Intent Versus Impact: Is the Choice Between Encampment and Out-of-Camp a Factual Hard Policy Dilemma?

Abstract

Executive summary

Recent debates in the migration field have indicated that the deficiencies of specific policies are attributable to so-called hard dilemmas: mutually exclusive options that are incapable of being reconciled. The central question posed in this study is whether the choice between refugee encampment and out-of-camp policies (OCPs) which allow for freedom of movement constitutes such a dilemma for policy makers. Proponents of encampment policies argue that the loss of freedom of movement is a small price to pay for the efficient provision of adequate shelter, protection, and basic aid. Moreover, they argue that encampment reduces the risk of heightened criminality; the risk of conflicts spilling over borders; and the risk of competition for scarce resources or labor with members of hosting communities. Conversely, advocates of out-of-camp policies argue that freedom of movement helps refugees to become self-reliant and to integrate into host communities. They consider encampment policies to be detrimental, leading to aid dependence and a life in limbo. Ethiopia has explored both options to host its large group of refugees. For a long time, encampment was the main approach. Since the 2019 revised Refugee Proclamation, out-of-camp settlement is allowed but only under specific circumstances and for specific groups. Until today, the overwhelming majority of refugees still resides in camps. This paper analyses Ethiopia’s refugee policies on paper, and its impact on the ground. For this, we specifically examine the experiences of Eritrean refugees in both camp and non-camp settings in Ethiopia. Our analysis demonstrates that discrepancies between policy intentions on paper and actual impacts have been resolved in practice. A combination of encampment and out-of-camp policies can be effective. There is no need to make an either/or decision about which policy to adopt. Policies that combine the best of both worlds should be guided by a bottom up assessment that considers the agency of refugees and host communities. Based on our findings, we formulate some concrete policy recommendations on how to develop refugee policies.

Recommendations:

- Refugee policy makers in general and Ethiopia’s Refugees and Returnees Services (RRS) in particular, should explore encampment and out-of-camp policies as complementary options and not as two mutually exclusive alternatives. Giving refugees more freedom to follow their preferences and building on existing practices will help to unleash refugees’ potential and to integrate better into the host society.

- It is detrimental to force beneficiaries of Ethiopia’s OCP to give up any institutional support they receive. The current practice contributes to unemployment and dependency in a country where around one in six citizens depend on humanitarian assistance. We therefore recommend that the OCP status should automatically include work permits.

- We recommend the Ethiopian Ministry of Labor and Skills to allow OCP beneficiaries to engage in business or seek employment opportunities and we recommend the Ethiopian RRS to consider allowing refugees to settle in their preferred locations, in accordance with the provisions of the 2019 Refugee Proclamation. This would avoid the negative effects of joining the informal economy and competing with Ethiopian citizens.

- To understand the impact of different refugee policies, researchers, and policy advisors should include both refugees and members of host and surrounding communities in their assessments, as the two are profoundly interdependent.

- Generally, focusing only a preconceived objective when determining the way forward is detrimental to a nuanced understanding of policy impacts. Instead, policymakers worldwide must assess the actual impacts of their policies on the ground in an evidence-based and unbiased manner.

Introduction

This is a prison, where the young’s dream is shattered, where they cannot work or bring any change in their lives. We are spared from the prisons in Eritrea that is all . . . the refugee camps in Ethiopia, [are] a big prison [where] you are allowed to see the sun and feel the wind (Eritrean refugee in Shimelba camp, Ethiopia, November 2019).

The words of this refugee, describing his situation in a camp in Ethiopia, are representative for a larger group of refugees. They are living in a state of indeterminate waiting in refugee camps. For them, the lack of freedom of movement and high levels of aid dependency are the most relevant shortcomings of encampment policies. Nowadays also policymakers agree that camps are not ideal places to accommodate refugees, especially when displacement is protracted. The Ethiopian government nevertheless used to promote encampment as a way to protect national interests, as non-governmental organization (NGO) staff members suggested to us (various interviews with experts and officials, Addis Ababa, October 2015; see also Proclamation No. 409/2004).

Ethiopia hosts 1.1 million refugees mostly from South Sudan, Somalia, Eritrea, and Sudan, mostly in 25 refugee camps, which are all located in the peripheral borderlands. Only a small number — 7 percent of them — live in Addis Ababa (UNHCR 2025) and about half of them do not have official registration to live in the city. Another four refugee camps located on the Eritrean border were destroyed and vacated during the Tigray war (2020–20220) that was fought between the Federal Government of Ethiopia and TPLF, the ruling party in the Tigray Regional State.

Drawing on our extended research among Eritrean refugees between 2019 and 2023 in six refugee camps and towns located in Tigray and Afar, and Addis Ababa, this paper examines the approaches, assumptions, and contradictions in the refugee policies vis-a-vis the everyday lived experiences of the refugees. The article argues that policy analysts’ arguments about the lack of coherence of refugee policies is not only rooted in contradictions between international and domestic political realities but also in perceptions — among both policymakers and analysts — that “[h]ard ethical policy dilemmas . . . involve a conflict between two morally worthy goals that cannot be hierarchically ordered and therefore cannot be easily reconciled.” (Bauböck, Permoser, and Ruhs 2022: 433). We show that in practice, staying in a camp or settling out of camps is not a hard dilemma for refugees if they are provided the freedom to navigate between the two options and to benefit from the best of both worlds at different moments in time. We argue that what seems to be a policy dilemma is in fact a choice that refugees can make themselves. Instead of creating barriers to move between options, policy can support refugees’ choices by allowing for flexible navigation between the two options. For this, it is important to look at actual impact of policies rather than at intentions.

Many policymakers struggle how best to protect refugees’ rights in line with internationally agreed upon legal and policy frameworks. The two main approaches of refugee hosting countries are encampment and Out-of-Camp Policies (OCPs) — a policy which allows refugees to live outside of camps. Encampment is seen as an effective way to guarantee physical safety to refugees, to provide humanitarian assistance, and to enable people’s mere survival (Agamben and Hiepko 1998; Jansen 2013). Camps are most suitable to provide medical and social assistance, nutrition, education, housing, and protection (Crisp 2017; Brankamp 2022). Following this policy preference, academic research also focused primarily on camp settings, but overlooked the self-settled ¹ refugees to a large extent (Maple 2024). Over time the former gained prominence over the latter: Encampment policies were criticized for undermining people’s rights to freedom of movement, economic development, social inclusion, political participation, and individual autonomy (Black 1998). OCPs are increasingly seen as a better approach, ensuring refugees’ freedom of movement, economic development, and social integration into the host society (Crisp, Morris, and Refstie 2012). With that turn, academic attention for self-settlement increased, especially in the urban space (Maple 2024).

Significantly, in 2014 UNHCR took a stance in the debate favoring OCPs: “UNHCR’s policy is to pursue alternatives to camps, whenever possible, while ensuring that refugees are protected and assisted effectively and are able to achieve solutions” (UNHCR 2014). This is in line with the dominant position among academics about the preferred solution (Bakewell 2014: 127; Betts 2021).

While the international community has reached an agreement laid out in the Global Compact for Refugees, individual nation-states’ internal politics did not necessarily yield the same outcome: On the contrary, domestically, refugee settlements are often perceived to be against the interests of the host communities; namely the interest to protect the national population against the loss of jobs, against increasing criminality, against rising social tension between refugees and members of host communities, against political infiltration, against spill-over of armed conflicts, and finally against transborder security issues (Loescher and Milner 2013). In addition, advocates of encampment policies argue that they allow a better provision of adequate shelter, protection, and basic aid. But costs of encampment are increasingly highlighted: as refugees are disallowed to seek their own means of subsistence, to create connections in host communities, and to integrate locally, they remain aid dependent. Despite maintaining that it might heighten security and economic risks for the hosting population, policy makers in hosting nations are gradually agreeing more to the freedom of movement or OCPs, hoping that humanitarian aid for refugees will also reach their own populations as foreseen in the Global Compact on Refugees (GCR; Dick and Rudolf 2019).

This article is an extension of the recent discussion amongst various policy experts on ethical policy dilemmas in the field of migration (Baubock, Permoser, and Ruhs 2022; Schmid 2024). The case of Ethiopia provides an ideal ground to scrutinize the assumptions of encampment versus OCPs. The country has consistently been amongst the African countries hosting the largest numbers of refugees and asylum seekers, mostly from South Sudan, Somalia, and Eritrea. For a long time, Ethiopia favored the encampment model for refugees, and prioritized security concerns (Proclamation No. 409/2004). This changed with the adoption of the 2019 revised Refugee Proclamation, that allowed the expansion of out-of-camp settlement under specific circumstances. ² Yet, thus far, 93 percent of the 1.1 million refugees in Ethiopia are still residing in camps (UNHCR 2025) making the policy discourse and practices of encampment versus OCP quite crucial.

In the following, we first present some background on the methods of data collection that we used for this paper. We then provide our theoretical approach, and a brief overview of relevant policy approaches at global and (east) African levels, and how they have been analyzed by scholars. From there, we dive into a presentation of our findings on Ethiopia’s refugee policy on paper and in practice, based on our data. We then discuss the consequences for policymakers of considering the lived reality of refugee policies.

Methods of Data Collection

This paper is based on a review of academic and policy documents and on empirical findings. Empirical data were collected from 2019 to 2023 by a team of local and international researchers and complemented with interviews conducted in 2015 before the official shift of encampment policy. Taking a qualitative mixed methods approach, the team conducted in-depth qualitative narrative interviews, combined with on-the-spot observations, focus group discussions, and quantitative surveys. Research addressed refugees, members of the host communities, experts, international humanitarian and development actors, local civil society organizations, and political organizations. Almost 90 percent of our respondents were Eritrean refugees; 60 percent of the respondents were male. The share of respondents living in camps was roughly the same as those living outside of camps. In addition we conducted expert interviews with practitioners, policy makers, and consultants from the top level to local staff on the ground, representing intergovernmental organizations, donors from various countries, governmental institutions, international and national NGOs, and community-based organizations. They provided insights into processes that usually take place behind closed doors. All names in the paper have been changed and all personal data has been anonymized to ensure the safety of the participants.

Our Theoretical Approach Toward Global Encampment Policies

This paper assumes that social reality is socially constructed. Concepts about challenges and policies addressing them are not detached from social reality, but co-produce it. Policy-relevant categories are as much constitutive for social reality as they — on the meta-level — constrain analysis. This is especially the case within the strongly policy-oriented field of refugee studies (Bakewell 2008). A person is, for example, identified and limited legally, socially, and politically by the category “refugee” in his daily life. At the same time, analysis of refugees is also shaped by the categories we commonly use and hence shape our understanding of reality. This two-way relationship, coined double hermeneutic (Giddens and Dallmayr 1982), in which the analyst is shaping the object of analysis, is not always reflected in refugee policies.

Yet, — almost by definition — policies translate theoretical analysis into practical action, even if that reality does not exactly fit the concepts we have. Refugees’ responses to policies may sometimes even be transformative, and change the way in which policies are foreseen to work in practice ,but only if we are open for that reality (Bakewell 2010). Even though the impact on the ground is what policymakers claim to care about (e.g., evaluating intended outcomes), self-reflection about the role of policy-related analysis in constituting the very practices targeted by policies remains the exception. As a result, the unintended practical impacts of policy interventions may be overlooked in this way (Koch and Schulpen 2018).

Is the choice between encampment and OCP a real dilemma if we consider not only policy intent, but also actual practices and the impact of policy on the ground? Does reality fit with the prescriptions of policy? Bauböck, Permoser, and Ruhs (2022). differentiate hard from “false dilemmas,” arguing that the latter emerge from “ . . . faulty theoretical analysis of a problem or from misinterpretation of facts” (p. 430). They therefore demand a bottom-up approach that is “ . . . value-based as well as fact-based and empirically grounded” (ibid.: 430). Looking at the policy debate on encampment versus OCP respectively as reflecting a protection versus freedom of movement dilemma underlines the necessity of fact-based assessments. We argue that these dilemmas — as many other policy issues — are often rather based on projections than on facts. We suggest to closely consider both intent and impact to understand policy options, as policies by definition strive for an impact.

Bauböck, Permoser, and Ruhs (2022, 435) distinguish two ways to analyze the dilemmas perceived by policymakers: “ . . . economists, political scientists, sociologists, and other social scientists ” look at “structural tensions . . . under the label of ‘paradoxes,’” while “ phenomenological, ethnographic, or discursive perspective . . . analysis seeks to identify and describe the morally worthy goals behind the horns of the dilemma as they are perceived, interpreted, and experienced by the actors in the field.” Once dilemmas translate into action through the imposition of policies, affected actors will react to them, and this may change the context, after which policies may need to be adjusted again, in what Bakewell (2010) describes as “morphogenesis.”

For us, this reifies the importance to approach dilemmas by looking at lived realities. We can then firstly identify the advantages and disadvantages of the elements of the measures; and secondly look at the way they work out, and how this then impacts on how people perceive the policy. While it is fruitful to identify the values directing action and framing policy making, these values need to be assessed as social reflections of reality constructed by those perceiving, analyzing, and addressing them. Identifying the reasons and mechanisms that underlie the perception of an issue as a dilemma, thus, help to reframe policies tackling them.

Global Encampment Policy in Perspective

Refugee policies are often a product of the cooperation between intergovernmental, governmental, or non-governmental organizations and their interests. International actors, foremost UNHCR, are responsible for providing protection and, at best, solutions to refugees. ³ The more countries (choose to) depend on international actors to finance and implement their policies, the larger the influence of these actors becomes and the more weight is given to their definitions and norms. ⁴ In consequence, the international debate has evolved into an asymmetrical relation: Countries hosting the largest numbers of refugees have only limited international bargaining power, whereas (donor) countries with lower numbers of refugees have more influence in formulating the global refugee regime.

Eighty percent of the world’s refugees are hosted in the Global South (UNHCR 2022b). Yet, many of them feel that international refugee policies are neither shaped in their interest, nor are they reflective of the national, political, economic, and sociocultural realities of the majority of hosting states (Dick and Rudolf 2019). Recognizing this asymmetry and taking into account the agenda and agency of actors in the Global South is a first step to better understand the gaps between intent and impact. The decision of the Ethiopian state to officially change the encampment policy was largely shaped by international actors pushing for policy changes and not by demands from the ground (ibid). Certain shortfalls in implementing these changes are due to the agency of actors that consciously or unconsciously react to the described asymmetrical relation. ⁵

Significantly, the UNHCR, — the main actor in the field of refugee protection — published a policy document in 2006, titled “Operational protection in camps and settlements,” and subtitled “A reference guide of good practices in the protection of refugees and other persons of concern” (UNHCR 2006). The reference guide was meant to “address the difficulties of translating policy into practice” (and hence not the difficulties of developing good policy). According to the guide, a good practice follows certain essential principles, such as a rights-based approach. In more recent years, UNHCR’s policy documents have shifted toward providing guidance on how to realize alternatives to encampment, underlining the need to respect people’s right to freedom of movement. The latest version of the UNHCR’s Emergency Handbook acknowledges the major shortcomings of camps: “they often limit the rights and freedoms of refugees such as their ability to move freely, choose where to live, work, or open a business, cultivate land, or access protection and services and their ability to make meaningful choices about their lives.” (UNHCR 2022a).

Also from an international law perspective the curtailment of freedom of movement can be criticized. It goes against article 26 of the 1951 UNHCR Refugee Convention, which reads:

“Each Contracting State shall accord to refugees lawfully in its territory the right to choose their place of residence and to move freely within its territory, subject to any regulations applicable to aliens generally in the same circumstances.” ⁶

In case of mass influx of refugees, governments can justify exemptions to the freedom of movement:

“ [R]estrictions on the right to move freely in the territory, such as into or out of certain areas or camps, should only be imposed when this is clearly in the interests of refugee security or overall national security.” (United Nations General Assembly 2001).

This declaration mirrors an increasingly widespread consensus within the field of refugee studies that propagate abolishing camps (Bakewell 2014; Brankamp 2022). Remarkably, Black pointed out already in 1998 that “justifications for a policy of encampment are rarely made in general terms by policymakers; rather, specific reasons are more often cited to justify its ‘exceptional’ use in particular circumstances” (Black 1998). In a way, this is in line with the above-cited note from the UN’s General Assembly, which stresses that limitations to the freedom of movement can only be imposed in case of specific security concerns.

Refugee Protection Policies in Africa

In 2022, Africa hosted a quarter of the world’s registered refugees, 6.58 million people, and nearly half of all internally displaced persons, 31.7 million people (UNHCR 2022b; IDMC 2023). Most African countries have ratified the UN Refugee Convention, and refugees are often recognized prima facie (i.e., on the basis of their group membership), which legalizes their protection status without long and tedious asylum procedures. In most cases, camps have been the preferred choice for humanitarian organizations and governments to host refugees. Logistical advantages to respond quickly, locally, and efficiently from a humanitarian point of view correlated neatly with political interests to isolate and control incoming outsiders and their relation with the local population. Fearing social, economic, and political tensions, threats to national security and possibly harm to diplomatic relations, several governments in Africa remained reluctant to allow freedom of movement (Dick and Rudolf 2019). Recently there are visible progresses as governments have become less restrictive in allowing refugees to settle outside of camps (Maple 2024). This is often a cost-motivated decision: Running camps is expensive, possibilities for local integration are largely lacking, displacements have become increasingly protracted, chances to find, or create employment diminish.

In sum, encampment policies seem to offer no long-term solutions but create and maintain dependency. Looking at empirical evidence, it is nevertheless striking that refugee studies and policies have a strong focus on protection, livelihoods, or quality of life inside camps. Only more recently analyses comparing camps with self-settlement became available: Outside of Africa, drawing on a quantitative comparative analysis of Syrian refugees in-camp and out-of-camp in Jordan for instance Obi (2021) argues that self-settled refugees are less likely to live below the national poverty line. Drawing on a political economy analysis, Omata (2021) compares refugees’ livelihoods in Nairobi and in Kakuma Camp, Kenya, and argues that apart from type of settlement, institutional factors may play a major role in shaping economic strategies, thus nuancing the impact of settlement type.

The promotion of self-reliance has been an argument to promote self-settlement of displaced people, and was for instance a key driver of Uganda’s out-of-camp policies, where it was argued that refugees are often highly entrepreneurial and use their agency to develop economic initiatives (Betts 2021). If such refugees are free to move, increased opportunities are not only advantageous for refugees themselves, but also strengthen local hosting communities (ibid. 2021). Uganda offered a refugee settlements model where refugees are given small plots of land in settlements located in sparsely populated areas and supposed to act as a catalyst for economic development of the larger area, from which also surrounding communities benefit (Ahimbisibwe 2019; Betts 2021).

More recently, the promotion of resilience has been added to the rationale behind such policies (Oliver and Boyle 2020). In the context of the DRC, Jacobs et al. (2020) show that urban Internally Displaced Persons add value to the market through their access to rural products that urban dwellers lack. Being able to move back and forth between their communities of origin and the city, they are able to mobilize rural resources, not only for their own benefit, but also for the benefit of the urban economy.

If benefits of out-of-camp settlement are clear, why then do camps continue to exist? Most discussions about this issue are rather state-centric, arguing for instance that communities lack the absorptive capacity to accommodate large groups of refugees, or that it is easier to provide protection and basic necessities in camp settings (Bakewell 2014). Empirical evidence also shows that the longer a camp exists, the more it resembles a city, with its own governance structures and dynamics, and with its own economic development (Jansen 2013; McConnachie 2014; Sahldengil and Adugna 2022). At the same time, the longer a camp exists, the more constrained the provision of food and other material aid may become, as intervening humanitarian actors often withdraw or reduce their contributions over time. This may push people out of camps, either permanently or on a flexible basis. Moving back and forth between a camp and urban setting is a proven strategy for people to take benefits from both locations (Rudolf et al. 2015; Krause and Gato 2019).

Ethiopia’s Refugee Policy on Paper

Despite a long history of hosting forcibly displaced people, Ethiopia’s first official displacement policy dates back to 2004 when Proclamation no. 409/2004 was signed as the first statutory asylum and refugee law of Ethiopia. Until then, Ethiopia had neither concrete national policies nor specific refugee legislation. Proclamation No. 409/2004 was enacted based on the international and regional refugee conventions. Ethiopia is signatory to the 1951 Refugee Convention, its 1967 Protocol and the 1969 OAU Refugee Convention, but it maintains reservations against Article 17 (Freedom of movement) and Article 19 (Right to work) of the 1951 Convention. A major point of critique against the Proclamation relates exactly to the restrictions it imposed to the rights to movement and work. The Proclamation assumed that refugees would return to their countries of origin and should be supported in camps until their return, without a need for self-reliance or local integration. The Proclamation prioritized security concerns; Refugee camps were identified to be potential security risks; a strict encampment policy was implemented to protect the own population.

Since 2010, a special OCP has been implemented, which permitted Eritrean refugees (but not refugees from other countries) to reside outside of refugee camps as long as they could support themselves financially or were sponsored by a relative or friend (see discussion below). This usually requires having a guarantor. Generally, we find that critics argue that this OCP has some benefits, but also point out that it is an important drawback that it does not allow formal employment for refugees.

In recent years, Ethiopia’s legal and policy frameworks on refugees and asylum seekers have changed in line with global policy changes. In 2016, the United Nations General Assembly (UNGA) unanimously adopted the New York Declaration for Refugees and Migrants, reaffirming the importance of international refugee rights and committing to strengthening protection and support for people on the move (UNGA, 2016). Based on the request of the UN General Assembly, UNHCR initiated and developed a Comprehensive Refugee Response Framework (CRRF) with four key objectives: (a) ease pressure on host countries, (b) enhance refugee self-reliance, (c) expand access to third-country solutions, and (d) support conditions in countries of origin for return in safety and dignity. Subsequently, in 2018, UNGA affirmed a non-binding Global Compact on Refugees (GCR), incorporating the CRRF and a plan of action (UNGA 2018). In Ethiopia it resulted in a new Refugee Proclamation in 2019.

Ethiopia’s new Refugee Proclamation has been praised as one of the most progressive refugee laws in Africa (UNHCR 2019) and Ethiopia became a CRRF pilot country. ⁷ Like its predecessor (Proclamation 409/2004), it maintains an open-door policy for refugee inflows and allows humanitarian access and protection to those seeking asylum. Moreover, the new Proclamation makes a number of specific changes facilitating local integration: According to article 26, sub-articles 1-10 of Proclamation 1110/2019, refugees can enjoy employment and property rights to promote self-reliance. This enhances possibilities to self-settle outside of camps. The new job compact in the Proclamation provides refugees the right to engage in wage-earning employment, in agriculture, industry, small and micro-scale enterprises, handicrafts, and commerce. The Proclamation also gives refugees and asylum seekers the right to a residence permit once they have formal employment, subject to renewal every five years. They furthermore have the right to open bank accounts, legally register births and marriages, and attend primary school. Last but not least, the new Proclamation has unambiguously outlined a comprehensive set of “rights and obligations” — mostly along lines provided under international instruments.

Due to the growing numbers of Eritreans seeking aid from Ethiopia’s Administration for Refugee and Returnee Affairs (ARRA) ⁸ and a cut of funds by 30 percent in 2020, the OCP was revised in regard to various issues at the beginning of the year: no guarantor is required anymore to reside outside camps; papers for OCP status (practically an equivalent to a residence permit) are to be handed swiftly, provided individuals waive aid claims; and a functioning procedure for working permits was scheduled to be in place by July 2020 before the COVID-19 crisis. ⁹

Ethiopia’s Refugee Policy in Practice

How do these policies and revisions work out in practice? We started this paper with a quote by a refugee residing in a camp. He emphasized feeling imprisoned and felt his life was on hold. Other Eritrean respondents expressed similar feelings, often drawing parallels between life in the camp and in the obligatory military service in Eritrea when their lives were similarly put on hold. Collected during the OCP implementation process, the quotes underline that people in encampment situations desire higher levels of freedom of movement but experience persisting constraints.

Although the large majority of refugees in Ethiopia still reside in camps, there is meanwhile a sizeable group of refugees and asylum seekers living out of camps. Refugees are allowed to live in urban centers on two grounds: Some benefit from the OCP, some qualify for the Urban Assistance Program (UAP). The OCP started in 2010 and has only been used for Eritrean refugees. Eritrean refugees enjoy historical, linguistic, religious, and ethnic commonalities with mainstream highland Ethiopian cultures. Due to their demographic characteristics (until 2018) — mostly young, male, and single — they are supposed to be self-reliant. However, to benefit from the OCP, Eritrean refugees had to either (a) produce evidence of their income, while they could not have any formal employment. In other words, those who take up this opportunity must have access to remittances from family members in the diaspora or have informal sources of income; or (b) present an Ethiopian citizen who serves as their guarantor. The UAP is a program for all refugee groups based on their specific vulnerabilities, such as health and special protection needs that cannot be met in camps. Thus, while the OCP refugees do not receive any support and are expected to be self-reliant, the UAP refugees receive financial assistance from UNHCR (Betts et al. 2019).

Refugees in camps often long for freedom of movement and feel deprived of economic opportunities. These are clearly two major concerns expressed by many camp residents. Significantly, 90 percent of survey respondents living in camps indicated not having worked for pay or profit in the last month (Tufa et al. 2021). Most of them lived on food rations provided by WFP and administered by UNHCR and ARRA. Depending on availability, these monthly rations consist of 10-16 kg of cereal, complemented with oil, sugar, and salt. All informants however indicated that this was often not enough. They regret being unable to complement their food rations with cash income through employment, especially when their stay in camps is protracted. Tadde for instance is staying in Shimelba camp for more than 20 years. He stated:

“Spending 20 years at a place without a job is simply devastating [. . .]. You don’t have the opportunity in the camps to go out and work with the knowledge and skills you have . . . We feel like we are blocked from all opportunities” (interview, November 2019).

Tadde and numerous others residing in the camp have indicated that they are unable to meet their basic daily needs, let alone pursue their aspirations. This resonates with the initial statement made by our respondent at the commencement of this paper that portrayed the camp as a prison. ¹⁰

Frictions Between Intent and Impact

Whereas refugees in camps are affected by “intractable waiting” (Tufa et al. 2021), they at least do not have to worry about shelter. This stands in sharp contrast to urban refugees. Finding proper housing is their major challenge, followed by challenges related to health care and to food. According to our survey 87 percent of urban refugees have to pay rent for their housing at relatively high prices. This means a need for income. Due to a lack of work permits finding a formal job is impossible. Remarkably, 93 percent of urban refugee respondents had not worked for pay or profit. It did not mean however, that they did not work. Most of them actually did, but in informal manners, making them vulnerable to exploitation (Tufa et al. 2021).

Eritreans had profited from the described rather informal exceptions from the encampment policy already starting from 2010. Until the start of the Tigray war in 2020, 28,000 Eritreans (out of 177,000) benefitted from OCP — including those on UAP. While this was significant as it included about 16 percent of Eritrean refugees, it was only 3 percent of the refugee population in the country. Moreover, under OCP, Eritreans cannot get work permits. They therefore register businesses under the name of an Ethiopian, or they are entirely dependent on Ethiopian landlords or employers for informal accommodation or work. This places refugees in vulnerable dependency relations. In Addis Ababa for instance, we met several refugees who had started small businesses, using the licences of friends or distant relatives with Ethiopian citizenship status. They mostly made only informal — and primarily oral — arrangements with these citizens and had to pay for the service. These arrangements often ended in disputes due to their informality and due to diverging expectations.

A good example is Tekeste, a thirty-year-old Eritrean refugee who lives in a large suburban apartment block. He has been in Addis Ababa as an OCP beneficiary since 2013. He survives on the remittances he occasionally receives from cousins living in South Sudan, Switzerland, America, and Canada, but feels uncomfortable about this dependency: “I am dependent on remittances from my cousins. . . . [yet ] until when should I depend on them?” Trying to overcome this dependence nevertheless led him into another dependency due to legal constraints:

A man I met here who has become a good friend got me a business license under his name. I agreed to pay him some money every month. After working smoothly for two years, he increased the monthly payment several times in a year. Then, we disagreed, and I was forced to stop the business. (interview, November 2019).

Starting from the second half of 2019, some of the promises in the new Proclamation ¹¹ were already fulfilled: Refugees could register births and marriages, open bank accounts, and buy mobile SIM cards. However, the right to work and access to residence permits remain mostly abstract because concrete implementation mechanisms are lacking. Several scholars have expressed concern about the practicability of the highly ambitious employment provisions (Betts et al. 2019; Adugna, Deshingkar, and Atnafu et al. 2021). These critics were sceptical about the possibilities for refugees to find jobs in a country strained by unemployment. Employment is still largely found through informal networks, and formal work permits in line with the Proclamation are very limited. Material assistance remains limited to camp settings.

Six years after the adoption of the 2019 Proclamation, the Ethiopian government has expedited the process of providing work and residence permits to refugees and asylum seekers. By mid-2025, over 14,000 refugees have been given work permits. This might have been prompted by the need to support bottom-up inclusion of refugees into the local and national socio-economic development in the face of declining international funding. However, in principle, both the work and resident permits do not include refugees who benefitted from OCP as they are considered self-reliant based on the conditions they accepted.

Capitalizing on Existing Opportunities

Helen is an Eritrean refugee in her twenties. Her story is an example of a refugee who managed to benefit in positive ways from the OCP. Our team met her in Shire, a city in the Tigray region of Ethiopia and had several conversations with her throughout 2019. Helen explored business options which would not require significant seed capital. She thought of making and selling coffee, drawing on her experience in a hotel in Asmara. She picked a location in a byway, which she considered suitable for her business because it was located close to a college and offered shady trees for some chairs and tables: “People enjoy drinking coffee under trees. . . . [yet] they are usually disturbed by the sound of cars. People prefer silence,” she told us during a conversation). She managed to get the permission to sell coffee on the veranda of a guesthouse by assuring her (Ethiopian) landlord that this would increase the clientele of his guesthouse. He became her guarantor and she got a confirmation from city council officials that she was eligible for a business licence. Within three months, Helen had employed a local person and another refugee to help in the store; successfully promoted the guesthouse; had learnt how to produce soap; started selling it; and was about to open an indoors coffee shop next to a game center. Seeing the benefit for the guesthouse and observing how she had changed the desolated street, an Ethiopian businessman proposed a deal to Helen where he provides for the expenses and she for customers and service. She rejoiced: “People always think money solves all problems — but if you work well chances come to you!” The profits were planned to be shared equally; the business partner would cover rent, kitchen equipment, and furniture, while Helen would run the daily business, supervising the staff to be hired. Her first shop would be run by her sister and a befriended refugee.

Helen’s case is a positive example of the empowerment taking place when entrepreneurially oriented refugees get the freedom to settle at their place of preference and are allowed to develop their own business initiative, and in doing so, contribute to urban economic development. In her case, the relation with the guarantor was a win-win relation and she made good use of it. ¹² Helen was not the only refugee we met who used informal opportunities to rebuild her live in an urban setting. Similar cases in fact could already be found prior to the mentioned change of policy.

One of these was Almaz, who joined Hitsat refugee camp in 2015. During her six months in the camp, she earned some income by making tea, preparing food, and handicrafts. Coincidentally, she learned about an Eritrean family flying to Addis Ababa from the United States for a three-month vacation. Longing for new opportunities, she came to Addis Ababa and started serving the family as a housemaid. She was excited when the family promised to take her to America. However, after a while, they became dissatisfied with Almaz because they wanted her to be in the house continuously. Almaz went to ARRA and UNHRC to follow up on her case, but nothing happened. Her Eritrean refugee boyfriend who was already living in Addis Ababa advised her to work at a nightclub whose owner he knew and acted as her guarantor. In the nightclub, she worked as a DJ for US $27 per month, a salary which did not suffice to make a living, so she also worked as a waitress in the same club for eight months. All was arranged in an informal way.

Yet working as a waitress didn’t pay well enough either. Her roommate complained about her small contribution to the household expenses. When her boyfriend left for Australia without telling her, she engaged in sex work, but felt increasingly uncomfortable with it. She quit and trained as a nanny for four months at the Jesuit Refugee Service (JRS), a humanitarian organization. The training was futile as employers would not accept employees with a refugee ID card. Moreover, her health deteriorated. When she went for a check-up, she tested HIV/AIDS positive. Almaz went to a protestant church and began studying the Bible. Her belief helped her to be strong. In church, she met a woman who employed her as a housemaid without a guarantee. Almaz worked in the woman’s house for five months before she joined another training offered to Eritrean refugees by JRS. She got a certificate in hairdressing but could not work in a saloon due to her deteriorating health. Creating her own business was not an option due to a lack of capital and/or work permit. Facing a continuous increase in rent, and high cost of food, Almaz currently only makes ends meet thanks to the help of her friends living in Europe.

Almaz’s case shows that not every refugee meets conditions that are favoring an entrepreneurial spirit such as Helen’s. We also encountered many refugees who did not manage to find an economic partner for a joint business. The road toward self-reliance and economic improvement is often a dead-end road. For those who are unsuccessful, costs for basic services such as housing, electricity, food, health care, and school fees outside the camps were often so high that they faced a downward spiral. They ran into debt and precarity. Out of these precarious cases in urban areas that we observed, the fortunate ones managed to return to the camps where their needs were covered by aid. This shows that out-of-camp options do not necessarily decrease vulnerabilities unless flexibly supported by work permits and or humanitarian aid, thus refuting the dichotomous and mutually exclusive approach to protection and OCP.

Encampment or Out-of-Camp Policy: A “Hard” Ethical Dilemma?

What do our findings tell us about the choice between encampment and OCP as a real dilemma? The bottom-up perspective of our empirical data shows that “faulty” analysis or “misinterpretation” might be inevitable to a certain degree. The closer the analysis gets to everyday realities the more above mentioned cases show that “facts” are dynamic and indeed ever-changing. Taking the dilemma as a prime issue, we see that the perception of facing only undesirable alternatives when receiving refugees depends on the way in which the issue is approached, namely a focus on theoretically anticipated negative impacts. Comparing the theoretical and the practical level of the perceived dilemma, our findings suggest that issues which theoretically seem hard, if not impossible to reconcile, either do not lead to the anticipated problems, are more nuanced in practice, or are even completely solved on the ground. De facto realities may differ from outcomes anticipated by policy makers.

Against the background of the described lived experiences of an OCP, the question whether the choice between encampment or freedom of movement policies constitutes a hard ethical dilemma becomes less easy to answer. Is there a conflict of morally worthy goals that cannot be resolved? The policy choice between encampment (protection, care, and maintenance) and freedom of movement (self-reliance and local integration) impacts not only on refugees, but also on the citizens of the hosting country. What may be most beneficial for one group may not be beneficial for the other. How to weigh the interests of different groups? Governments have a responsibility to protect people seeking refuge on their territory, but they also act in the interest of their citizens. Both for refugees and for hosts, the two options provide morally worthy goals, with different trade-offs. This means it may not always be easy to reconcile both options and that resolving one dilemma might lead to another.

Our cases show how self-reliance can increase and how refugee entrepreneurs can contribute to local economies under an OCP. Helen set up her business in partnership with a member of the host community without government or NGO-support. Her initiative contributed to the creation of employment opportunities for both refugees and members of the host communities and fostered her integration. The case of the urban refugees that managed to settle out of camp showed that, rather than creating new opportunities, the OCP expanded the already existing arrangements and reduced some of the vulnerabilities when out of camp settlement was an option for a small group only. When encampment was the rule for most, refugees outside camps engaged mostly in informal work, heightening abuses and exploitation, and creating social tensions or economic competition between locals and refugees. Encampment policies, in other words, did not avoid but fostered certain perceived risks. The new Ethiopian Proclamation — with its work and residence permits — has not yet corrected these shortfalls. The vulnerability of refugees persists. Impact hence does still not match the intent.

Both OCP and encampment have their own pros and cons. Analyzing this co-existence enables us to see more nuanced interrelations and interdependencies. Numerous studies criticize the prison-like conditions in camps (Brankamp 2022). Yet our cases showed that many refugees appreciate camps for providing basic services, which they cannot access elsewhere. For many, camps have become homes through personal connections based on socio-cultural communalities with the host communities. Camps are not necessarily run and experienced as prisons, even if national policies favor such conditions (ibid.): The case of Shimelba camp we mentioned above underlines this: Camp residents enjoy a high level of local mobility and supplement their livelihoods through sharecropping with the host community.

Conclusion and Policy Recommendations

In this paper we have taken an empirical approach to gain an understanding of how the policy dilemma between encampment and freedom of movement/OCP works out in practice. As a case study for such policies we looked into the case of Eritrean refugees in Ethiopia, where both policy approaches exist alongside each other, but as mutually exclusive alternatives. Paying attention to the policy development and dominant policy discourses, our empirical description started from an “oblique approach” (Bakewell 2008), in which we looked into what people actually do and how they feel about the options at hand. Rather than differentiating politics, policy analyses, and policymaking, we lumped them together and focused more on unpacking the subjects of the policies on the ground. We agree with the need for a detached analysis, focusing on actors affected by policies, as suggested by Bauböck, Permoser, and Ruhs 2022, 436), yet propose a slightly different approach. We turned the pledge to take “a step back from the field and entering again the realm of the detached observer” in order to cross “the bridge between empirical analysis and theoretical conceptualization” upside down — and combine an analysis of policy with an analysis of empirical realities. In general, we argue that focusing only on the preconceived objective when determining the way forward is detrimental to a nuanced understanding of policy impacts. Instead, we recommend that policymakers worldwide assess the actual impacts of their policies on the ground in an evidence-based and unbiased manner. For this, they need to consider both refugees, host communities, and their surroundings, in their assessments.

Top-down discussions on desirable policies often disregard everyday outcomes that are unconsciously or unintendedly addressed in policies — for example, less vulnerability and heightened resilience — yet which are realized by people themselves (Rudolf 2022). Such outcomes were achieved independent of, or even despite, policies seeking these very outcomes. Existing practices sometimes catch the attention of policymakers only because of the lack of alternatives: It was only due to overwhelming lack of solutions and ever diminishing resources combined with the prospect of rising expenses — caused both by the number of new refugees and demographic growth of existing refugee populations — that existing de-facto alternatives drew the attention of policymakers, not the other way round (ibid.).

In addition to pointing at the risks of faulty misinterpretation of facts, we hold that the top-down conceptualization of facts constitutes a practice in which existing differentiations and alternatives will remain overlooked: Refugees, like other people, exert agency and take decisions regardless of policy frameworks. Policies, however, can encourage or discourage some options. Various studies suggest that the above described informal practices also exist elsewhere (Rudolf et al. 2015; Sanyal 2017; Krause and Gato 2019). Also in these places current policies might have placed people in a vulnerable in-between position, created by the policy that intended to prevent or overcome this vulnerability.

In the past, both refugees and policymakers in Ethiopia chose the OCP not because it was the best option, but because it was the only alternative to encampment. Even then, there were overlaps of camp and out-of-camp realities. Freedom of movement, access to services, markets, shelter, or protection were not as rigidly blocked or guaranteed in practice as in theory. Policy on paper did not neatly correspond with policy in practice. Governments do not have full control over policy implementation, and people find ways to alter practices of policies in unintended ways or circumvent policies altogether.

The Ethiopian government is currently (summer 2025) accelerating the implementation of the provisions of the 2019 Proclamation by providing the refugees with the right to work and residence permit. RRS and the Ministry of Labor and Skills thus have the opportunity to increase the practical impact of their policies and can rectify the previous faulty dichotomies between encampment and OCP. Our evidence shows that it is detrimental for Ethiopia’s OCP to force beneficiaries to give up any institutional support they receive, while not providing work permits. The current practice contributes to unemployment and dependency in a country where around one in six citizens depend on humanitarian assistance. We therefore recommend that the Ethiopian Ministry of Labor and Skills allows OCP beneficiaries to engage in business or seek employment opportunities and we recommend the Ethiopian RRS to consider allowing refugees to settle in their preferred locations, in accordance with the provisions of the 2019 Refugee Proclamation. This would avoid the negative effects of joining the informal economy and competing with Ethiopian citizens.

Our findings suggest that the assessment of hard policy dilemmas should put a greater emphasis on impact than intent. Dilemmas in policy-making are as much related to the values behind the policy as to their expected implementation. Encampment is perceived as detrimental to values such as the freedom of movement, economic autonomy, or social integration of the refugees. Camps are, thus, portrayed and assessed as open-air prisons. OCP is perceived as detrimental to values such as the right to economic, political, and social security of the host population. Refugees then, are, portrayed as merely a burden and a risk. Our examples show a more differentiated picture in which both in-camp and out-of-camp conditions bear risks and benefits, both for refugees, host communities, and governments. Camps provide access to basic services, health care, education, and a level of protection which is appreciated by refugees. Hosting refugees can be advantageous for the labor market, political recognition, and security, due to the presence of international actors in the region. Benefits are nevertheless associated with downsides such as labor exploitation, human rights violations, economic precarity. The alternatives are hence neither clear-cut dilemmas nor straightforward win-win situations.

Our analysis clearly indicates that the argument by Bauböck, Permoser, and Ruhs (2022) on fact-based approaches needs to be extended, firstly by a discussion of the impact of hermeneutics on asymmetrically shaped policy realities. For this, policymakers need to be more self-reflective when shaping realities. Secondly, by a reinterpretation of false versus hard dilemma that leaves room for a diachronic assessment of the very dynamics caused by policies derived from such models. In other words, our analysis indicates that hardness of dilemma — just as the other elements believed to be constitutive for a policy — might change over time with the answers actors found on the ground: Changes that again can only be identified by a continuous empirical feed-back methodology.

The assessment of dilemmas depends on whose perspective is taken. It has been the aim of this paper to provide glimpses into the lived realities of policy options and their impact; realities which may not be in sight of policymakers. We have shown that taking aspects such as livelihoods, security, basic services, and humanitarian assistance into consideration changes the assessments of the dilemmas identified. According to the cases described, policies did not always have the intended impact, sometimes on the contrary. Not focusing on equally undesirable options, we saw that dilemmas have a flip-side. An assessment of de facto practices rather than anticipated outcomes diluted or even dissolved theoretically hard dilemmas. Rather than mutually exclusive or strict alternatives, the two sides of the dilemma can be complementary options. This is an important lesson for Ethiopia’s RRS and for refugee policy makers more generally: Giving refugees more freedom to follow their preferences and building on existing practices can help to unleash refugees’ potential and to integrate better into the host society. Encampment and out-of-camp policies should hence not be treated as two mutually exclusive alternatives. A bottom-up policy could combine impacts of OCP and encampment policies to promote the best of the two worlds: