – experiencing severe economic disadvantages. These disparities are largely driven by restricted access to employment opportunities, which perpetuates dependence on aid and hinders refugees’ paths toward self-reliance and economic integration. To improve refugee welfare and support economic integration, it is essential to expand both de jure (legal) and de facto access to employment opportunities for refugees. Allowing refugees to work outside camps can foster self-reliance, reduce aid dependency, and stimulate local economic growth. Developing formal employment avenues will also protect refugees from the risks associated with informal labor markets. Together, these initiatives provide a foundation for sustainable livelihoods and economic inclusion. In this respect, Ethiopia’s recent Directive to operationalize the legal rights to work for refugees in Ethiopia is an important step towards self-reliance of refugees in the country. Yet, whether the new Directive for refugees to seek employment in Ethiopia is a signifi- 15 cant step forward towards improving welfare, not only for the refugees themselves but also for the host communities, remains to be seen. This Directive is crucial as it provides a legal framework that enables refugees to transition from aid dependency to self-reliance, fostering their integration into the national economy. By addressing structural issues such as lack of employment or informal employment, the Directive aims to create a more inclusive and sustainable economic environment. This policy shift aligns with Ethiopia’s commitment to inclusive development and recognizes that work permits alone are insufficient for positive welfare outcomes unless accompanied by measures to tackle underlying employment chal- lenges. While the Directive offers hope for overcoming previous implementation gaps, its full impact remains to be seen. The success of Ethiopia’s model will depend on the effective execution of this Directive and the creation of economic opportunities that allow refugees to contribute meaningfully to their host communities, thereby benefiting both refugees and the broader society. Our study further indicates that many in-camp refugees seek employment outside camps due to the urgent need to escape poverty and gain more autonomy. However, off-farm or formal job opportunities within host communities remain limited, and there is no evidence that households with members working outside camps fare better in terms of welfare than those working within camps. Those refugees who work outside camps often come from poorer households and have lower levels of education. Addressing this requires policies that allow refugees the freedom to move to areas with higher economic potential, which could enhance job prospects, increase local demand, and contribute to rural transformation—an essential step for Ethiopia’s structural development. In this context, area-based policy interventions for refugees and their hosts are essential for fostering social cohesion and economic develop- ment in regions affected by forced displacement, to ensure that the benefits of such policies are equitably distributed (Schell et al., 2020). Lastly, the study underscores the linkage between host community characteristics and refugee welfare. Refugees benefit more when host communities have favorable socio-economic conditions and positive attitudes toward refugees. Strengthening social and economic inter- actions between refugees and hosts can create mutual benefits, fostering social cohesion and enabling positive spillovers from infrastructure investments. Although poverty levels in host communities are generally lower than among in-camp refugees, these communities still face persistent poverty and inadequate access to essential public services. Addressing these needs through targeted investments can yield long-term benefits for both refugees and hosts, supporting more inclusive and resilient communities. 16 7 Tables & Figures 17 Table 1: Descriptive statistics Household-level Host communities In-camp refugees Variables N. of Mean Std. Dev. N. of Mean Std. Dev. households households Total consumption (pc) 1718 36535 25953 1287 14052 8741 Poverty (USD2.15 per day) 1718 0.249 0.432 1287 0.837 0.369 Rely on aid 1718 0.007 0.083 1287 0.779 0.415 Access to public services (index) 1718 0.780 1.371 1287 -0.411 0.788 Arrived before 2000 0 1269 0.040 0.196 Arrived between 2000-2009 0 1269 0.749 0.434 Arrived between 2010-2019 0 1269 0.209 0.407 Arrived between 2020-2022 0 1269 0.002 0.043 Control Index (internal) 1718 2.931 0.707 1287 2.654 0.833 Control Index (chance) 1718 1.741 0.821 1287 1.649 0.829 Control index (others) 1718 1.509 1.100 1287 1.694 1.141 Host attitudes (index) 1718 2.799 0.403 0 Individual-level Host communities Refugees Variables N. of Mean Std. Dev. N. of Mean Std. Dev. 18 individuals individuals Age 15-29 8165 0.132 0.338 7357 0.166 0.372 Age 30-44 8165 0.476 0.499 7357 0.552 0.497 Age 45-64 8165 0.288 0.453 7357 0.252 0.434 Age above or equal 65 8165 0.104 0.305 7357 0.030 0.172 Literacy (read and write) 7469 0.764 0.425 6662 0.607 0.488 Work outside camp 0 1028 0.402 0.490 Labor force participation 4385 0.599 0.490 3502 0.311 0.463 Employed (past 12 months) 4385 0.562 0.496 3502 0.290 0.454 Formal employment (past 12 months) 2367 0.602 0.490 1028 0.277 0.448 Off-farm employemnt (past 12 months) 2385 0.858 0.350 1039 0.823 0.382 Monthly wage (log) 893 8.425 0.860 249 7.327 0.781 Note : Poverty is measured based on households living under the international poverty line of USD2.15 per day. Access to services measured based on the composite index of access to electricity, improved water and improved sanitation. The access to public services index was calculated using a weighted-average, with weights derived from principal component analysis. This control index is the unweighted average of 10 questions about feelings of control over one’s fate. The index ranges from 0 to 4, where more positive indicates greater control. The internal control index uses four questions regarding personal control over destiny. The chance index uses five questions regarding the role of chance or determinism. The role of the powerful others index is 1 question on whether others determine fate. Host attitudes towards refugees is measured based on the Attitudes Index – the average of 10 questions regarding beliefs about refugees’ character, the rights they should receive, and their impact on the host community, standardized to a mean of 0 and SD of 1, where positive indicates better attitudes. Figure 1: Regression outcomes: Baseline Note : The figure shows the regression coefficients with the 95% confidence intervals. All these regressions control for location/camp fixed effects. Table 2: Welfare regression among in-camp refugees. Welfare Public services Labor outcomes Dependent Consumption Poverty Electricity Improved Improved LFP Employed Off-farm Earning variable Water Sanitation emp. (log) (1) (2) (3) (4) (5) (6) (7) (8) (9) Host outcome 0.363* 0.231* 0.202* 0.268* 0.776* 0.065 0.314 -0.075 -0.029 (0.104) (0.082) (0.084) (0.11) (0.182) (0.173) (0.189) (0.148) (0.312) Travel time to city -0.034 0.041* 0.014 -0.005 0.033 -0.017 -0.023 -0.054* -0.077 (0.02) (0.017) (0.018) (0.006) (0.018) (0.013) (0.013) (0.016) (0.044) Host attitudes 0.186* -0.092 0.026 -0.029 -0.074 0.026 0.045 -0.163 0.082 (0.073) (0.048) (0.015) (0.063) (0.101) (0.052) (0.057) (0.103) (0.239) Household size -0.085* 0.042* 0 -0.004 0.007 -0.005 -0.007 -0.003 0.027 (0.007) (0.006) (0.001) (0.004) (0.007) (0.005) (0.004) (0.006) (0.021) Literacy 0.118* -0.084* -0.015 -0.026 0.005 -0.216* -0.198* 0.043 0.092 (0.038) (0.027) (0.012) (0.026) (0.041) (0.021) (0.022) (0.033) (0.106) N. of observations 1269 1269 1269 1269 1269 3351 3351 1013 243 Note : All regressions control for age and years of arrival. In this table, host outcome refers to the average values of yic among nearby host communities (see Equation 2). For example, when the dependent variable is household consumption (Model 1), we compute the mean of that among host households living in or near camp c and include it as one of the predictors. In these regressions, camp-fixed effects are not included because doing so would not allow us to include host environment characteristics as a covariate. Standard errors clustered at the EA level. * denotes statistical significance at the 0.05 level. 19 Table 3: Regression results on out-of-camp employment and refugee welfare and well-being Welfare Labor outcomes Control index Dependent variable Consumption Poverty Off-farm Formal Earning Internal Chance/Fate Powerful emp. emp. (log) others (1) (2) (3) (4) (5) (6) (7) (8) Work outside camp -0.087* 0.045 -0.222* -0.084* 0.363* 0.132* -0.17 -0.205* (0.035) (0.027) (0.048) (0.037) (0.136) (0.06) (0.135) (0.086) Household size -0.084* 0.04* 0.001 -0.004 0.014 -0.014 0.064* 0.019 (0.009) (0.01) (0.004) (0.005) (0.019) (0.014) (0.027) (0.015) Literacy 0.086 -0.058 0.035 0.097* 0.253* 0.042 -0.07 0.054 (0.047) (0.037) (0.032) (0.032) (0.102) (0.084) (0.097) (0.09) Access to services 0.089* -0.018 0.019 0.016 0.02 -0.036 0.124 0.087 (0.025) (0.023) (0.026) (0.028) (0.082) (0.065) (0.085) (0.066) N. of observations 596 596 1002 1002 243 596 596 596 Note : All regressions include controls for age and years of arrival as well as camp-fixed effects. Standard errors clustered at the EA level. Note that the sample here is only limited to those who indicated working in the past 12 months. * denotes statistical significance at the 0.05 level. Table 4: Regression results on the socio-economic profile of refugees working outside camps Model (1) (2) (3) (4) Female -0.149* -0.163* -0.163* -0.179* (0.052) (0.052) (0.05) (0.049) Literacy -0.113* -0.105* -0.108* -0.135* (0.036) (0.038) (0.039) (0.042) Total consumption -0.078* -0.063 -0.087* (0.037) (0.04) (0.042) Access to services -0.006 0 -0.013 (0.025) (0.029) (0.027) Other worker in the same household 0.007 0.01 0.003 (0.055) (0.056) (0.057) Travel time to city 0.021 (0.02) Host attitudes -0.021 (0.112) N of observations 1028 1028 1028 1028 Household-level characteristics No Yes Yes Yes Camp-level characteristics No No Yes No Camp-fixed effects No No No Yes Note : All regressions include controls for age and years of arrival. Model 4 includes camp-fixed effects. Standard errors clustered at the EA level. Note that the sample here is only limited to those who indicated working in the past 12 months. * denotes statistical significance at the 0.05 level. References Alam, A., Dutta, I., Haque, M. E., and Nogales, R. (2022). Impact of rohingya refugees on food prices in bangladesh: Evidence from a natural experiment. World Development, 154:105873. 20 Alloush, M., Taylor, J. E., Gupta, A., Rojas Valdes, R. I., and Gonzalez-Estrada, E. (2017). Economic life in refugee camps. World Development, 95:334–347. Ayenew, A. B. (2021). Welfare impact of hosting refugees in ethiopia. World Bank Policy Research Working Paper 9613, World Bank, Washington, DC. License: CC BY 3.0 IGO. Betts, A., Bloom, L., Kaplan, J., and Naohiko, J. (2016). 40Refugee economies. In Refugee Economies: Forced Displacement and Development. Oxford University Press. Betts, A., Chaara, I., Omata, N., and Sterck, O. (2019). Refugee economies in uganda: what difference does the self-reliance model make? Refugee Studies Centre. Betts, A., Flinder Stierna, M., Omata, N., and Sterck, O. (2024). The economic lives of refugees. World Development, 182:106693. Betts, A., Stierna, M. F., Omata, N., and Sterck, O. (2022). Social cohesion and refugee-host interactions evidence from east africa. Betts, A., Stierna, M. F., Omata, N., and Sterck, O. (2023). Refugees welcome? inter-group interaction and host community attitude formation. World Development, 161:106088. Black, R. (2001). Environmental refugees: myth or reality? Carpi, E. (2019). Towards a neo-cosmetic humanitarianism: Refugee self-reliance as a social- cohesion regime in lebanon’s halba. Journal of Refugee Studies, 33(1):224–244. Ceritoglu, E., Yunculer, H. B. G., Torun, H., and Tumen, S. (2017). The impact of syrian refugees on natives’ labor market outcomes in turkey: evidence from a quasi-experimental design. IZA Journal of Labor Policy, 6:1–28. Crawford, N., Cosgrave, J., Haysom, S., and Walicki, N. (2022). Protracted displacement: uncertain paths to self-reliance in exile. ODI. De Vroome, T. and Van Tubergen, F. (2010). The employment experience of refugees in the netherlands 1. International Migration Review, 44(2):376–403. e, M. (2021). UN Sustainable Development Goals and the “Refugee Denaro, C. and Giuffr´ Gap”: Leaving Refugees Behind? Refugee Survey Quarterly, 41(1):79–107. Evans, W. N. and Fitzgerald, D. (2017). The economic and social outcomes of refugees in the united states: Evidence from the acs. Working Paper 23498, National Bureau of Economic Research (NBER). 21 Ferris, E. (2010). The Politics of Protection: The Limits of Humanitarian Action. Brookings Institution Press, Washington, D.C. Hahn, E., Richter, D., Schupp, J., and Back, M. D. (2019). Predictors of refugee adjustment: The importance of cognitive skills and personality. Collabra: Psychology, 5(1):23. Hoogeveen, J. and Obi, C., editors (2024). A Triple Win: Fiscal and Welfare Benefits of Economic Participation by Syrian Refugees in Jordan. World Bank, Washington, DC. License: CC BY 3.0 IGO. Hussam, R., Kelley, E. M., Lane, G., and Zahra, F. (2022). The psychosocial value of employ- ment: Evidence from a refugee camp. American Economic Review, 112(11):3694–3724. IPCC (2019). Special report on the ocean and cryosphere in a changing climate. Retrieved from. Jacobsen, K. (2002). Livelihoods in conflict: the pursuit of livelihoods by refugees and the impact on the human security of host communities. International migration, 40(5):95–123. Kaiser, T. (2006). Between a camp and a hard place: rights, livelihood and experiences of the local settlement system for long-term refugees in uganda. The Journal of Modern African Studies, 44(4):597–621. ¨ and Siegel, M. (2019). Considering the benefits of hosting Loschmann, C., Bilgili, O., refugees: evidence of refugee camps influencing local labour market activity and economic welfare in rwanda. IZA Journal of Development and Migration, 9(5). Malevolti, G. and Romano, D. (2024). Poverty dynamics and poverty traps among refugee and host communities in uganda. The Journal of Development Studies, 60(3):380–405. Martin, S. F. (2010). The causes and consequences of forced migration. Masaki, T. and Madson, B. (2023). Data gaps in microdata in the context of forced dis- placement. Policy Research Working Paper WPS 10631, World Bank Group, Washington, D.C. Neikirk, A. and Nickson, R. (2023). Towards a spatial analysis of refugees working outside camps. Journal of Human Trafficking, pages 1–14. Nguyen, N. T. V., Savadogo, A., and Tanaka, T. (2021). Refugees in Chad: The Road Forward. World Bank, Washington, DC. 22 Pape, U. J., Petrini, B., and Iqbal, S. A. (2018). Informing durable solutions by micro-data: a skills survey for refugees in ethiopia. Technical report, The World Bank. Pape, U. J., Petrini, B., and Iqbal, S. A. (2021). Informing durable solutions by micro-data: A skills survey for refugees in ethiopia. Technical report, World Bank Group, Washington, D.C. Ruiz, I. and Vargas-Silva, C. (2015). The labor market impacts of forced migration. American Economic Review, 105(5):581–586. Sanghi, A., Onder, H., and Vemuru, V. (2016). Yes in My Backyard? The Economics of Refugees and Their Social Dynamics in Kakuma, Kenya. World Bank Group, Washington, D.C. Schell, J., Hilmi, M., and Hirano, S. (2020). Area-based approaches: An alternative in contexts of urban displacement. Forced Migration Review, 63. ¸ eker, B. D. (2023). Attitudes of the host population towards syrian refugees: A new S theoretical perspective. Journal of Psychological Research, 5(4):36–49. Thum, A.-E. (2014). Labor Market Integration of German Immigrants and Their Children: Does Personality Matter? Technical report. Tsionis, A., Pantoglou, D., and Kasvikis, Y. (2024). Mental health locus of control in refugees with clinically established psychopathology. Psychiatriki, 35(2):103–111. Epub 2022 Nov 11. UNHCR (2020). Global trends: Forced displacement in 2019. Retrieved from. UNHCR (2023a). Ethiopia country refugee response plan 2023. UNHCR (2023b). Mid-year trends. Geneva. UNHCR (2024). Refugee policy review framework update as at 30 june 2023. [accessed 15 September 2024]. Vemuru, V., Sarkar, A., and Fitri Woodhouse, A. (2020). Impact of refugees on hosting communities in ethiopia: a social analysis. Verme, P. and Schuettler, K. (2021). The impact of forced displacement on host communities: A review of the empirical literature in economics. Journal of Development Economics, 150:102606. 23 Walelign, S. Z., Wang Sonne, S. E., and Seshan, G. (2022). Livelihood Impacts of Refugees on Host Communities: Evidence from Ethiopia. The World Bank. Weiss, D. (2018). A global map of travel time to cities. Weiss, D. J., Nelson, A., Gibson, H. S., et al. (2018). A global map of travel time to cities to assess inequalities in accessibility in 2015. Nature, 553:333–336. Werker, E. (2007). Refugee camp economies. Journal of Refugee Studies, 20(3):461–480. Wieser, C., Tesfaye, W. M., Abeje, F. G., Lebow, J. A., and Yonis, M. B. (2023). Expanding development approaches to refugees and their hosts in ethiopia. Technical report, World Bank Group, Washington, D.C. Wieser, C., Tesfaye, W. M., Abeje, F. G., Lebow, J. A., and Yonis, M. B. (2024). Expanding development approaches to refugees and their hosts in ethiopia. World Bank (2017). Forcibly displaced: Toward a development approach supporting refugees, the internally displaced, and their hosts. World Bank (2019). Informing the refugee policy response in uganda: Results from the uganda refugee and host communities 2018 household survey. Technical report, World Bank Group, Washington, D.C. Zetter, R. and Ruaudel, H. (2016). Refugees’ right to work and access to labor markets–an assessment. World Bank Global Program on Forced Displacement (GPFD) and the Global Knowledge Partnership on Migration and Development (KNOMAD) Thematic Working Group on Forced Migration. KNOMAD Working Paper. Washington, DC: World Bank Group. Zetter, R. and Ruaudel, H. (2018). Refugees’ right to work and access to labour markets: Constraints, challenges and ways forward. Forced Migration Review, 58. Zhou, Y.-Y., Grossman, G., and Ge, S. (2023). Inclusive refugee-hosting can improve local development and prevent public backlash. World Development, 166:106203. 24