Refugee livelihoods: a comparative analysis of Nairobi and Kakuma Camp in Kenya

The quantity of literature on the livelihoods of refugees is increasing, yet there is a lack of comparative analysis of their strategies in this regard vis-à-vis their host communities and living locations. Drawing on fieldwork conducted in the city of Nairobi and Kakuma Camp in Kenya in 2016 – 17 , this paper narrows the gap by presenting an examination of the economic activities in which refugees in different contexts engage, and highlighting some of the institutional factors that distinguish refugees’ economic lives from those of members of the host community. The findings suggest that while some Kenyans are encumbered by challenges comparable to those faced by refugees, the myriad political, legal, and policy factors that characterise refugeehood are none-theless particular in the ways that they influence livelihoods. By emphasising the multiple actors and aspects that shape refugees’ economic strategies, this paper also reveals the political economy within which refugee livelihoods are embedded. of on multi-sited fieldwork the livelihoods of refugees living in a camp and an urban area, with attention paid to differences across these contexts and between refugees and their receiving national populations. Furthermore, as a theoretical contribution, it describes a ‘political economy’ approach to refugee livelihoods, which are shaped by complex interactions of multiple actors and refugees’ agency and assets.


Introduction
Forced displacement has become one of the most significant challenges confronting the world. According to the latest statistics of the United Nations Refugee Agency (UNHCR), the size of the refugee population across the globe was a record 25.9 million at the end of 2018 (UNHCR, 2019). In addition to the staggering number of refugees, the duration of their exile has also been prolonged. The vast majority in the Global South remain in protracted displacement long after the initial emergency phase of a crisis has ended. The average time period is now estimated to be more than two decades (UNHCR, 2016).
However, the volume of international assistance for long-term refugees is inadequate given dwindling interest among donor countries in supporting populations living in a situation of protracted displacement (Aleinikoff, 2015;Omata, 2017). The daunting scale of extended cases, coupled with the decreasing availability of relief aid, has made it vital that the global refugee regime understand better refugees' economic activities and seek ways to enhance their integration into countries of first asylum, particularly by ensuring access to livelihoods and economic opportunities (Milner, 2014;Jacobsen and Fratzke, 2016).
While the quantity of literature on refugee livelihoods is increasing, the extent and volume of such research is still small and not synthesised as compared to other issues in the field of refugee studies. In particular, there is a significant lack of comparative empirical analysis of refugees' livelihoods in relation to their host communities and their locations of exile, including camps and urban settings. Drawing on multi-sited fieldwork in Kenya, this paper explores therefore the livelihoods of refugees living in a camp and an urban area, with attention paid to differences across these contexts and between refugees and their receiving national populations. Furthermore, as a theoretical contribution, it describes a 'political economy' approach to refugee livelihoods, which are shaped by complex interactions of multiple actors and factors as well as refugees' agency and assets.
Kenya has been hosting refugees for some three decades. As of the end of 2018, with more than 400,000 residing within its territory, it is the thirteenth largest refugeehosting country in the world (UNHCR, 2019). Kenya has also developed an unflattering reputation for enforcing draconian refugee policies. Milner (2009) characterised the country's refugee policy with the words 'abdication' and 'containment': the abdication of responsibility for refugee protection to UNHCR and the containment of refugees on the periphery of the state.
A number of researchers conducting fieldwork in Kenya have produced seminal works on refugees' economic strategies (see, for example, Jaji, 2009Jaji, , 2014Pavanello, Elhawary, and Pantuliano, 2010;Campbell, Crisp, and Kiragu, 2011;Lindley, 2011;Oka, 2011Oka, , 2014Newhouse, 2015;Jansen, 2016;Tippens, 2017;O'Callaghan and Sturge, 2018). Notably, the World Bank commissioned a large-scale study in 2016 to assess the long-term impacts of hosting refugees on local populations (Sanghi, Onder, and Vemuru, 2016). This paper contributes to this pool by contrasting the ramifications of 'refugeehood' on the livelihoods of different groups of refugees attempting to make a living in a camp (Kakuma) and an urban environment (Nairobi). By highlighting the factors that influence their economic strategies and outcomes, it also points up programmatic implications for policymakers and aid agencies working in the area of refugee livelihoods.
The data for this submission is based on a total of seven months of fieldwork in Kakuma Camp and Nairobi between 2016 and 2017. Qualitative methods were employed, comprising mainly semi-structured interviews, focus group discussions, and non-participant observations. The research centred on the most prevalent nationalities of refugees in each site, namely Congolese and Somali refugees in Nairobi, and Congolese, Somali, and South Sudanese refugees in Kakuma Camp. To select refugee interviewees, the study relied on stratified snowball sampling based on country of origin. In addition to 74 refugee respondents, 32 non-refugee stakeholders were interviewed across two research sites, including officials of the Government of Kenya, UNHCR staff and members of partner agencies, and individuals from host communities. To protect their anonymity and to avoid any potential negative consequences from involvement in this research, anonymised names were applied to all refugee interviewees. This paper is composed of seven sections. The next section provides an overview of current scholarship on refugees' economic strategies and identifies analytical gaps in the literature. The third section describes the contexts in which refugees make their lives in Nairobi and Kakuma Camp. The fourth and fifth sections present the main findings regarding the economic activities of refugees from different nationalities, in relation to those of local host populations in these two sites. The final two sections flag some research implications of this inquiry and the need for a political economy lens on refugee livelihoods.
Ample research describes how refugees engage in economic activities and participate in labour markets in their host states, demonstrating the way in which their livelihoods are shaped by various aspects of 'refugeehood'. Broadly understood as the legal, social, political, and/or institutional contexts in which refugees find themselves in exile (Kibreab, 2003;Betts et al., 2016), refugeehood is marked by a deviation from the 'normal' model of state-citizenship relations (Haddad, 2008). Pushed into the gaps of a global system of sovereign territories, refugees pursue their livelihoods in ways that are institutionally distinctive from those of nationals. Whereas existing scholarship has produced important knowledge of key aspects of refugees' economic activities, there are no overarching conceptual frameworks for understanding their economic lives; this remains a lacuna in 2021 (see also Hammar, 2014;Jacobsen, 2014;Betts et al., 2016).
Nevertheless, the existing literature points to three key avenues to explore the economic lives of refugees vis-à-vis their host populations. First, having been excluded from state protection, refugees in exile hold a particular status and position, partly governed by an asylum country and partly by international institutions, primarily UNHCR. This means that the economic lives of refugees occur in a distinctive institutional context regulated by the refugee agency of the United Nations (UN) and the host government, which is often framed as a 'double chain of administration' (Colson, 2004, p. 108).
Second, as non-citizens of the host country, refugees suffer due to the legal and administrative constraints associated with their refugee status (or even because of an inability to obtain formal refugee status). According to the UN's Convention Relating to the Status of Refugees (1951 Refugee Convention), refugees must be accorded the same status as nationals with regard to the right to engage in wageearning employment. Typically, however, their right to work is significantly hampered by various regulatory impediments imposed by the host government, including a lack of access to work permits, proscriptions against formal employment, and restrictions on freedom of movement (Werker, 2007;Kibreab, 2012;Jacobsen and Fratzke, 2016;Janmyr, 2016;Zetter and Ruaudel, 2016;Baban, Ilcan, and Rygiel, 2017;Brown et al., 2018). As well as formal regulations, refugees' access to economic resources, such as land, forests, and rivers, is frequently restricted by informal regulations imposed by local host populations (Bascom, 1993;Bakewell, 2014;Omata, 2017;Betts et al., 2018).
Third, refugees' economic strategies are regularly shaped by the political economy of asylum. As widely discussed, the securitisation of refugees by host states not only limits their access to labour markets (Fábos and Kibreab, 2007;Lee, 2014;Turner, 2015), but also it fosters negative perceptions of refugees among host populations (Obi and Crisp, 2000;Horst, 2006;Crisp, Morris, and Refstie, 2012). These viewpoints can feed into discrimination, harassment, or even open extortion and unwarranted detention by the police ( Jacobsen, 2006;Crisp, Morris, and Refstie, 2012;Betts et al., 2018).
Notwithstanding these general constraints emanating from refugeehood, empirical evidence also shows that different groups of refugees navigate obstacles with varying degrees of success. For instance, refugees' social networks and ethnic bonds often facilitate their access to formal and informal labour markets and wider economic opportunities (Monsutti, 2005;Andrews, 2006;Porter et al., 2008;Van Hear, 2012;Palmgren, 2014;Brown et al., 2018). In addition, 'mediating factors' (Zetter and Ruaudel, 2016)-that is, a common language, similar culture, or contacts with local host populations-can facilitate refugees' access to work (see also Carrier, 2016). In some cases, these lines of belonging lead local authorities to loosen the implementation of restrictive policies by 'turning a blind eye' to certain economic activities of refugees.
As the existing literature indicates, the livelihoods of refugees are constrained, mediated, and at times enabled by a multitude of actors and factors inherent to the particular environments in which they find themselves. Yet, owing to the relative scarcity of comparative studies centred on different localities and nationalities of refugees, little is known regarding to what extent and how the identified factors of refugeehood influence the economic strategies of refugees across different living environments. For instance, as Jacobsen (2006) and Crisp, Morris, and Refstie (2012) note, the impacts and nature of refugeehood can be differentiated, depending on the location of refugees. While camp-based refugees normally face regulatory constraints, such as restrictions on their freedom of movement, refugees outside of camps can encounter different vulnerabilities, including xenophobic attitudes and harassment, arbitrary arrest and detention, and exploitation by employers. By juxtaposing findings from Kakuma Camp and the city of Nairobi, the empirical sections that follow illustrate the ways in which 'being a refugee' places them in a particular economic position in relation to the host state and the host populations in different settings.

The research context: refugees in Kenya
Kenya ratified all of the key international treaties on refugee protection decades ago, including the 1951 Refugee Convention, the 1967 Protocol Relating to the Status of Refugees, and the Organisation of African Unity Convention Governing the Specific Aspects of Refugee Problems in Africa (1969 Refugee Convention). Although the Government of Kenya made no reservations in regard to these accords, it did not develop national legislation in line with them until 2007, when the Refugees Act 2006 entered into force (O'Callaghan and Sturge, 2018).
Kenya has often been portrayed as a 'tough country for refugees' because of restrictions on their right to work and to move freely. Until the end of the 1980s, however, refugees and asylum-seekers had the right to move freely and to settle anywhere in the country, and they had access to work in the formal sector with employment permits issued by the central government (Wagacha and Guiney, 2008).
Kenya's refugee policy became increasingly restrictive from 1990, though, owing to the sharp increase in the number of refugees arriving from adjacent countries. Since the early 1990s, the host administration has required refugees to reside in the Dadaab or Kakuma Camps, which are located in remote, underdeveloped, and insecure areas. Movement out of them has been constrained; refugee residents wishing to travel have been required to obtain a 'movement pass'-a written permission issued by camp authorities (Campbell, Crisp, and Kiragu, 2011;Betts, 2013).
Kenya's refugee policy has been inseparable from the issue of national security. Refugees and asylum-seekers have continually been perceived as a security threat to Kenya from the 1990s to the present day (Refugee Consortium of Kenya, 2015). Over the past several years, refugees and asylum-seekers have been connected with terrorism; Somali refugees, in particular, are victimised because of securitisation processes. A string of attacks by al-Shabaab, such as the Westgate shopping mall siege in Nairobi on 21 September 2013 and the massacre of 148 students at Garissa University in the northeast on 2 April 2015, has led to growing intolerance and even hostility towards Somali communities, including refugees (ReDSS, 2015). The Government of Kenya declared immediately after the campus attack that if UNHCR did not close Dadaab Camp (situated about 100 kilometres from Garissa) within three months, it would do so (BBC, 2015).
Alongside the rise in negative perceptions of (Somali) refugees in recent years, the Government of Kenya has been reinforcing its encampment policy. The Ministry of Interior and Coordination of National Government stated in December 2012 that all refugees and asylum-seekers in the capital, Nairobi, should move to the Dadaab and Kakuma Camps or be forcibly relocated to them (Human Rights Watch, 2013;ReDSS, 2015). The central administration again cited security concerns in March 2014 when it gazetted the camps as the exclusive refugee-hosting areas (Refugee Consortium of Kenya, 2015;Tippens, 2017).
The Government of Kenya made an abrupt announcement in May 2016 that the refugee camps in the country would close within six months due to insecurity. This statement was delivered alongside an announcement that the Department of Refugee Affairs, the central agency responsible for the implementation of the country's refugee policy, would be disbanded and replaced by the Refugee Affairs Secretariat, mainly composed of staff members from security and intelligence departments of government (Hargrave, Pantuliano, with Idris, 2016;O'Callaghan and Sturge, 2018). The Government of Kenya later removed Kakuma Camp from the closure agenda, but it maintained its intention to shut down Dadaab Camp. The High Court of Kenya blocked the move in February 2017, describing it as 'illegal discriminatory and therefore unconstitutional' (BBC, 2017).
This link between insecurity and the hosting of refugees has exacerbated public perceptions of Somali refugees, especially in urban areas. Refugees are often perceived to rely on money acquired illicitly through Somali pirates and al-Shabaab to finance their stay in Kenya (Refugee Consortium of Kenya, 2015). This has led to outright aggression towards refugees in some areas; hostility driven not by facts, but rather by fearmongering, rumours, and stereotyping. The political climate has also limited refugees' access to justice in the country (O'Callaghan and Sturge, 2018).
At the outset of this study in 2016, Kenya was hosting approximately 490,000 refugees in three locations: Dadaab Camp; Kakuma Camp; and the city of Nairobi-the latter two settings were selected as research sites because of insecurity in the former. Kakuma Camp was set up in 1991 in Turkana County, the homeland of the Turkana ethnic group. Given its proximity to South Sudan, more than one-half of the roughly 200,000 camp residents are South Sudanese. Refugees from Somalia constitute the second largest sub-population, followed by those from the Democratic Republic of the Congo (DRC), Ethiopia, and Burundi.
Despite the reinforcement of the encampment policy on a number of occasions in recent years, Nairobi plays host to more than 65,000 refugees of various nationalities. It is no secret that in excess of 30,000 Somali refugees reside in the neighbourhood of Eastleigh, which is widely known as a predominantly Somali section of the city. Refugees from the DRC trail Somali refugees in number, with about 18,000 living in the capital as of 2016. The number of Congolese refugees in the city has been on an upward trend in the past few years, nearly tripling from some 6,500 in 2012, according to the statistics of UNHCR (2017).
The following sections present findings concerning the economic livelihoods of refugees from Somalia, the DRC, and South Sudan in these two settings.

Impacts of refugeehood on livelihoods in Kakuma Camp
In this long-standing site of displacement, the camp economy has become a 'normal' part of the local landscape (Oka 2011(Oka , 2014Jansen, 2016). While farming is limited due to the arid climate and poor soil quality, refugees engage in a range of commercial activities, giving the camp an 'urban' quality. Nonetheless, as described below, refugeehood continues to shape the livelihood options of camp residents.

Incentive workers
Refugee camps are created by a host government and the international humanitarian regime as 'exceptional spaces' to contain refugees, who are viewed as aberrations of the 'national order of things' (Malkki, 1995). According to Kenyan law, refugees must obtain a work permit to access formal employment. Within the exceptional space of Kakuma Camp, however, they are allowed to work for UNHCR and its partner agencies without one. Although organisations such as UNHCR enjoy formal status in Kenya and refugees undertake full-time 'official work' defined by terms of reference composed by these agencies, refugees are engaged on a voluntary basis and are paid 'incentives'-not a salary-which are significantly lower than the wage received by Kenyans for equivalent work (Horn, 2010;Betts, Omata, and Sterck, 2018).
A series of focus group discussions with Congolese and South Sudanese refugees revealed that one of the most important livelihoods of refugees is work for UNHCR and its implementing partners. Steven is a South Sudanese refugee who works as a community mobiliser at a UNHCR partner non-governmental organisation (NGO). He has been in that role since 2015 and responds to community complaints about aid distribution. He works six days a week from 08:00 to 15:00, and earns KES 5,500 (approximately USD 55) per month, whereas his Kenyan colleagues who are doing the exact same work make KES 45,000 (about USD 450). Steven explained how difficult it is for his household to survive on his incentive payments: With my current salary, I am not able to afford basic items such as a mattress and food. But I cannot find any better job in Kakuma, so I have no choice.
This example illustrates how refugeehood places camp residents in a 'grey zone' ( Jansen, 2016) within a dual governance system, consisting of international and national administrations.

Taxation on camp businesses
While access to full, formal employment is virtually denied in the camp owing to the national refugee law, the local government in Kakuma tries to formalise businesses. In the commercial areas of Kakuma Camp, there are many outlets providing a wide range of items and services, including banking, carpentry, clothing, electrical equipment, groceries, and medicines, as well as bars, internet cafes, restaurants, and salons (see also Oka, 2014). These businesses used to operate on an informal basis without the official recognition of and registration by the state government. Since around 2014, though, the County Government of Turkana has been requesting that refugee enterprises inside the camp pay a fee for a business licence and register their 'informal' shops with the local authority. Yousif, a Somali refugee who runs a retail shop inside Kakuma, was perplexed as to why such a 'tax' has to be paid to the authorities: In an anonymous interview, a senior Kenyan staff member of a UNHCR implementing partner confirmed the exploitative arrangement: Since around 2014, the local government started collecting taxes from refugee business owners in the camp. Their logic is that Kakuma Camp is located within the influence of the County Government, so refugees there must follow the same tax regulations as Kenyan nationals.
He added that the registration fees provide revenues to the local government in Kakuma. This is why officials attempt to formalise refugee enterprises through the payment of registration fees, even inside the exceptional space of the camp.
The examples of incentive workers and the collection of business tax inside the camp indicate the inconsistent enforcement of regulations by national and local government bodies.

Local restrictions on refugee livelihoods
Refugees in Kakuma face an additional set of exclusionary regulations: they are not allowed to keep livestock such as camels, cattle, and goats. Furthermore, they are prohibited from cutting down trees for charcoal production.
These local regulations are part of a legacy of previous conflict between refugees and Turkana host communities. In the early years of the camp, the absence of restrictions on collecting natural resources led to the exploitation of these resources by refugees and put considerable pressure on the forage needed by Turkana pastoralists (Newhouse, 2015). These tensions generated various violent clashes between them.
The local restrictions on refugees were initially introduced as a mitigation measure in response to the conflicts, yet, currently, they provide alternative economic opportunities for indigenous populations. Kakuma Camp is in a historically marginalised area, which is largely devoid of any major investment or development activities by the national government or private enterprises (Crisp, 2000;Ohta, 2005). Turkana people have traditionally relied on pastoralist livelihoods, but their nomadic lifestyles are in transition. A representative of one UNHCR implementing partner, who has been working in Kakuma for several years, explained: Most Turkana people are poor. They used to have a lot of livestock but with bad climate, raising animals is getting much harder now. But they do not have alternative livelihoods. Most of them are poorly educated and many are illiterate.
The local rules imposed on refugees have in turn generated markets and economic opportunities for Turkana residents who have established a monopoly on livestock, firewood, and charcoal markets fuelled by the presence of refugees. Many locals sell firewood, charcoal, and building materials in the camp, but refugees are prohibited from doing so.
Put differently, the state of refugeehood spawns an interdependent relationship between locals and camp inhabitants in this historically marginalised and poor region (see also Jansen, 2016;Sanghi, Onder, and Vemuru, 2016). The County Government of Turkana has provided land for refugees on which to reside, and in exchange, this area has been enriched by the new demand and markets created by the presence of some 200,000 refugees.

Encampment, mobility, and networks
Refugees frequently point to the restrictions on their right to move freely outside of the camp as a major obstacle. In one focus group discussion with Congolese refugees, they depicted Kakuma as a 'half prison'. A senior officer with one of UNHCR's partner NGOs noted how refugees in Kakuma are affected by state regulations:

Refugees cannot move around freely due to the encampment policy. Refugees can go to Kakuma town and other nearby areas. But going beyond these areas requires a movement pass [ from the Government of Kenya].
Owing to the encampment policy of the Government of Kenya, refugees are not allowed to sell their products and services outside of the camp. Some UNHCR partner agencies provide technical skills training programmes for refugees. However, there are very few places in which participants can utilise the vocational techniques acquired. In the words of a representative of one UNHCR partner: In the current conditions, no matter how much livelihood support we provide for refugees, success will be very limited because refugees have no place to use their learned skills.
This NGO teaches refugee women how to weave textiles and to craft handbags. The products produced by the trainees were very sophisticated and marketable, yet they were piled up in boxes for several months because the organisation cannot furnish them with market access.
Mobility without Kakuma Camp is severely constrained, but it is not isolated from external economies. There is frequent trade between the camp and nearby Kakuma town, to which refugees can walk freely. Some refugee business owners purchase their merchandise directly from Kenyan counterparts in other major commercial centres, such as Kitale and Nairobi, and there is a commercial bus service-the Dayah Bus Company, owned by a Kenyan business person-between Kakuma and Nairobi. Many items including clothing, cosmetics, medicines, and stationery, enter the camp thanks to this transportation service.
In particular, several Somali shops in Kakuma Camp are better stocked and organised than those owned by people of other nationalities, largely because of extensive There are many business transactions between Kakuma Camp and Nairobi, especially among Somali business operators. Owing to restrictions on refugees' mobility and difficulties in obtaining permission to travel, most refugee shop owners rely on Kenyan brokers to purchase their merchandise.
Some refugee households send family members outside the camp to seek gainful economic opportunities, even without permission to travel from the Government of Kenya. This is a common strategy among Somali refugees in particular. Some of them connect with the larger Somali-Kenyan society in Kitale, Mombasa, and Nairobi, and try to find work there as porters, restaurant staff, and shopkeepers in order to send money back to Kakuma. This split-family strategy is an example of how certain groups of refugees capitalise on their networks and mobility to negotiate constraints on their livelihoods.

Harassment and policy change
The Kenya Police Service is another institutional actor that shapes refugees' livelihood activities in the camp and the city (Newhouse, 2015). Refugees in Kakuma regularly experience harassment by police, even inside the camp. According to refugees, some police officers target 'unauthorised' businesses run by refugees, such as breweries, clinics, and pharmacies, as well as motorbike taxi drivers (locally called boda boda), all of which are formally required to acquire specific licences from the authorities. Many lack such certification, and so refugee business owners must either pay unofficial charges and bribes or face the closure of their enterprise.
During one focus group discussion with Congolese refugees in Kakuma Camp, a participant in his late twenties expressed frustration when asked by the author about the behaviour of the police. He came to Kakuma in 2013 with his wife and child, and now works as a boda boda driver. He underscored that police systematically collect 'toll fees' from motorbike drivers operating within the camp: There are two specific stopping points in the camp. Every day, we have to pay KES 50 [about USD 0.5]. If we pay once, they will not stop us again in the same day. . . . If we resist the police or refuse to pay, they will take our motorbike and keep it at the police station. The police will then give us a 'fine' of KES 10,000 [around USD 100].
This contrasted starkly with the way in which members of the local host community were treated by police. During the research process, tens of Turkana people were asked whether they had experienced any police harassment or received requests for bribes. The response was almost always 'never'. The systematic targeting of refugees for bribes imposes a significant handicap on their economic endeavours.
In addition to these longstanding challenges, the Government of Kenya's sudden announcement in May 2016 of the closure of the Dadaab and Kakuma Camps, owing to deteriorating security and environmental issues, has affected refugees' economic planning. Kakuma-based refugees routinely reported that they were 'shocked' by this declaration, and subsequently were fearful. Although the authorities later exempted Kakuma from the plan, the psychological impact of this intention lingered among many residents, especially Somali refugees. The following comment by a Somali interviewee is demonstrative: Still we are anxious about our future. After they close Dadaab, they may come back to Kakuma and close the camp here. . . . We know their target is Somali refugees, not other refugee groups. . . . The Kenyan government wants to get rid of us in this country.
Numerous Somali refugees stated during a focus group discussion that they have lost confidence in the government and remain concerned that it will announce again the closure of Kakuma. Some Somali refugee business owners in the camp said that they have halted further investment in their enterprises and are considering moving to Uganda, since action by the host administration may force them to relocate anyhow in the future.

Impacts of refugeehood on livelihoods in Nairobi
Nairobi is home to approximately 15 per cent of Kenya's refugee population, hosting more than 65,000 refugees. Despite limited access to international assistance, they reside in the capital to pursue the socioeconomic opportunities available there. These self-settled refugees navigate a city in which multiple stakeholders exert influence.

Making a living under multiple layers of authority
Although official policy requires that refugees live in either of the two camps, enforcement at the local or community level in Nairobi has been relaxed. 1 When a local government official in Kayole (an area with a high concentration of Congolese refugees) was asked about the encampment policy of the national government, he replied: Yes, we know the Kenyan national government has a tough policy against refugees, but at a community level, we are accommodating them. . . . I think most local governments do not really pay attention to the legal status [of refugees living] in their areas.
A quintessential example of refugeehood is evident in the fact that the national government's regulations have largely denied refugees access to the formal employment sector. The Refugees Act 2006 stipulates that they are allowed to work in Kenya and should be able to apply for formal employment with a 'Class M' work permit. In reality, however, these passes are rarely provided, and refugees are virtually excluded from the formal labour market. According to a staff member of the Refugee Consortium of Kenya, a well-known supporting agency in Nairobi: The Class M work permit is intended for refugees seeking formal employment. The Ministry of Interior is in charge of this work permit. . . . But refugees are requested to submit many other documents to get this permit and in reality, very few permits are given to refugees.
Although work permits are difficult to acquire, many refugees are able to obtain business licences in the capital and embark on their own enterprises. During fieldwork, I interviewed a number of refugees who were granted business licences by the City Council of Nairobi. Among them was Anne-Marie, a Congolese refugee who has run a hair salon since 2015. When asked whether it is officially registered or not, she responded: Yes. I paid KES 9,500 [approximately USD 95] for a business licence. In addition,I paid KES 5,000 [about USD 50] and KES 4,500 [around USD 45] for trade and fire licences.
A legal officer with Kituo Cha Sheria, a legal aid and human rights NGO in Nairobi, explained the contradiction between the Government of Kenya and the City Council of Nairobi with respect to refugees' right to work:

Provision of business licences falls under the authority of Nairobi City
Council. This is a completely separate process from work permits issued by the Ministry of Interior. . . . Fees from business licences are a part of the City Council's revenue so they do not exclude refugees as long as they pay it. In a way, the Council does not discriminate between refugees and locals.
Similar to Kakuma Camp, there are differences between the national and local governments with respect to access by refugees to the formal economic sector. According to a UNHCR senior officer in Nairobi, Kenya's refugee policy is characterised by 'multiple standards' and it does not always reflect reality. Whereas these inconsistencies complicate the situation on the ground, they also create the space for refugees' livelihood activities.

Money for police is part of our daily budget
While national refugee policy is somehow diluted by the local authority in Nairobi, refugees still face numerous challenges in making a living in the capital. The most notable is undoubtedly abuse by public authorities, especially police officers (see also Campbell, 2006;Pavanello, Elhawary, and Pantuliano, 2010;Lindley, 2011;Kibreab, 2012). Bribery, stemming from persistent corruption throughout the county, is a common trait of Kenya's informal sector (Hope, 2014). With precarious legal status, refugees in Nairobi confront increased risk of harassment and extortion by the police as compared to members of the host population.
Somali refugees in particular have been targeted by the police in Nairobi. Many are identifiable owing to their features and limited command of the Swahili language, whereas Congolese refugees tend to blend in more. Furthermore, Somali refugees have been perceived locally and regionally as targets in the fight against terrorism, stemming from suspicion about Somali links with al-Shabaab ( Jaji, 2014). Kenyan nationals in Nairobi, though, claim that the increased targeting of Somali refugees by the police is not necessarily attributable to securitisation. A local Kenyan government official who has lived in Eastleigh for decades pointed out that:

Police see Somali refugees as 'a source of money'. In Nairobi, police harassment toward Somali refugees is deliberate and systematic [owing to their limited access to formal justice mechanisms].
Giving cash to the police in Nairobi has become so routine for many refugees that they include the cost in their daily budget. For instance, Najat, a Somali woman who runs a retail shop in Eastleigh, is registered in the city as a refugee and her business is officially registered with the City Council, yet she constantly experiences police harassment: Limited access to the justice system is one of the major differences between refugees and local Kenyan hosts. Ahmed, a Kenyan shop owner of Somali origin, described in an interview how Somali refugees suffer in Nairobi because of their status: In recent years, unfortunately, Somali-Kenyans are often targeted by Kenyan police and government officials. But Somali refugees suffer more than we do. We are essentially Kenyan nationals. We can speak the local language and have national ID [identification] cards, which usually resolve cases of potential harassment.
Paying bribes is an essential survival strategy of refugees in Nairobi, if they are to avoid arrest and imprisonment. The practice, aside from its emotional toll, increases the cost of goods and services provided by refugee entrepreneurs, making their livelihoods less profitable and more precarious.
As of 2017, refugees from Somalia and the DRC constitute the two largest refugee demographics in Nairobi. However, their settlement patterns differ significantly. A large number of Somali refugees navigate their way to the Eastleigh district of Nairobi, popularly portrayed as 'Little Mogadishu' (Carrier, 2016).
The presence of a large Somali-Kenyan population has offered Somali refugees a space for 'segmented assimilation' (Lindley, 2011). Taking advantage of ethnic and religious bonds, many have found income-generating opportunities in Eastleigh. According to a series of focus group discussions with Somali refugees, the most popular activities are selling clothing, vending cooked meals such as pasta and rice, and a range of casual labour roles, including working as porters, shopkeepers, and waiters/waitresses. Almost all of these refugees engage in these endeavours only within Eastleigh, sometimes employed informally by Somali-Kenyan residents living in the neighbourhood.
In addition, some Somali refugees have embarked on joint business ventures with Somali-Kenyans. As reported above, Somali refugees are particularly disadvantaged by securitisation policies and police harassment, so collaborative partnerships with Somali-Kenyans offer a 'cover'. Shukuria, a Somali refugee woman in her thirties who jointly owns a wholesale clothing business in Eastleigh with her Somali-Kenyan partner, offered a good example: Interviewer: Where did you meet your partner? Shukuria: She is my neighbour. We met each other in Eastleigh. Interviewer: Why did you start this joint business? Shukuria: We used to hawk in Nairobi, but hawking is now prohibited. So we decided to start this shop together. I trust her a lot though we are not related. She has been my longtime friend. Interviewer: How much money did you invest? Shukuria: In total we put USD 3,000. It is a 50-50 ownership. Interviewer: How do you share profits? Shukuria: In proportion to shareholding.
Most of the large-scale Somali refugee businesses highlighted in interviews relied on joint ownership with Somali-Kenyans and were registered with the City Council of Nairobi. They all emphasised the advantage of having a Kenyan 'front side' of the business, avoiding cumbersome documentation issues and other procedures imposed on refugees while decreasing vulnerability to police harassment. In the case of Somali refugees in the capital, their ethnic bond can attenuate somewhat the impacts of refugeehood on their economic activities.
In contrast to the high concentration of Somali refugees in Eastleigh, the approximately 18,000 refugees from the DRC are scattered throughout the city in suburbs such as Githrai, Kasarani, Kayole, and Umoja. The nature of relations between Congolese refugees and their neighbouring Kenyan hosts is quite different from those of Somali refugees, principally because Congolese refugees do not have ethnic counterparts among Kenya's citizenry. Instead, they take advantage of their command of Swahili-the national language of Kenya-to find traction in the urban economies of Nairobi. What is more, they can 'blend in' more easily, as they are not profiled in the same way as Somalis.
Interviews with a number of Kenyans in areas where large numbers of Congolese refugees reside revealed that perceptions of them are generally positive and cordial. A Kenyan elder, who holds a local government position and has lived in Kayole for more than 30 years, said: Our place [Kayole] has hosted so many refugees for many years. . . . Our relationship is generally cordial. Previously we had some conflicts between refugees and hosts, but there is no longer such tension. We have no problem hosting refugees.
Given their generally peaceful relationships with Kenyan hosts, as well as their language advantage, most of these Congolese refugees find casual labour among their Kenyan neighbours. Benjamin, one such refugee, commented: I often go to Kenyan construction companies in search of available work. If I find some, I make KES 500 [about USD 5] per day. Also, I do many other ad hoc tasks for Kenyans like working as a security guard and taking care of their animals.
As Congolese refugees tend to accept less payment as compared to nationals, Kenyan business owners often prefer to hire them for casual labour tasks. Many Congolese refugee interviewees appreciated the economic opportunities provided by Kenyans, but they were dissatisfied with their lower salary. As with incentive workers in Kakuma Camp, this is a frustrating consequence of their refugeehood.
A majority of Somali and Congolese refugees are making their living in the informal sector of Nairobi. This is not surprising since a large number of Kenyan people are also absorbed into the capital's informal economy owing to high unemployment and a lack of formal job opportunities (Hope, 2014). An important difference, though, is that refugees are largely forced to eke out a living in the informal economy and, in the case of Congolese refugees, frequently to accept payment at lower than ordinary rates. A representative of the Congolese refugee community in Kitengela summed it up as follows: 'we accept cheaper employment because we have no alternative'.

Comparative perspectives on refugee livelihoods in Kenya
This section draws on the findings presented above to provide a comparative analysis of refugees' livelihoods in different contexts of refugeehood. Furthermore, it points up some pragmatic implications for policymakers engaged in refugee assistance.
First, refugeehood requires that people make a living under complex regulatory assemblages, involving multiple institutional actors. Even though Kakuma Camp and the city of Nairobi come under the same refugee law, myriad stakeholders are active in shaping refugees' economic spaces in each site, with repercussions for their daily livelihoods. While the national government enforces restrictions on refugees' right to work and to move freely within Kakuma, UNHCR, as another camp authority, provides a hybrid of formal and informal economic space for refugees with compromised financial remuneration, as demonstrated by the example of incentive workers. Alongside these formal regulatory bodies, even Turkana host populations impose local restrictions on refugees' economic activities by prohibiting them from selling charcoal, firewood, and meat. In Nairobi, beyond the 'exceptional space' of the camp, the influence of UNHCR is significantly diminished. Instead, the national and local government play a key role in formulating refugees' livelihood strategies. Again, however, despite the encampment policy of the Government of Kenya, local host communities are accommodating refugees and the City Council of Nairobi offers registration of the business activities of refugees.
Second, relatedly, the impact of refugeehood on the livelihoods of refugees differs depending on whether one seeks employment or self-employment. Refugee incentive workers in Kakuma Camp cannot enjoy the full benefits of formal employment offered to Kenyan nationals by UNHCR or its partners. Yet, the County Government of Turkana is attempting to 'formalise' refugee businesses to derive tax revenue. Similarly, in Nairobi, although the national authority rarely issues work permits to refugees, a number of refugee businesses are officially registered with the City Council of Nairobi. Hence, these enterprises enjoy formal endorsement by the municipal government, regardless of whether or not refugees have proper status. Such different treatments show that the application of restrictions by authorities vary in degree depending on the potential benefits that can be derived from the presence of refugees. Employment takes jobs away from Kenyan nationals, which may be unpopular among national and local authorities. But self-employment presents financial incentives, through the collection of registration fees, and might even create employment opportunities for Kenyans.
Third, aside from the multiple regulation standards, refugeehood also entails specific vulnerabilities to unofficial forms of discrimination and harassment. Whereas the practice of bribery owing to police corruption is a feature of the Kenyan economy affecting many (Hope, 2014), refugees, especially those from Somalia, are much more susceptible to police abuse owing to their precarious legal status and limited access to justice in Nairobi. This particular form of harassment also affects nonrefugee Somali populations, but Somali-Kenyans do have some recourse thanks to their citizenship. While harassment by authorities is a common challenge to urban refugees, police abuse is also a problem for refugees living in Kakuma Camp, regardless of nationality. In both settings, bribes and 'fees' charged by the local authorities also generate additional transaction costs for refugees' livelihoods.
Fourth, in addition to an evaluation of Kakuma Camp and Nairobi, this study allows for a comparison across different nationalities of refugees living within the same site, revealing differences in their livelihood strategies and relations with host communities. The contrastive patterns of settlement between Somali and Congolese refugees in Nairobi highlight different kinds of networks established with Kenyan hosts and how these contacts serve as 'mediating factors' (Zetter and Ruaudel, 2016) in their pursuit of livelihoods. The existing Somali-Kenyan community in Eastleigh provides Somali refugees with ethnic protection and access to a 'safe space' to run commercial activities. Congolese refugees cannot rely on ethnic community networks to the same extent, but their competency in the Swahili language permits them to generate connections with ethnically unrelated Kenyan nationals and explore casual employment opportunities. However, although these networks can facilitate access to informal economic opportunities by Congolese refugees, they do not fully eradicate the impacts of refugeehood, as indicated by the lower wages paid to them in Nairobi.
Fifth, these findings serve as practical lessons for aid practitioners and policymakers working to strengthen refugee livelihoods. While the nature of the support of aid organisations tends to be the provision of technical assistance, typically vocational training, these actors need to take into account the economic obstacles associated with refugeehood. Not considering these refugee-specific handicaps means that livelihood support alone will be ineffective. In addition, refugee-supporting organisations developing assistance plans should take into account the diversity of refugees' economic strategies. Currently, they often do not capture adequately these differences in their refugee livelihood programmes. As shown in Nairobi, the programmatic needs of Congolese and Somali refugees may differ significantly owing to the disparity in their economic strategies and relationships with host populations.
Sixth, refugeehood entails persistent forms of uncertainty and unpredictability, which was more observable in the camp context. The closure announcement by the Government of Kenya demonstrated to camp residents the precariousness of their life in exile, and left many questioning whether they should invest in their camp businesses, given the fear that they may be forced to pack their things and leave. Refugees' suspicion of the Kenyan administration, coupled with a lack of access to information concerning their future, can affect refugees' future economic planning and decision-making.

Conclusion
This papers draws on comparative analysis to offer an overview of the diverse economic activities employed by refugees in both the camp and urban setting in Kenya, highlighting a number of institutional factors that distinguish the economic lives of refugees from those of members of the host population. Some of the livelihood issues addressed are not applicable exclusively to refugees, yet they constitute acutely inhibitive handicaps to their economic lives. The empirical findings suggest that while some Kenyans are encumbered by challenges comparable to those facing refugees, the myriad political, legal, and policy-related factors that characterise refugeehood are nonetheless particular in their impacts on livelihoods.
The findings also point to the need to focus a political economy lens on refugee livelihoods, demonstrating how they are constrained, mediated, or enabled by a range of variables and numerous formal and informal actors according to their agendas and motives. In Kenya, while security is a key priority guiding the national government's regulatory policies, adherence to these regulations by other actors varies, and they may even pursue conflictual strategies, such as collecting official licencing fees from refugee businesses.
Aside from economic interests, the economic spaces within which refugees implement livelihoods are also shaped by the historical contexts of marginalised host populations, as illustrated by the example of Turkana locals at Kakuma Camp. Similarly, refugees themselves attempt to contest restrictions by capitalising on their language skills, ethnic and social networks, or mobility to engage in their livelihoods, or in some cases, although not ideal, by offering their labour at a reduced rate of income.
It is well-known that refugee livelihoods, especially those that centre on informal economic activities, are often tolerated despite numerous regulatory constraints, but relatively little is known about the mechanisms by which such tolerance is generated. A political economy approach can address an assortment of aspects that influence the economic spaces of tolerance and provide a deeper understanding of mediating factors that can augment or dilute the ramifications of refugeehood.
Looking forward, this paper suggests more research on the consequences of different regulatory settings on refugees' livelihoods. As emphasised, forced migration scholarship lacks comparative empirical analysis of the livelihoods of refugees and their local host communities across different countries. The collection of additional empirical data in diverse environments would contribute to a more nuanced and systematic understanding of refugeehood and the ways that it can affect the economic lives of refugees.
Correspondence Naohiko Omata, Associate Professor, Refugee Studies Centre, Department of International Development, University of Oxford, 3 Mansfield Road, Oxford OX1 3TB, United Kingdom. E-mail: naohiko.omata@qeh.ox.ac.uk Endnotes 1 Refugees are given special permission if they are invited for resettlement interviews, need specialised medical treatment, wish to pursue further education, or face serious security threats in the camps. A considerable number of refugees in Nairobi do not have such permission, however, and thus reside unauthorised in the city.